Information Retrieval in Digital Environments [1 ed.] 9781119015147, 9781848216983

Citation preview

W698-Dinet.qxp_Layout 1 01/07/2014 08:54 Page 1

FOCUS SERIES in INFORMATION SYSTEMS, WEB AND PERVASIVE COMPUTING

Jérôme Dinet is a Psychologist and Assistant Professor in cognitive psychology and ergonomics at the University of Lorraine in France.

Information Retrieval in Digital Environments

The author of this book presents a summary of work undertaken over several years relative to the behaviors and cognitive processes involved in information retrieval in digital environments. He presents several examples of theoretical models and studies to better understand the difficulties, behaviors and strategies of individuals searching for information in digital environments.

Jérôme Dinet

Information retrieval is a central and essential activity. It is indeed difficult to find a human activity that does not need to retrieve information in an environment which is often increasingly digital: moving and navigating, learning, having fun, communicating, informing, making a decision, etc. Most human activities are intimately linked to our ability to search quickly and effectively for relevant information, the stakes are sometimes extremely important: passing an exam, voting, finding a job, remaining autonomous, being socially connected, developing a critical spirit, or simply surviving.

FOCUS INFORMATION SYSTEMS, WEB AND PERVASIVE COMPUTING SERIES

Information Retrieval in Digital Environments Jérôme Dinet

www.iste.co.uk

Z(7ib8e8-CBGJID(

Information Retrieval in Digital Environments

FOCUS SERIES Series Editor Fabrice Papy

Information Retrieval in Digital Environments

Jérôme Dinet

First published 2014 in Great Britain and the United States by ISTE Ltd and John Wiley & Sons, Inc.

Apart from any fair dealing for the purposes of research or private study, or criticism or review, as permitted under the Copyright, Designs and Patents Act 1988, this publication may only be reproduced, stored or transmitted, in any form or by any means, with the prior permission in writing of the publishers, or in the case of reprographic reproduction in accordance with the terms and licenses issued by the CLA. Enquiries concerning reproduction outside these terms should be sent to the publishers at the undermentioned address: ISTE Ltd 27-37 St George’s Road London SW19 4EU UK

John Wiley & Sons, Inc. 111 River Street Hoboken, NJ 07030 USA

www.iste.co.uk

www.wiley.com

© ISTE Ltd 2014 The rights of Jérôme Dinet to be identified as the author of this work have been asserted by him in accordance with the Copyright, Designs and Patents Act 1988. Library of Congress Control Number: 2014941992 British Library Cataloguing-in-Publication Data A CIP record for this book is available from the British Library ISSN 2051-2481 (Print) ISSN 2051-249X (Online) ISBN 978-1-84821-698-3

Printed and bound in Great Britain by CPI Group (UK) Ltd., Croydon, Surrey CR0 4YY

Contents

CHAPTER 1. INFORMATION RETRIEVAL IN DIGITAL ENVIRONMENTS: DEBATE AND SCIENTIFIC DIRECTIONS . . . 1.1. Information retrieval, current and future challenges . . . . . . . . . . . . . . . . . . . . 1.2. What are we talking about? . . . . . . . . . . 1.3. Interaction and navigation at the heart of information retrieval . . . . . . . . . . . . 1.4. Why should we be interested in information retrieval? . . . . . . . . . . . . . . . . . 1.4.1. Economy: maximize profitability and minimize risks. . . . . . . . . . . . . . . . . . . . . 1.4.2. Information technology: mathematical concepts of the relevance of information. . . . 1.4.3. Robotics: improving movements and interactions. . . . . . . . . . . . . . . . . . . . . . .

1

........ ........

1 3

........

7

........

9

........

10

........

12

........

14

CHAPTER 2. CONCEPTUAL AND METHODOLOGICAL APPROACHES TO INFORMATION RETRIEVAL IN DIGITAL ENVIRONMENTS . . . . . . . . . . . . . . . . . . . . . . . .

19

2.1. The approaches of information sciences: the precursors . . . . . . . . . . . . . . . . . . . . . . . 2.2. The Marchionini sequential iterative model 2.3. The holistic model of Kuhlthau . . . . . . . . . 2.4. The first studies of psychology and cognitive ergonomics . . . . . . . . . . . . . . . . . . .

....... ....... .......

19 21 23

.......

26

vi

Information Retrieval in Digital Environments

2.5. The cyclic model of David, Song, Hayes and Fredin . . . . . . . . . . . . . . . . . . . . . 2.6. The skills-centered model of Brand-Gruwel . 2.7. Kitajima’s predictive model . . . . . . . . . . . 2.8. The hyper-specialized model of Sharit, Hernandez, Czaja and Pirolli . . . . . . . . . . . . . 2.9. The Landscape Model “diversion” by Dinet .

....... ....... .......

31 33 36

....... .......

39 42

CHAPTER 3. INFORMATION RETRIEVAL: PSYCHOERGONOMIC APPROACH . . . . . . . . . . . . . . . . . . . . . . . . .

49

3.1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . 3.2. Identifying difficulties in modifying interfaces . 3.2.1. Hierarchical task analysis . . . . . . . . . . . . 3.2.2. Analysis of the end users’ behavior . . . . . . 3.2.3. Implications for the (re)design of interfaces 3.3. Anticipating the needs of users . . . . . . . . . . . 3.3.1. “If we built it, they will come” . . . . . . . . . 3.3.2. The analysis of users’ expectations and behaviors . . . . . . . . . . . . . . . . . . . . . . . . 3.3.3. Prospective ergonomics and technological innovation. . . . . . . . . . . . . . . . . . . . . . . . . . . 3.3.4. Anticipating and understanding the needs of users: the method of staff made up of community experts . . . . . . . . . . . . . . . . . . . . . 3.3.5. An example of application of the method of staff made up of community experts . . . . . . . . 3.4. The motor dimension . . . . . . . . . . . . . . . . . . 3.4.1. Motor ability and information retrieval in digital environments. . . . . . . . . . . . . . . . . . . . 3.4.2. Toward a lexicon of intuitive gestures . . . . 3.5. The social dimension and collaborative . . . . . . 3.5.1. From individual research to collaborative information retrieval . . . . . . . . . . . . . . . . . . . 3.5.2. Benefits and limitations of collaborative information retrieval . . . . . . . . . . . . . . . . . . . 3.6. Impact of emotional ties between collaborators 3.6.1. Ties between collaborators and impact on information retrieval . . . . . . . . . . . . . . . . . . . 3.6.2. “RCI-Web”: software to assist information retrieval . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

49 51 52 55 61 63 64

.....

65

.....

70

.....

73

..... .....

76 79

..... ..... .....

79 85 88

.....

89

..... .....

90 92

.....

94

.....

97

Contents

3.7. The cultural dimension . . . . . . . . . . . . . . . . . . . 3.7.1. About the importance of the home page . . . . . . 3.7.2. Culture and design of Websites’ home pages: an ergonomic inspection . . . . . . . . . . . . . . . . . . . . . . 3.7.3. Information retrieval culture and behavior navigation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.8. The visual exploration strategies. . . . . . . . . . . . . 3.8.1. Impact of the typographical marking (bottom-up approach) . . . . . . . . . . . . . . . . . . . . . . 3.8.2. Impact of the mental model (top-down approach) . . . . . . . . . . . . . . . . . . . . . . .

vii

.. ..

102 102

..

105

.. ..

107 109

..

112

..

117

CONCLUSION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

123

BIBLIOGRAPHY . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

125

INDEX . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

165

1 Information Retrieval in Digital Environments: Debate and Scientific Directions

1.1. Information challenges

retrieval,

current

and

future

Information retrieval is a central and essential activity. It is indeed difficult to find a human activity that does not need to retrieve information in an often increasingly digital environment: moving and navigating, learning, having fun, communicating, informing, making a decision, etc. Most human activities are intimately linked with our ability to search quickly and effectively for relevant information, the stakes are sometimes extremely important: passing an exam, voting, finding a job, remaining autonomous, being socially connected, developing a critical spirit or simply surviving. From the psychological point of view, the activity of information retrieval presents several characteristics that make them both unique, complex and fascinating [DIN 12a]: – Information retrieval in digital environments is a necessity in a growing number of human activities. Daily, we must search for information (an address, a phone number, a time slot, a name, etc.) in various digital environments of

2

Information Retrieval in Digital Environments

constant evolution (GPS, mobile phone, Website, electronic terminal, etc.). – Information retrieval is a composite activity, which typically requires reading, memorizing, writing, taking decisions, etc., these activities being cognitively complex and corresponding to areas of scientific research on their own. – Information retrieval is a dynamic activity in the sense that the environment in which the activity is carried out evolves independently of the actions of the user (the same query in a search engine on the Web gives different results in an interval of a few seconds). – Information retrieval is an iterative process, since the behavior and knowledge of the user are constantly changing at the pace of the information returned by the information system. In 1998, the first Google index had 26 million Web pages. In 2008, Googles engineers counted a trillion Web pages (or a thousand billion). Since then, the number of pages has no longer been counted, and only the number of Internet sites is recognized. In June 2011, there were 346,004,403 Internet sites (source: Netcraft.com). It is virtually impossible to know with accuracy the number of Internet sites and/or Web pages (sites protected by password, military sites, invisible Web, etc.). However, this is no matter, the only important data are the incredible progression of this mass of information, which everyone can easily access today. But, more than the mass of information, it is especially the relationship between humans and information, that has significantly evolved and is still evolving.

Debate and Scientific Directions

3

1.2. What are we talking about? The term “information retrieval” (“recherche d’information” in French) includes both the scientific field and the related human activity. Some professional Frenchspeaking organizations, such as the French association of information and documentation professionals (ADBS 2004) or even the AFNOR (French national organization for standardization 1993), proposed distinguishing information retrieval from information research: – Information retrieval (recherche d’information in French) is the “totality of methods, procedures and techniques that allows, based on search criteria specific to the user, the selection of information in one or more repository of documents more or less structured” [AFN 93]. – Information research (recherche d’information in French) is the “totality of methods, procedures and techniques that aim at retrieving the relevant information from a document or a set of documents” [AFN 93]. The difference remains subtle and above all too superficial regarding the place of the end user in the activity. Because the English lexicon concerning the areas of computing and new technologies is more extended and more precise, it is logical to find more words (and therefore more accuracy) on topics related to these areas, such as information retrieval. In the English language, several specific terms coexist for mentioning the scientific fields related to information retrieval: – Information retrieval: generally noted “IR”, these terms refer to the scientific field concerned with everything that falls under the search for documents or information, research in documents and whatever environments are considered (physical environment, off-line digital environment, such as CD-Roms, digital library, Internet, etc.). IR is a multidisciplinary field primarily involving disciplines such as computer science, mathematics,

4

Information Retrieval in Digital Environments

information science and communication, psychology or even economics. Initially, one of the first objectives of researchers in the field of IR was the creation and development of technical systems and interfaces allowing quick access to relevant information in the increasingly large and more dynamic informational bodies. – Information behavior: this refers to all of the human behavior related to all sources and channels of information (television, telephone, paper documents, Internet, face-toface communication, etc.). In this case, human behaviors are not necessarily motivated by a need for information. For example, studies interested in the impact of exposure to advertisements (i.e. the “passive” consumption of information) are part of this scientific sub-domain. – Information-seeking behavior: this refers to all of the human behaviors explicitly oriented toward obtaining information, regardless of the media (e.g. textbooks paper, magazines and Websites). In this case, a need for information is, therefore, at the basis of the studied behaviors, whether this need is conscious or not. – Information-searching behavior: this refers to human behavior directly concerned with information retrieval, from a sensorimotor point of view (e.g. using the mouse or the keyboard) or intellectual (e.g. evaluation of the relevance of information). – Information use behavior: this corresponds to human behavior involved in the processing and the assimilation of information found to knowledge already stored in memory. Here again, the sensorimotor aspects are sometimes studied (e.g. note-taking and annotation of texts) as well as the psychological aspects (e.g. memorization of information found). If it is difficult to define with precision the exact shape of the different sub-areas related to IR, the task becomes even more difficult in the case of human–machine interactions (HMIs).

Debate and Scientific Directions

5

In order to define the different scientific sub-fields related to HMI, Wania, Atwood and McCain [WAN 06] have investigated the “genealogy” of different productions (articles and conferences proceedings). More specifically, these authors have studied the citations and co-citations in various publications based on the 64 authors most frequently quoted in HMI. After a review of bibliographic databases for the Institute of Information Science (ISI), Wania, Atwood and McCain [WAN 06] were able to identify the relationships existing between the articles; these relations take the form of citations and co-citations of the authors of reference. In addition, seven distinct groups could be identified in the studies conducted on HMI (Figure 1.1).

Figure 1.1. Affiliations between the articles in HMI, according to [WAN 06]

6

Information Retrieval in Digital Environments

The two axes along which the seven sub-areas of HMI are divided and have been translated by [WAN 06] in the following manner: – The horizontal axis distinguishes the studies according to the main center of interest, from technical systems to the human individual. – The vertical axis distinguishes the studies according to the degree of involvement of end users in the design and/or evaluation process. For example, the work of Jacob Nielsen is located at the right end of the horizontal axis indicating clearly that the end user is the main factor of interest; however, since his work rarely results in experiments or empirical studies, this author is located in the lower half due to the involvement of individuals in his research. This classification of studies in HMI according to these two axes allows us to address differently the conceptual and methodological frameworks in HMI. In fact, generally, research about HMI is presented by distinguishing works at the design level (i.e. conception) from works at the assessment level. However, it is difficult to distinguish studies according to this dichotomy, since most of the studies address the two dimensions (to various degrees of course). Thus, not only does information retrieval cover a multitude of studies, sometimes very different, but works on HMI also cover a plurality of disparate studies. It is appropriate to add a third difficulty in the apprehension of scientific studies interested in human behavior in information retrieval in digital environments: the level of granularity chosen by the researcher. According to Fidel et al., [FID 04], the “size of the magnifying glass” used by the ergonomist researcher or practitioner determines its theoretical and methodological choice. And this “size of magnifying glass” is itself determined by the questions underlying the research or the intervention (Figure 1.2):

Debate and Scientific Directions

7

Impact of organizational factors? Strategies of visual exploration? Task analysis? etc.

Figure 1.2. The concentric approach of Fidel et al. [FID 04]

1.3. Interaction and navigation at the heart of information retrieval As we will see in detail in the following chapters, the individual has been quickly (re)placed at the center of interests of scientific studies focusing on the activity of information retrieval (for a summary, see: [DIN 12]). According to Kelly [KEL 07], most of these studies can be located on a continuum (Figure 1.3). At one end of this continuum, studies are mainly focused on systems and technical devices related to information retrieval; at the other end of this axis are found studies mainly focused on behaviors and human factors. According to the approach

8

Information Retrieval in Digital Environments

considered, the theoretical frameworks and methodologies used differ.

Figure 1.3. Scientific research on information retrieval according to the continuum proposed by Kelly [KEL 07]

The significance of Kellys design [KEL 07] is to consider that the balance and/or break between these two extremes is the interaction: indeed, the human dimensions appear from the moment where the interaction becomes the center of interest (“IIR” for “Interaction Information Retrieval”). Some authors also propose a historical reading of the evolution of theoretical frameworks and methodological studies, for example, noting the evolution of topics discussed during important international scientific events [BEA 96, OVE 01]. The links between information retrieval and navigation are so narrow that the terms are often used interchangeably. From a formal point of view, navigation encompasses the whole range of techniques and procedures that allow, on the one hand, finding the position of a moving object in relation to a reference system or a fixed point and, on the other hand, calculating and measuring the route of a given point to another point. The moving object can be a human individual or an object, depending on the case.

Debate and Scientific Directions

9

For some, navigation is one of the elements of information retrieval: for example, it is by navigating from one Web page to another Web page that information is retrieved. For others, it is information retrieval that is one of the navigation elements: it is by searching for and retrieving the relevant information in the environment (physical or digital) that we can navigate. This design corresponds to that presented in studies about driving cars, piloting aircraft, operating machinery or even the movement of pedestrians or users in physical or digital environments. In all these areas, navigation is the final objective and information retrieval is one of the processes and/or one of the underlying phases to achieve the objective. We can point out that many metaphors borrowed from aerial or maritime navigation (in physical environments) are always applied to the design of digital environments [AHU 01, CHE 11, EDW 89, MCD 98, SMI 96]: the navigation bar, the portal, the compass, etc. 1.4. Why should we be interested in information retrieval? Ergonomic psychology and ergonomics are not the only disciplines in the field of social and human sciences (SHS) that are interested in information retrieval. The activity of information retrieval is also the subject of many studies in the disciplines of management sciences or even engineering. Although that may seem very distant from the approaches of psychology and ergonomics, we present very briefly the contributions of three disciplines whose knowledge is related to the activity of information retrieval in digital environments: economics, computer science and robotics. Most importantly, we present why these disciplines are very interested in the activity of information retrieval.

10

Information Retrieval in Digital Environments

1.4.1. Economy: maximize profitability and minimize risks In the business world, information retrieval is a crucial activity. Indeed, corporate decision-making can only be done after a substantial work of information retrieval. The increase of pressure related to competition, time constraints, information load as well as the geographical dispersion of information have made the phases of information retrieval even more dominant [DAV 05, KAI 91]. In a survey conducted in 2009 by an important group at the junction of marketing and technology (Delphi Group: http://www.delphigroup.com), 1,030 employees of 15 US companies of medium size responded that they spent more than a third of their working time searching for information. Yet, these employees did not belong to a document and/or information survey service. These results coincide with those obtained by McDermott [MCD 05], which show that 38% of employees of large companies spend the majority of their time searching for information. The importance of information retrieval not only affects large firms: it appears that professionals from other sectors spend increasingly more time searching for information, such as teachers, health professionals or even lawyers [COO 06]. According to an internal report of Google© of 2008, the majority of employees of the major groups are now “Information Workers”, insisting on the importance of activities related to information in some sectors, such as banking or pharmaceutical businesses. Still, according to this report, only a quarter of information searches conducted within large corporations would be successful (“Successful Search”) and almost half of the activities would be nonproductive (and therefore, non-profitable), since they would recreate existing information, convert information into other formats or even simply collect without really analyzing

Debate and Scientific Directions

11

documents and/or information. Of course, the results of this report are to be taken with caution. Nevertheless, in 1964, Drucker had already predicted that the majority of employees would have a function of knowledge managermanipulator (“Knowledge Management”). If information retrieval interests businesses and professionals, it is because it is an activity that represents a huge cost in many ways: – Overall, the time spent searching for information is increasingly important [COO 06, MCD 05]. – The time spent searching for unnecessary information is also increasingly important. Even if it is difficult to make accurate evaluations, some economic agencies (e.g. IDC: http://www.idc.com) estimate that 90% of the documents and information produced by enterprises could already exist elsewhere. – The cost generated by “bad” information retrieval can be considerable, the cost can be material, human or both (accidents, incidents, brand image violation, commercial failure, errors on the target audience, infringement of industrial property, etc.). In this case, information retrieval aims at minimizing the risk linked with the decision-making processes important for businesses [BON 06, ZHA 06]. Among the many sources of information they have, employees overwhelmingly use Internet and especially the Web to search for information. Even if the information is sought in other media (e.g. off-line documentary databases, business applications and e-mails), these are almost exclusively digital environments. Also, it is logical that economists, entrepreneurs and managers have been closely involved in the activity of information retrieval in digital environments. Finally, do not forget that more and more human resource officials admit searching for information on

12

Information Retrieval in Digital Environments

open social networks in order to obtain the maximum of information on candidates. 1.4.2. Information technology: mathematical concepts of the relevance of information When referring to the activity of information retrieval, information technology is sometimes viewed as the most relevant scientific discipline to study. As a matter of fact, some do not hesitate to consider that the interest for the information retrieval activity was born with information technology: “The field of information retrieval dates back to the early 1950s, shortly after the invention of computers […]. The name “information retrieval” was given by Calvin N. Mooers in 1948 for the first time when he was working on his Master’s thesis […]. It was during this period that the domain of IR was born.” [NIE 11]. “Information retrieval (IR) was founded shortly after the advent of the first computers, and is certainly the oldest application of computer technology to access electronic documents.” [CHI 07]. It is true that IT had to quickly address the problem of indexing documents and information, such that humans could, on the one hand, store them and, on the other hand, find them quickly. But, concerning the relationships between IT and information retrieval, there are usually two levels of confusion: – The first level of confusion exists between, on the one hand, the activity of information retrieval and, on the other hand, the domain of information research. In fact, the domain of information research (usually noted “IR”) is the scientific field that encompasses the theories and methods used to design tools (mathematical, technological, etc.)

Debate and Scientific Directions

13

dedicated to the activity of information retrieval, typically to facilitate it. – The second level of confusion exists between, on the one hand, the activity of information retrieval as an intellectual approach and, on the other hand, the tools used to achieve said activity. Only a small part of the work done in IT related to information retrieval aims to design and/or evaluate interfaces and environments designed to help humans (particularly in HMI). Indeed, the bulk of the work done in IT is not interested in interfaces, but studies systems supporting information retrieval. This branch of the work relies primarily on the use of mathematical theories and aims to propose extrapolable models in information retrieval [KUR 04]: – The set approach (based in a large part on Boolean logic) considers that all data (documents and information) maintain links of union, of intersection, and can therefore be found by means of logical operations. The direct translation of this approach is the presence and/or explicit or implicit use of Boolean operators to query databases and databanks (“AND”, “OR”, “EXCEPT”, etc.). We can point out that if these Boolean operators were present on many interfaces of research a few years ago, today they are mainly present on tools aimed at specific audiences (e.g. young users and researchers) and concern only documentary research. – The algebraic approach considers that all the data belong to the same vector space. – The probabilistic approach is primarily interested in the concept of relevance and has largely inspired the work of researchers on information science and communication, such as [GAB 99]. In other words, these works are essentially of fundamental research and are relatively distant from

14

Information Retrieval in Digital Environments

problems related to users, to usages and interfaces. Nevertheless, behind a technological device linked with information retrieval there is always a theoretical mathematical design, certainly unknown to most users, but which inevitably determines part of the related interactions and usages. 1.4.3. Robotics: improving movements and interactions For cyberneticists and roboticists, the activity of information retrieval is central, since it is intimately linked to the locomotor movements of robots. A robot is “an engineered machine that senses, thinks, and acts”; [LIN 11]. A robot is, therefore, a machine that has at least three components: – materials and equipment, which enable it to receive information from its near or distant environment (microphones, cameras, external and internal sensors, buttons, etc.); – effectors, which allow it to act in and/or on its environment through motor behaviors (arms, legs, tracks, etc.) and verbal and/or non-verbal language behaviors (synthesized speech, postural attitude, etc.); – a central decision-making system, which saves, processes and manipulates inbound and outbound data. To be qualified as autonomous, a robot must be able to effectively search the relevant information with a view to making “good” decisions mainly in the area of locomotor movements. Three processes are then necessary [BAL 99, LEV 90]: – The robot must possess and/or construct a representation of the environment in which it evolves (“Maplearning”). – It must be able to position itself in this representation (“Localization”).

Debate and Scientific Directions

15

– It must be capable of planning a move within this representation before efficiently moving (“Path-planning”). However, retrieving information in its environment is essential during these three phases. Indeed, to move, a robot must collect and use allothetic indices (i.e. collected in its environment: smell and detection of objects by sonar) or idiothetic indices (i.e. internal: speed and acceleration). In psychology, these two types of indices correspond to the exteroceptive and proprioceptive indices. It is interesting to note that works of ethology have largely inspired and still inspire models of robotics when the issue of information retrieval inside its environment to move is mentioned [MEY 03, MIT 80]. The analysis of behaviors and human cognitive processes related to information retrieval to implement them in robots is a concern in a growing number of studies of robotics. For example, on the basis of ethnographic observations and audio-video recordings from older residents, Quan et al., [QUA 11] have developed a robot capable of searching for, detecting and taking into account the different information presented in the environment (eye orientations, postural attitudes, movements, etc.). An experiment conducted on 21 participants has shown that the individuals, overall, felt that “the robot understood their request”. Also, the opinions of the participants in the study were overwhelmingly favorable and positive with respect to this robot capable of perceiving these sometimes subtle non-verbal clues (e.g. eye blinking and light smiles) that any human perceives naturally. It is in the area of information retrieval with special focus on navigation that studies of robotics are the most numerous. Many research laboratories are interested in the movements of robots in physical environments, in order to make them more efficient and reliable once placed in buildings open to public or private homes. Alongside the

16

Information Retrieval in Digital Environments

mechanical and computing aspects, some authors have found that the way in which a robot approaches a human was decisive in the perception and acceptance of this robot by this human. Concretely, Qian et al. [QIA 10] have determined four rules that a robot must observe during its physical interactions with a human individual, if we want this person to accept and trust this robot: – proxemics: the robot must remain relatively far away, for security (risk of collision) and psychological (respect for the “vital space” of the human) reasons; – visibility: the robot must know how to draw attention discretely but effectively to avoid surprising the individual; – compliance with passing conventions, determined (from the right or from the left);

culturally

– human priority: whatever the context and the situation, the human individual must always be the priority during its movement. Also, the information that a robot must search for must no longer only be about the spatial information present in its environment (e.g. potential obstacles) but also finer information so that the interaction is accepted by the human individual. Furthermore, the first three rules (proxemics, visibility and passing conventions) correspond to norms shared socially and that apply to interactions between humans. On the basis of these rules, the authors have developed and tested a robot that should move in physical environments with a human individual. Not only does their prototype give excellent results in terms of locomotor movements, but it also gives satisfaction to the participants in the study, since the latter declare to have appreciated the “courtesy” of the robot during their meetings in the environment. In other words, it is an implementation of a human psychological characteristic (courtesy), which positively influences the opinions of users, this courtesy only

Debate and Scientific Directions

17

being possible if the right information is collected from the environment. In other words, because we are moving from industrial static, robots, performing automated and repetitive tasks to robots coexisting alongside human individuals in their physical environments (home, shopping center, hospital, etc.), taking into account the psychological parameters related to information retrieval has become paramount in robotics [DIN 12b]. Today, studies aim to design robots capable of searching for and collecting information useful for their movements both from their direct physical environment and from interpreting the behavior of human individuals encountered during their movement. In other words, we see that information retrieval is at the heart of the concerns of many disciplines and of several sectors of activity and for different reasons (e.g. minimizing risks, improving the performance of automatic or robotic systems). If these concerns can lead to very important scientific breakthroughs, it is the study of information retrieval from the point of view of the end user that seems to us the most captivating. Furthermore, although “technical or technological barriers” still exist, human and psychological factors essentially explain most of the behaviors, difficulties and emotions of individuals searching for information in digital environments. It is, therefore, an approach resolutely focused on the human dimension that we are defending here.

2 Conceptual and Methodological Approaches to Information Retrieval in Digital Environments

2.1. The approaches of information sciences: the precursors Historically, information and communication sciences (ICS) were among the first to focus on the activity of information retrieval. One of the concepts that has most benefited from the inputs of ICS is relevance. According to Gabrielli and Mizzaro [GAB 99], the degree of relevance of the information depends on the interaction between three dimensions. Graphically, the degree of relevance of information can be positioned according to these three axes (Figure 2.1): – the representation of the need for information of the end user (axis “InfNeeds”): considering information retrieval as a problem solving activity, the closer the information found is to the need of the user (i.e. meets his expectations) the more significant the relevance. According to [GAB 99], the need for information can be real (“RIN”), perceived (“PIN”), translated into natural language (“EIN”) and/or translated into a query (“END”);

20

Information Retrieval in Digital Environments

– information resource (axis “InfRes”): this axis distinguishes the different types of information associated with a document considering that some are more relevant: the document itself (“D”), a set of linked documents (“SD”), the metadata (“MD”) and a representative of the document such as a documentary note (“S”); – the context of the activity of information retrieval (axis “InfuseCo”): this axis is the most difficult to understand, since it encompasses the subject of the search (“To”), the task to be performed (“To + Ta”) and the set of physical and psychological attributes of the end user (“To + Ta + UA”) such as his domain expertise, his preferences or even his sensorimotor capacity.

Figure 2.1. The three dimensional space of relevance, according to Gabrielli and Mizzaro [GAB 99]

Even if Gabrielli and Mizzaro [GAB 99] asked 26 participants to corroborate the existence of the three axis of the space of relevance, their theoretical framework and its graphical representation are above all training tools destined

Conceptual and Methodological Approaches to Information Retrieval

21

to documentalists and librarians. In addition, despite their attempt to classify the impact factors of relevance, many dimensions remain very blurry and unusable. For example, the authors do not precisely define what the attributes of the individual noted “To + Ta + UA” are, which encompass the entire spectrum of physical, physiological and psychological characteristics. Although interesting, these studies, however, did not provide much information about the behavior, difficulties and the human strategies involved in information retrieval. However, other authors close to ICS proposed at an early stage to focus primarily on the end user. 2.2. The Marchionini sequential iterative model Based on a series of empirical studies [MAR 92, MAR 87, MAR 91, MAR 93, MAR 95], Marchionini proposed a model that distinguishes three types of information retrieval: directed, semi-directed and non-directed. During a directed search, the individual has a specific goal and his behavior is oriented to achieve this goal under a strict procedure. During a semi-directed search, the individual has only a rough idea of the information that he is seeking and his behavior will tend to find the information as close as possible to what he thinks he must obtain. During a non-directed search, the individual has no specific goal and will usually simply gather nectar (so-called “browsing” activity). Regardless of the type of information retrieval (directed, semi-directed and non-directed), the activity goes through eight successive stages according to Marchionini (1995; Figure 2.2: 1) recognition and acceptance that information is necessary to solve a problem; 2) definition and delimitation of the problem to be solved; 3) choice of a system and/or of a resource; 4) formulation of a query; 5) effective execution of the information retrieval; 6) review of the results and the

22

Information Retrieval in Digital Environments

information proposed by the system; 7) extraction of useful information from the information found; 8) evaluation of the result obtained and resumption of an earlier stage if this result is not satisfactory.

Figure 2.2. The eight sequential steps of information retrieval according to Marchionini [MAR 95]

The main contributions of the first version of Marchionini’s model [MAR 95] and the corrected versions [MAR 04, MAR 06] are the following: – to have laid down the activity of information retrieval as a natural human activity of fundamental importance, particularly when it is carried out in digital environments; – to have insisted on the fact that information retrieval is an interactive process involving on the one hand, a

Conceptual and Methodological Approaches to Information Retrieval

23

technological system and on the other hand, a human system via an interface within a shared information environment [KNI 08]. Also, regarding Marchionini [MAR 95, MAR 04, MAR 06], it is necessary to apprehend the information environment in which the two systems interact (technological and human) if we want to be able to understand the difficulties of users. 2.3. The holistic model of Kuhlthau Since the first versions of the ISP model (“Information Search Process”) published in 1985 and 1988, many changes and improvements have been made by Kuhlthau and her collaborators [KUH 99a, KUH 99b, KUH 01, KUH 04, KUH 07, SHA 02] to a model supposed to describe the “internal” behaviors and processes when searching for information. Since the first versions of the model, the intention of the principal author was to design a model capable of accounting for the total of the dimensions involved and/or affecting all stages of the activity of information retrieval. In addition, in the context of aid and remediation with pupils and students in difficulty, Kuhlthau is particularly interested in information retrieval mediated by an expert (a researcher or a librarian), the retrieval of information in a learning process, the latter being apprehended under the angle of socio-constructivist theories. According to Khulthau [KUH 85, KUH 88, KUH 99a, KUH 99b], two independent axes are to be distinguished (Figure 2.3): – on the one hand, the six [KUH 88] and then seven

[KUH 99a] successive steps present in all searches for information: - the initiation, namely the identification of an information requirement that is at the very source of all activity,

24

Information Retrieval in Digital Environments

- selection of a source and/or information resource, - exploration in order to extract details to better delineate the contours of the information need, - formulation of concepts generally translated into keywords, - collection of information, - presentation, which is the restitution of information gathered and collected, - assessment not only of the final product (e.g. an oral presentation and a written report) but also the procedure and of the activity of information retrieval. – On the other hand, the three types of factors likely to influence these seven steps are: - physical aspects, corresponding more precisely to the motor aspects, - affects, referring to the sensations and feelings experienced by the user during the different stages (worry, fear of not finding relevant information, apathy, despondency, etc.), - cognition (included in the term thoughts). The ISP model thus presents two major interests: on the one hand, the activity of information retrieval is considered in its entirety, from the initiation phase (i.e. definition of information required) to the use – restitution of the information collected and its evaluation; on the other hand, the factors regarding motor functions as well as affects and intellectual dimensions are regarded as likely to facilitate and/or disrupt one or more stages of the information retrieval. The ISP model is, therefore, a holistic model that proposes apprehending all the behaviors and intellectual processes during all the phases of information retrieval. For this

Conceptual and Methodological Approaches to Information Retrieval

25

reason, today, the ISP model is a tool that is still very often used in the training of documentalists and librarians, since it pretends to be generalistic and global. Moreover, understanding the difficulties of learners and helping them to overcome their difficulties has been – and remains – one of the main motivations of the authors who developed this model.

Figure 2.3. The seven steps and three dimensions (physical, cognitive and emotional) of information retrieval according to Kuhlthau [KUH 99a, KUH 99b, KUH 04]

However, from the point of view of psychology and ergonomics, the focus of the ISP model remains limited even if other authors have attempted to clarify it [ELL 89, HYL 06, HYL 09a, HYL 09b]. The ISP model has been and is generally still “valid” due to in situ observations of pupils or students looking for information in digital environments (mainly, the Web). From a descriptive and qualitative point of view, these studies provide very valuable information. However, many concepts remain relatively unclear, and no experimental validation has actually been proposed. For example, thoughts correspond to cognitive processes, according to the authors supporting the model; however, no details are given about the mental mechanisms that are referred to. Similarly, the sensations and feelings of end users (e.g. uncertainty and doubt) are not clearly defined.

26

Information Retrieval in Digital Environments

Finally, the applications of this model are generally considered as learning and/or training initiatives aimed at users without any real prospect in terms of design of digital environments. 2.4. The first studies of psychology and cognitive ergonomics Although studies on ICS provide interesting insights as to the behavior of users looking for information, they are primarily descriptive (i.e. what an individual does) and prescriptive (i.e. what an individual should do to efficiently search for information). In addition, the willingness to translate the comments collected during the proposal of learning and/or training activities tends to consider information retrieval as an activity composed of successive, sequential and clearly distinct phases. However, the existence of parallel and automatic processes were quickly identified. Finally, from a methodological point of view, very little experimental data have been used to support the theoretical conceptions. Also, works from psychology and cognitive ergonomics emerged as a relevant alternative to, not only describe, but also explain the behavior of individuals retrieving information in digital environments. Cognitive approaches focus on the processes and mental mechanisms underlying the behaviors involved in the activity of information retrieval. Guthrie and his colleagues [GUT 88, GUT 90, GUT 87, GUT 91] are often regarded as the first to have experimentally studied the mental processes involved in information retrieval. Their studies were then related to “traditional” paper environments (e.g. manuals and directions). According to these authors, information retrieval goes through four successive stages: defining a goal;

Conceptual and Methodological Approaches to Information Retrieval

27

selecting a text or a part of a text; extracting and integrating into prior knowledge of the relevant information found in the text; and evaluating the result obtained. If this assessment is not satisfactory, the individual restarts at the first step until the goal is reached. The mass arrival of digital environments in all our working and living spaces, and the development of the Internet and the Web in particular, have contributed to increase the number of studies truly focused on the mental mechanisms involved in information retrieval. Since the early 1990s, Ingwersen [ING 92, ING 96] has proposed considering human behavior related to information retrieval as a series of dynamic and interactive processes between the cognitive space of an individual and the information space offered by the digital system. The originality of his model lies in the fact that he no longer considers the interaction between two entities (the human vs. the technical system), but an interaction between three entities (Figure 2.4): – the cognitive space of the individual, including the representation of the need for information, the representation of goals and sub-goals or even how the instructions are understood; – the technical space linked with the information retrieval system, comprising the documentary languages implemented, the structure of the information, the indexation rules or even the underlying formal logic; – the information stored in the memory of the individual (i.e. knowledge) and the information contained in the technical system (i.e. images and texts). According to Ingwersen’s model [ING 92, ING 96], the state of the three entities mutually alters during the information retrieval. Since 1996, we can also observe that Ingwersen envisages the involvement of the social and organizational environment, and that he will only integrate

28

Information Retrieval in Digital Environments

it in the three spaces previously cited several years later [ING 05].

Figure 2.4. The model of the interactive processes of information retrieval according to Ingwersen [ING 96]

Finally, one of the notions that Ingwersen’s model [ING 96] has introduced is that of “polyrepresentation of information objects”: by these words, Ingwersen wants to emphasize the inter-individual differences concerning understanding – mental representation of information present in an informational environment (such as a Website). Indeed, the mental representation of information may be different depending on the individual considered: the designer – creator of the said information; the diffuser (e.g. creator of the Website); or even the reader – end user. During the same period, Saracevic [SAR 96, SAR 97] developed his “stratified interaction model”, which offers a global vision of the process of information retrieval. Further, the vision can be qualified as global in the sense that Saracevic attempted to apprehend multiple components involved in information retrieval (Figure 2.5):

Conceptual and Methodological Approaches to Information Retrieval

29

– the situational and contextual component (i.e. contexts of practice, tasks and instructions, definition of the problem and goals to achieve); – the emotional and intentional component (i.e. beliefs, emotions, expectations and motivations of the individual that retrieves information); – the cognitive component (i.e. domain knowledge of the end user); – the query component (i.e. characteristics of document queries and/or end user keywords); – the man–machine interface component (i.e. aspect of the interface and surface “ergonomics”); – the technical component linked with the software to retrieve information (power, memory, etc.); – the software and algorithmic component architecture of the information retrieval system);

(i.e.

– the holding components (i.e. size, volume, density, structure and characteristics of the document and information holdings).

Figure 2.5. The stratified interaction model by Saracevic [SAR 97]

30

Information Retrieval in Digital Environments

As Simonnot emphasized [SIM 08], one of the main concerns of Saracevic’s stratified interaction revised model [SAR 97] is to distinguish between different types of relevance: – the relevance system or algorithmic outcome of the evaluation by the system of the adequacy between the information provided and the request generated by the end user; – the subject or thematic relevance (topicality), which affects the semantic distance between the theme of the query generated by the user and the theme of the information proposed; – the cognitive relevance, which affects the “cognitive” distance, in the problem space given, between the state of the knowledge stored in the end user’s memory and his need for information; – the situation relevance that corresponds to the utility of the information proposed by the system to achieve the goal (final state); – the affective or motivational relevance that corresponds to the capacity of the information to satisfy and please the user. However, although interesting, these different types of relevance are difficult to operationalize, especially when Saracevic himself recognizes that all these types of relevance are interdependent and interrelated. We can point out that Saracevic’s model of stratified interaction [SAR 96, SAR 97] has several points in common with Ingwersen’s model [ING 96]: – the context (social and/or physical and/or temporal and/or emotional) must be taken into account if we want to understand the behaviors and mental processes related to information retrieval;

Conceptual and Methodological Approaches to Information Retrieval

31

– three levels (or strata) of interaction exist: - at a superficial level, the individual interacts with a technical system by using commands and/or producing queries, - at a cognitive level, the individual interacts with the products of the system (i.e. processes the outputs such as images or text), - at a situation level, the individual interacts with a problem space containing an initial state (the demand and information need) and a final state (the information sought). 2.5. The cyclic model of David, Song, Hayes and Fredin David, Song, Hayes and Fredin [DAV 07] have proposed a model entitled ISC (for Information Seek Cycle) in which the emphasis is on the cyclical nature of the activity of information retrieval. Because it is intended to be simpler and more accurate at the same time, the ISC model focuses on the only moment during which the individual interacts with a technological system offering information, without integrating the previous (e.g. definition of information need and elaboration of the problem space) or later stages (e.g. restitution of information). Also, the ISC model assumes a very limited number of steps (three), which are repeated as long as the aim is not achieved, i.e. as long as the relevant information is not found (Figure 2.6): – during a preparation phase, the individual makes choices among the links on which it is possible to click and takes decisions; – during an exploration phase, the individual examines and navigates among the information that his choice has generated;

32

Information Retrieval in Digital Environments

– during a consolidation phase, the end user processes and integrates the information that seems useful and relevant to him. If this cycle of information retrieval is satisfactory (i.e. if the information found is relevant), then the activity stops; if this cycle is not deemed satisfactory, a new cycle begins.

Figure 2.6. The cyclic pattern of David et al. [DAV 07]

As said above, the ISC model focuses on the only moment during which the individual interacts with a technological system. But, paradoxically, one of the main interests of the ISC model of David et al. [DAV 07] is to include the activity of information retrieval in a period of time much longer than the period of time during which the individual seeks information on the screen. For example, past experiences lived by an individual can influence the perception that this individual has of the level of difficulty of the task to achieve [LOC 02]: this concept corresponds to the concept of “user experience”. Similarly, individual characteristics, such as the sense of self-efficiency when operating computers or intrinsic motivation [COM 95, MEY 04], can influence the preparation phase, consequently influencing the other two successive phases (exploration and consolidation). More precisely, according to the ISC model, past experiences would be a looking-back component that would influence the investment

Conceptual and Methodological Approaches to Information Retrieval

33

and persistence behavior in future information queries (looking ahead). 2.6. The skills-centered model of Brand-Gruwel As with the authors mentioned previously, Eisenberg and her colleagues [EIS 90, EIS 02] consider information retrieval as being composed of different successive phases. However, unlike the other authors, they propose an approach focusing on skills, in distinguishing between six major skills in their model (Big6TM skills). This approach has found a favorable resonance with educationalists and learning specialists [BOE 00, FED 03]. Based on this skills-centered approach, Brand-Gruwel and her collaborators [BRA 08, BRA 06, BRA 05, BRA 09, WAL 08] propose focusing, not on the steps of information retrieval as such, but on the skills necessary to effectively carry out the following six steps: – task definition; – choice and the definition of the “best research strategy” to adopt; – location of and access to information; – use of collected information; – synthesis (of information already existing in memory);

found

with

information

– evaluation. According to these authors, information retrieval is a “problem solving” activity and must therefore be dealt with as such: the end user starts from an initial state generally poorly defined and must reach a final state (i.e. the goal) respecting the physical and/or temporal constraints and reaching sub-intermediate goals. Going from one sub-goal to

34

Information Retrieval in Digital Environments

the next sub-goal is only made possible by mobilizing a number of skills, of cognitive and metacognitive processes. Based on this assumption, Brand-Gruwel, Wopereis and Walvaren [BRA 08, BRA 09] proposed the IPS-I model (Figure 2.7) for “Information Problem Solving on the Internet”, a model that is currently being further refined [BRA 11]. In the latest version of their model, the authors propose that the regulation activities are central and transversal, since they should oversee all of the mental processes and skills involved in information retrieval. More specifically, the regulation of related behaviors and processes should intervene in order to: – manage the achievement of the task of information retrieval as a function of the material and temporal conditions (after defining the information need); – manage and adjust its researching and information analysis behavior (e.g. to change its behavior if the information found is not satisfactory); – assess the relevance of the information found on the Internet (i.e. credibility, recency and thematic relevance); – evaluate the product resulting from the information retrieval after processing the information (e.g. to disseminate it and to draft a plan). From a cognitive point of view, the regulation activities correspond to metacognitive processes (planning, control and regulation). In addition, for Brand-Gruwel, Wopereis and Walvaren [BRA 08, BRA 09, BRA 10], three basic skills are crucial during information retrieval on the Internet: reading capacity (in the sense of reading – understanding the language), evaluation capacity (e.g. to judge the credibility of a source of information) and abilities related to the mastery of computers and digital environments (e.g. typing with a keyboard and going from one Website to another).

Conceptual and Methodological Approaches to Information Retrieval

35

Figure 2.7. The skills-centered model (IPS-I) according to Brand-Gruwel, Wopereis and Walraven [BRA 09]

As previously mentioned, the main focus of the IPS-I model is to address information retrieval under the perspective of the skills required. From a pedagogical point of view, the formalization of this model allows us to target the skills to transmit or develop. From a cognitive point of view, the IPS-I model places the cognitive and the metacognitive processes at the center of the activity and proposes to distinguish the “main skills” (such as reading or evaluating) and subordinated skills (“sub-skills” such as inferring keywords or coming up with a problem from a research topic). This distinction explains nearly all the studies conducted by Brand-Gruwel and his colleagues who seek to determine the strategic differences between experts (the domain vs. the procedure) and novices, the participants usually being students.

36

Information Retrieval in Digital Environments

2.7. Kitajima’s predictive model The models presented previously attempt to describe the underlying behaviors and mental processes involved during information retrieval. In parallel to these descriptive models, other authors develop models aiming to simulate and predict these same behaviors and mental processes. Kitajima and his collaborators [BLA 02, BLA 07, KIT 03, KIT 08, KIT 00, KIT 05] have developed one of these predictive models of the cognitive processes involved in searches for information in digital environments: the CoLiDeS model (for “Comprehension-based Linked model of Deliberate Search”). This model relies on the one hand, on the theory of construction – integration of Kintsch (1988 and 1998), namely a theory outcome from textual psycholinguistics and, on the other hand on several works initiated in the mid-1990s on certain specific graphical interfaces or GUI (for Graphical User Interface; [KIT 95, KIT 97]. According to Kitajima and his colleagues, all the actions that an individual can achieve on and/or from a Web page are under the influence of only two processes (Figure 2.8): – the process of attention: the user visually partitions the screen that is displayed before him into zones, i.e. the information is distinguished and grouped by the end user – reader. These distinctions and groupings made by the user do not necessarily correspond to the cutting done by the page designers into frames, in menus, etc.; – the selection process: the user decides to concentrate his attention (his reading and his processing) on some of the areas previously distinguished and grouped. In addition, the end user infers the content which he will/would access if he clicks on the hypertextual links present on the Web page.

Conceptual and Methodological Approaches to Information Retrieval

37

Figure 2.8. The Kitajima CoLiDes model [KIT 03]

According to the designers of the CoLiDeS model, these two processes of attention and selection are involved in several phases of the retrieval of information: – on the one hand, on the basis of information and environmental constraints and on the other hand, on his own knowledge stored in long-term memory; the individual develops a mental representation of the goal and of the different sub-goals. In the CoLiDeS model, the mental representations are in the form of macro and micro-proposals composed of predicate-argument pairs [KIN 88]; – in front of a Web page, the individual analyses the structure of the page and the physical characteristics of the content (the general organization, the typographies, the paratextual cues, etc.). On the basis of this analysis, the user “cuts” the page into different zones; – in parallel to this analysis of the shape of the Web page, the individual discriminates and brings together the

38

Information Retrieval in Digital Environments

different elements semantically linked that make this page. This content analysis is, therefore, closely linked to the level of declarative and procedural knowledge stored in long-term memory; – once these two analyses are done (form and content), the individual focuses his attention on the part of the page that seems the most relevant, i.e. on the area that seems to allow him to reach the final goal. At this time, the user must activate his knowledge to understand the content but particularly to infer the type of information that he is likely to obtain subsequently (e.g. as he clicks on a hypertext link); – the individual, therefore, builds a second mental representation of the content that he will access and compares this second mental representation with his first mental representation of the goal. If the semantic distance between these two mental representations is small, the user then selects the hypertext link, this selection being translated into a click. One of the main strengths of the CoLiDeS model is that it is based on the theory developed by Kintsch [KIN 88, KIN 98], which is an extremely robust formalization. It takes into account the mental processes of understanding and the construction of mental representations of complex linguistic content. In addition, Kinstch’s theory places the productions of inferences and elaborations in the heart of the process of understanding. Precisely, when we use the Internet, we must very often infer and/or develop (i.e. foresee and predict) the contents, which we could access if we clicked on such or such hypertext link. In addition, associated with the latent semantic analysis (or LSA), the behavior and performance predicted by the CoLiDeS model are very highly correlated with the behavior and performance of individuals in real situations of information retrieval [BLA 07, KIT 08, KIT 05].

Conceptual and Methodological Approaches to Information Retrieval

39

The CoLiDeS model thus presupposes the presence of joint processes ascending and descending during the search for information – surfing the Web: on the one hand, the user relies on the physical characteristics and/or surface of a Web page; on the other hand, this same end user uses his knowledge (of the domain and procedural) to make his choice and decisions. With the CoLiDeS model, information retrieval and navigation are thus closely linked if not inseparable. 2.8. The hyper-specialized model of Sharit, Hernandez, Czaja and Pirolli Most of the models related to behavior and mental processes involved during the search for information in digital environments do not distinguish the different tools and are supposed to apply to all types of electronic environments (e.g. CD-Rom encyclopedia, offline or online databases and documentary data banks, digital library, Web, etc.). However, because any information retrieval typically involves the use of an internal or external search engine (e.g. Yahoo and Google), some authors have recently proposed to focus almost exclusively on the behavior of users during their use of these search engines as a priority [PAK 08, SHA 08, STR 06]. The main argument of these authors is that the behavior and performance related to the use of a search engine largely determine the success or failure of any information retrieval. In other words, browsing a page or between Websites is “secondary” and is conditioned by the use of search engines. Drawing heavily from the work of Sutcliffe and Ennis [SUT 98] and Marchionini [MAR 95], and borrowing multiple concepts of Borgman [BOR 86] and Kuhlthau [KUH 99a, KUH 04], Sharit, Hernandez, Czaja and Pirolli [SHA 08] have created one of these new models centered on the behaviors and psychological processes involved in the use of

40

Information Retrieval in Digital Environments

search engines to acknowledge the difficulties of the elderly (Figure 2.9). Among the models of this type that we call hyper-specialized because centered on search engines and a specific population, that of Sharit et al. [SHA 08] appears to be the most accurate.

Figure 2.9. Processes and behaviors involved when using a search engine, according to Sharit et al. [SHA 08]

Just as other models previously described, the model of Sharit et al. [SHA 08] considers the activity of information retrieval to be a cyclic activity, which simultaneously involves declarative knowledge, procedural knowledge and several cognitive abilities (working memory, visuo-spatial skill, reasoning ability, mental flexibility, etc.). Moreover, many studies prior and/or subsequent to the publication of this model confirm that individual factors, such as cognitive flexibility, the ability to inhibit certain inferences, spatial skills, processing speed or even the capacity of the working memory, significantly affect the performance and behavior of users searching for information in digital environments

Conceptual and Methodological Approaches to Information Retrieval

41

[CHE 07, CZA 01, KUB 99, LAB 05, MAR 09, SLE 09, VAN 10]. These studies have particularly focused on the behavior and performance of the elderly, because this population has significant weaknesses and/or weaknesses on these cognitive dimensions. Depending on the hyper-specialized model of Sharit et al. [SHA 08], information retrieval is assimilated to a “blurred” problem solving situation whose contours are poorly defined [CHI 85], a situation in which three successive moments can be distinguished: – the individual develops a mental representation of the problem to be solved, the different states of the problem (i.e. the problem space: sub-goals, initial state and final state) being encoded in the form of multiple representations. The construction of this mental representation is under the direct influence of knowledge of the area and of the nature of the information need (complex, simple, highly structured, “blurred”, etc.); – during the planning phase, the user develops a procedure in order to achieve the final state. This development depends closely on the procedural knowledge and the language skills of the individual; – during the execution, the individual reflects and effectively implements the procedure he has planned. The execution produces keywords to query the search engine. These three moments are repeated as long as the ultimate goal is not reached. Based on the approach of Sutcliffe and Ennis [SUT 98], Sharit et al. [SHA 08] postulate the existence and intervention of four types of knowledge when using a search engine: knowledge of the domain (or declarative); knowledge related to the technical tool actually used; knowledge related to the activity of information retrieval (which corresponds from a cognitive point of view to procedural knowledge); and knowledge related to

42

Information Retrieval in Digital Environments

informational resources, i.e. specific to the databases or Websites queried. 2.9. The Landscape Model “diversion” by Dinet Most of the models presented above try to apprehend and describe the behavior involved in the activity of information retrieval in digital environments. Some allow us to predict the behavior and performance of users. Finally, others are focusing on the skills or even the knowledge, the capabilities and the abilities involved. However, all these models share a common point: they seek to address the activity of information research in a comprehensive manner in order to propose an “overview” of the factors involved, whether these factors are facilitators or disruptive. Another approach focuses on the impact of cognitive mechanisms, studied in other contexts and/or tasks, in connection with certain specific difficulties. Therefore, the difficulty of young users to select relevant information among all that proposed by search engines has been interpreted calling upon activation mechanisms – knowledge inhibition in working memory [DIN 03, DIN 02]. Whatever the environments and tools used, any information retrieval has generally two crucial phases: first, the individual must question the database (e.g. CD-Rom encyclopedia, digital library and Web) issuing a query; secondly, this individual must select the relevant information from that proposed by the tool in response to his request. A large number of descriptive studies conducted since the beginning of the 1990s have acknowledged the difficulties of end users, especially the younger ones, to select relevant information from large information bodies [BAR 94, BAR 98, BRU 94, CHA 00, HIR 96, PEI 98, SCH 91, SCH 94, WAL 94]; for a synthesis, [MIZ 98]. However, beyond the simple observation, no explanation was then proposed. The

Conceptual and Methodological Approaches to Information Retrieval

43

use of some theoretical models derived from cognitive psychology, such as the Landscape Model, seems to allow us the understanding of the causes of difficulty to assess information relevance. Among the computational models supposed to take account of the mental processes involved in reading, the Landscape Model developed by van den Broek and his collaborators since the 1990s [LIN 00, LIN 02, VAN 93, VAN 95, VAN 96, VAN 99] has particularly caught our attention for three main reasons: – The Landscape Model describes and predicts both the online process of reading – understanding of the offline consequences of memorization of information content. – This model is based on the one hand, on declarative knowledge stored in memory in the form of semantic networks and on the other hand, on mechanisms of cognitive research and on maintaining consistency when reading. – Numerous studies have demonstrated the very strong predictive power of this model by comparing the data obtained from computational processing and behavioral data of individuals when reading [LIN 04, TZE 05, VAN 05, VAN 01]. The Landscape Model is a neo-connectionist model based on the theory of activation: as for all models based on this theory, memory is thus conceived as a network comprising a large number of basic units of representations (lexical and conceptual units) massively interconnected. The basic units are called network “nodes”, these nodes corresponding to concepts in the Landscape Model. According to this model, the processing of linguistic material (reading or hearing) causes the activation of basic units (conceptual units). These basic units, or concepts, then reach a level of maximum activation. In other words, the concepts activated first are those explicitly read and/or heard by the individual in the

44

Information Retrieval in Digital Environments

statement being processed. But, other concepts can then be activated, by propagation and diffusion of the activation. These other concepts can be enabled from long-term memory (LTM) or since previous processing cycles (i.e. concepts momentarily activated and stored in working memory). These other concepts can be activated for two reasons: on the one hand, they constitute an aid to understand the language units being processed: in this case, the individual performs a controlled and explicit search of concepts related in his memory; on the other hand, they are closely connected (lexically or semantically) to the activated concepts: in this case, the activation is automatic, implicit and irrepressible. If the activation of concepts in memory is central in the Landscape Model, their inhibition and deactivation are just as important. Indeed, two cases can occur according to the Landscape Model when an individual reads a new statement: – either this new statement causes the activation of concepts related to those present in the already activated cohort; in this case, the mental representation of the instructions and of the theme remains stable and consistent, since the cohort is only very slightly modified, some constituent concepts simply see their activation level significantly fluctuate; – or, this new statement causes the activation of concepts that are not related to those already active in the original cohort; in this case, a new cohort, composed of new “incoming” concepts, is created. However, because of his limited memory and attention capacity, an individual cannot keep several conceptual cohorts that do not share the same concepts active simultaneously. A “competition” between the newly activated and the original cohort then begins. According to the Landscape Model, the activation of a new conceptual cohort “drives out” the original cohort to replace it in memory: indeed, the activation pool is limited, the

Conceptual and Methodological Approaches to Information Retrieval

45

activation of the incoming concepts happens to the detriment of the concepts present in the original cohort, the activation level of the latter decreases until they disappear. Van den Broek and his collaborators have often used the metaphor of the magnetic tape to account for this phenomenon: a newly activated cohort “clears the previous cohort by over recording”. The product resulting from the activation of different concepts (i.e. the mental representation of the content) fluctuates, therefore, at the pace of the reading activity. This fluctuation is graphically represented by a landscape in 3D, thus justifying the name Landscape Model (Figure 2.10).

Figure 2.10. Illustration of the fluctuation levels of activations of the concepts according to the Landscape Model [VAN 01]

46

Information Retrieval in Digital Environments

Initially, the Landscape Model was developed to account for the process of reading-memorization of narrative texts [VAN 93, VAN 95, VAN 96]. However, the authors have gradually expanded the application areas of their model to other types of textual material and to situations more complex than reading-memorizing (e.g. understanding scientific debates from readings of descriptive scientific texts; [VAN 08]). In the same way, and in agreement with the designer of the model, we assumed that the Landscape Model could also explain the behavior and performance in information retrieval tasks because, after all, reading (in the sense of processing linguistic material) and memorizing are at the heart of these activities [DIN 03, DIN 09]. According to the predictions of the Landscape Model, processing the subject of retrieval of information (e.g. the thematic statement, the instructions, the objectives to be achieved and the conditions to comply with) resulted in a first process of activation diffusion of conceptual units in memory, whose aim is the construction of a stable and consistent representation of the theme. For example, according to the stock of conceptual units stored in memory by the individual, the theme “global warming” can activate, through the bias of automatic or controlled inferences, an associated set of concepts such as “pollution”, “temperature”, “ecology”, “planet”, etc. This set of concepts primarily activated (from instructions or retrieved from LTM or working memory) forms a “conceptual cohort”, this cohort being synonymous for the designers of the Landscape Model with mental representation. The size of this cohort (i.e. the number of active concepts) as well as the level of activation for concepts depend on a broad set of factors, which include at the forefront the initial declarative knowledge that the individual has. In other words, the level of familiarity of the theme of information retrieval determines the size of the cohort, as well as the level of activation of concepts that it involves.

Conceptual and Methodological Approaches to Information Retrieval

47

Always in accordance with the postulates of the Landscape Model, reading information provided by a digital system (e.g. a Website and a list of results from search engines) can lead to two cases: – either this information is closely semantically related to the concepts already active; in this case, the mental representation of the instructions and of the theme remains stable and consistent, since the cohort is only very slightly modified; – or, these information are not related to the theme of the research of information and in this case, they lead to the activation of new concepts. Since it cannot hold several active cohorts, content related to new processed information “erases” the initial representation developed during the processing of the instructions and of the subject of the information retrieval. In practical terms, the “erasing” of the initial mental representation (i.e. the purpose) must lead to difficulties in selecting the information quickly and efficiently, since the individual must then reactivate this initial representation. This is what we have demonstrated with two experiments [DIN 03, DIN 09]. In the first experiment conducted on 69 children of 10–11 years of age, we have demonstrated that the representation that these children have of the instructions and the subject of their search for information gradually and quickly degrades as they go through lists of references of Websites and/or bibliographic records. This degradation translates into a significant lengthening of the selection time and a greater number of errors in the selections. In addition, some factors accelerate this degradation such that the lack of familiarity with the subject of the search for information. In other words, the children appear to gradually “forget” the subject of their search for information (by disabling the relevant initial concepts) as they keep on searching. In the second experiment conducted on 21 children of 10–11 of age, we have demonstrated that the difficulties of selecting relevant information were

48

Information Retrieval in Digital Environments

actually due to a gradual deactivation of the relevant concepts in memory, since a rearticulation of these initial concepts (i.e. reactivation of the instructions and the subject) was accompanied by an increase of performance of students during the selection task. In other words, an “external” help, here a re-explanation of the instructions and the subject, seems to help young users to reactivate the relevant concepts, as well as helping them to assess more quickly and effectively the relevance of information and bibliographical references that they are presented with. If the need to remember the instructions and the theme of the information retrieval appears to be fairly trivial to an adult, maintaining the objective and the subject requires a certain amount of knowledge and/or experience of information retrieval tasks that young learners and novice adults do not seem to possess [DIN 02, HIR 96, KUH 99a, KUH 99b]. Moreover, some studies show that it is only after a long apprenticeship that individuals gradually learn to pay special attention to the instructions and the subject of their information research [DIN 09, HIR 99, KUH 99b]. Also, the question of the students’ autonomy to carry out information retrieval, alone in front of the computer, still remains. This question is all the more crucial when young users are carrying out literature and/or information research in environments that display an immeasurable body of information, decorated with animations and attractive “objects” (e.g. the Web). From a theoretical point of view, the important question is: Is information retrieval a specific activity that must be approached with specific models? Or is information retrieval an activity that must be apprehended with theoretical models borrowed from other activities (e.g. problem solving and reading-memorizing)? In other words, it is the status of the activity of “information retrieval” that is questioned here.

3 Information Retrieval: Psycho-Ergonomic Approach

3.1. Introduction As described in previous chapters, several disciplines and several activity sectors are concentrating on the activity of information retrieval, especially when the latter is performed in digital environments. In this chapter, we intend to present and defend the psycho-ergonomic approach of information retrieval: on the one hand, this approach combines psychology and ergonomics and is the most relevant for describing and anticipating the behavior of human individuals, since it centers on individuals interacting with socio-technical systems; on the other hand, this approach relies both on robust theoretical models (particularly from cognitive psychology) and on scientifically proven methods. The psycho-ergonomic approach of information retrieval is relatively recent in comparison to other approaches. It also appears urgent to continue with works in this direction in order to respond to certain limitations as follows: – there are a large number of studies that focus on the behavior and the mental processes involved during the

50

Information Retrieval in Digital Environments

search for information in digital environments. However, the participants in these studies are generally adults who have a certain level of experience and/or knowledge in specific areas: students, adults in professional activities, experts in specific sectors (e.g. aeronautics, training, medicine). Comparatively, very little is aimed at young users, ordinary users, the general public or socially isolated people or even “digitally excluded” people [JOC 08]. Yet, it is the general public that we must assist as a priority; – works focusing on information retrieval behaviors in digital environments essentially come under clinical approaches and/or case studies. The main objective of these studies consists of describing the behaviors, the difficulties and the performance of users when searching for information; those that trying to interpret these behaviors from the point of view of the underlying mental mechanisms are much less frequent; – the large majority of studies that concentrate on the behavior of individual humans when searching for information in digital environments fall into three main areas: education sciences, information and communication science (ICS) and information technology. These different approaches are namely characterized by differences in the objects of study (e.g. the organizational context, the teaching process, the technical tool), the theoretical frameworks, the methods and of course the objectives (e.g. to develop a new technical tool, to assess the impacts on teaching processes). In addition, it is nowadays very difficult to generalize the collected data. On the basis of these findings, the most promising recent works are guided by three convictions: – although less numerous in the field of information retrieval, the approaches stemming from psychology and ergonomics are perfectly complementary, legitimate and relevant, because they allow an approach that is at the same

Information Retrieval: Psycho-Ergonomic Approach

51

time analytical and holistic to information retrieval in digital environments: - an analytical approach since it offers models and methods that allow a fine-grained apprehension of a number of underlying behaviors and mental processes, - a holistic approach since it enables the situating and integration of the elements studied in more global contexts; – the (re)design and modification of technical tools and interfaces must not only be based on the analysis of the actual behaviors but must also be constantly requestioned in the light of new behavioral contexts data; – the dialogue between fundamental and applied research

must be constant since these two aspects mutually enrich each other; on the one hand to increase scientific knowledge, and on the other hand to respond to problems and societal issues (e.g. struggle against socio-technological exclusion, prevention of accidents). To illustrate this threefold positioning, the following sections present several studies that exemplify the desire that animates researchers in ergonomic psychology, namely to reconcile fundamental research and applied research (primarily from psychology and cognitive ergonomics) and to take into account several human dimensions in the (re)designing of technical tools and/or methodologies. 3.2. Identifying difficulties in modifying interfaces Some tools dedicated to information retrieval are aimed at specific audiences. In addition, we must absolutely take into account the specificities of these specific audiences (at the sensory, motor, cognitive levels) if we want to be able, on the one hand to understand the possible difficulties for these users, and on the other hand to provide the appropriate responses.

52

Information Retrieval in Digital Environments

For example, the range of BDCI© tools (developed by the CRDP of Poitou-Charentes) covers a collection of software applications for young learners, from kindergarten and primary school (BCDI-maternelle© and BCDI-ecole©) to junior high school and secondary high school (BCDI©), of Web versions currently existing (BCDI-Web©). Regarding the youngest learners (BCDI-maternelle© and BCDI-ecole©), this software is more of a learning tool, whose approach is related to literature research, than a formal tool of information retrieval. Indeed, this software is part of an educational device, which aims to build and develop certain skills related to information retrieval: keyword definition, documental query formulation, distinction of the various fields of a bibliographic notice (“author”, “title” “summary”, etc.), or even understanding the semantics environment of a term. The team that designed this software desired an ergonomic study be carried out in order to, on the one hand better understand the difficulties of users (young students), and on the other hand provide recommendation elements for future versions of the product. Users’ apprehension of the difficulties has gone through the association of two complementary phases: a hierarchical task analysis and a real activity analysis. These two analyzes are “traditional” methods of ergonomic psychology. 3.2.1. Hierarchical task analysis The hierarchical analysis of tasks is commonly used in work psychology and ergonomics. The objectives and aims of this analysis are multiple: – to lead to a descriptive model in which a task (i.e. a goal) is decomposed into subtasks (i.e. subgoals); – to better understand the prerequisites and constraints that may intervene during the general process;

Information Retrieval: Psycho-Ergonomic Approach

53

– to better understand the relationships (namely temporal) existing between different subtasks and procedures; – to identify the different a priori pathways of an end user; – to detect the points or “nodes” of the process where the end user may find difficulties; – to obtain a task model from the point of view of the digital device, or more precisely from the point of view of the device designer. Regarding the literature research carried out with version 2 of the BCDI-ecole© software program, by applying the hierarchical task analysis proposed by Hollnagel [HOL 91], three major subgoals have been identified: launch the application and the information base; query the information base; select a bibliographic notice. For each of these subgoals, the pathways and the related procedures can be described by diagrams. With the task and the hierarchical process analysis, some users’ problematic points (here, of young readers) have been identified: – during the information database query, the end user can perfectly come up with multiple keywords without necessarily using the Boolean operators “AND”, “OR” and “XOR” (Figure 3.1). However, the use of Boolean operators was evident for the designers of the software, from the moment the end user wanted to input several keywords. – the proposed research interface does not enable the field that the retrieval focuses upon to be distinguished: indeed, the goal of an end user is very different whether he/she searches for books about “Alphonse Daudet” (i.e. books whose theme is Daudet) or from “Alphonse Daudet” (i.e. books whose author is Daudet). A query conducted with

54

Information Retrieval in Digital Environments

version 2 of BCDI-ecole© using the keyword “Daudet” used to therefore cover simultaneously all the fields (title, summary, author, etc.);

Figure 3.1. Possible actions to query the information database with BCDI2-ecole©

Information Retrieval: Psycho-Ergonomic Approach

55

– there was a large number of opportunities offered to the end user to consult bibliographic records (all these actions resulting in fine in the same outcome; – many actions permitted by version 2 of BCDI-ecole© corresponded to expert procedures associated with the literature research task. Therefore, the presence of parentheses to write a search algorithm or even the title of some fields (e.g. the field “sequence” corresponding to pages numbers) were justified by the fact that experts in documentation had essentially implemented their knowledge within the software. The changes on these issues have been easily taken into account and carried out by the designers. Although essential, it is impossible to apprehend the real activity of end users through the task analysis. This is the reason why in situ observations were conducted in order to, on the one hand to identify accurately the differences between the task and the actual behaviors, and on the other hand to determine some of the difficulty factors in the execution of the activity. 3.2.2. Analysis of the end users’ behavior “What do end users really do with the software?”, “Do they use it the way that the designers had foreseen that they would use it?”. Here are two questions to which it is absolutely necessary to answer if we want to understand the possible differences between the task prescribed (i.e. what the end user was supposed to do) and the actual activity (i.e. what the end user does). There are multiple ways to analyze the behavior of the end users of a socio-technical system. In the case that we are interested in, it is an ecological environment observation (i.e. in class) that seemed to us the most relevant, on the one hand, to disrupt students as little as possible, and on the other hand, to take into account the various factors likely to influence behavior.

56

Information Retrieval in Digital Environments

Also, multiple observations accompanied by audiovisual recordings were conducted with children performing bibliographic searches with version 2 of BCDI-ecole©. The main objective of these observations was to get the most faithful representation possible of the actual behaviors of end users in context. For example, during one of these observations, 12 students from the same fifth grade class of the Academy of PoitouCharentes, a school located in a peri-urban zone, were followed. The 12 students were all French speakers and had been carrying out bibliographic searches every weeks, for two years, with the same software (version 2 of BCDI-ecole©). The students always used the software in an autonomous way. Nevertheless, they had the possibility to ask the education assistant for help, who was notably in charge of the proper functioning of the Documentation Center Library (B.C.D. – Bibliothèque Centre Documentation) as well as of the computers that were there. In the context of this observation, each student was to search for bibliographic records relating to two themes, one familiar (“blood”) and the other unfamiliar. The students worked for two weeks on this theme in the context of biology sessions. In addition, in the hour that preceded the session of bibliographic research, the teacher used a pre-activation method of thematic knowledge [LAN 84]. This pre-activation session involved three phases: (1) free associations carried out by the students concerning the theme of “blood”; (2) joint deliberations from the entire class around the free associations pupils came up with and (3) a reformulation by the teacher of the reflections carried out by the students. The unfamiliar thematic (“the animals of the sea”) had never been the subject of any class work. The unfamiliar thematic was tackled only for the sole purpose of preparing a school trip at the end of the year. No knowledge pre-activation session has been conducted. The students were warned of the

Information Retrieval: Psycho-Ergonomic Approach

57

title of the unfamiliar theme 15 minutes before they effectively carried out their bibliographic research while they were warned two weeks earlier that they would have to search for bibliographic records relating to “blood”. The only tool used by the students to complete their bibliographic research was version 2 of BCDI-school. The 12 fifth-grade students had to therefore complete their bibliographic research, individually and with the aid of their usual software (i.e. version 2 of BCDI-school). Each student had to simply note down on his/her personal notebook the titles or the codes of the bibliographical references evaluated as relevant, in order to be able to find and exploit the corresponding document the following week. The spatial configuration of the B.C.D. and the location of the computer did not allow the student who was doing his/her research to see what was happening in the rest of the room without looking back. Before the recording of the bibliographic research session, the observer (always the same) had been to the school regularly and had made many digital audiovisual recordings of students during their bibliographic research. Therefore, all the students were accustomed to the presence of the camera and the experimenter in the B.C.D. The camera was located behind the pupil and filmed from above (3/4 rear) in order to catch both the screen and the behavior of the user. In addition, a number of microphones allowed the collection of any possible spontaneous verbalizations from students. During the time slots reserved to the observations, the teaching assistant was present but had received instructions to help students only in cases of important technical difficulties. In addition, the teaching assistant was usually busy with a small group of students from another class at the back of the B.C.D. In other words, each student carrying

58

Information Retrieval in Digital Environments

out the bibliographic research was alone in front of the computer as usual. The recordings coding was conducted based on the coding used by Goldman, Zech, Biswas, Noser and the CTGV [GOL 99], these authors had studied objective planning in subjects supposed to solve mathematical problems with a computer. In our observation protocols, each piece of bibliographic research carried out by the pupils is decomposed into research cycles, a research cycle beginning when a query was starting to be input using the keyboard and ending when all or part of this query was modified. For each research cycle, five indices have been identified: – the relative times: they allow capturing the total time devoted to bibliographic research, as well as the time devoted to each research cycle; – the queries formulated by the pupil: they were recorded as they were made, namely with prospective spelling mistakes, – The actions of the pupil: they could be of two kinds: (1) a mouse click on a icon present on the software screen (the actions are then put between brackets); (2) the use of the alphanumeric keypad. – The reactions of the software to each of the actions: they could also be of two types: (3) the software responded to the action of the student by displaying a dialog box; (4) the reaction did not cause the display of a dialog box but instantly changed the screen display (e.g. returning to the previous screen). – the comments and paraverbal indices of the pupil. The examination of the queries carried out by the 12 students has shown significant intraindividual differences depending on the level of familiarity of the theme of the bibliographic research.

Information Retrieval: Psycho-Ergonomic Approach

59

Overall, students come up with few documental queries, especially when the theme of the bibliographic research is unfamiliar to them. In addition, six queries in (or 46.1%) correspond to the title of the theme of the bibliographic research in the case where this theme is unfamiliar (against 17.6% for the familiar thematic). In other words, when the bibliographical research theme is unfamiliar, the pupils merely make “reproductions” of the research subject to compose their queries. In addition, the level of familiarity of the theme seems to have an influence on the ability of users to infer associated terms: in the case of a familiar theme (“blood”), several specific terms related to the main title of the research are produced (e.g. “blood development”, “blood coagulation”, “blood group” and “albumin”); in the case where the theme is unfamiliar (“animals of the sea”), there is only a single specific request (“crustaceans”) and many spelling errors. In other words, the referential expertise influences the production of documentation queries. Indeed, when the theme of the bibliographic research is familiar to the students, the queries are more numerous, contain more specific terms and correspond less often to simple “reproductions” of the theme. This finding is perfectly compatible with the notion that the activation of a conceptual cohort in memory depends on the familiarity (i.e. semantic proximity) of concepts. The examination of the chronometric data also provided possible answers to the following question: “what do students do when they use BCDI-ecole©?”. For all students, the display time of each screen proposed by BCDI-ecole© was been recorded, namely the mean time of exposure of students to the various screens of BCDI-ecole© was recorded. The data shows that young users using BCDI-ecole© spend nearly half of their time in front of the search screen

60

Information Retrieval in Digital Environments

(45.8% of the total average time), and spend more than a third of their total activity time in front of a blank search screen (i.e. without formulating any query). The other significant mean exposure timings correspond to the exposure time in front of bibliographic entries (42.21% of the total average time), either presented one at a time or in list form. It is reasonable to think that the exposure times in front of bibliographic records correspond to the time dedicated to reading and to the processing of the entry content. In other words, during bibliographic research conducted with version 2 of BCDI-ecole©, a young end user spends most of his/her time in front of the search screen and in front of the bibliographic entries (presented in a list or individually). We can reasonably assume that the time spent in front of a blank search screen is devoted to the planning of formulations that is planning the creation of queries. At least, this is the optimistic assumption! Regarding the exposure to the documental entries, the data shows that students spend more time in front of the entries when the theme is familiar. In other words, the students in our study spend more time to process the results of bibliographic research when the theme is familiar to them. Finally, an analysis of the protocols of the bibliographic research carried out by the students in our sample has allowed discriminating against the eight main types of difficulties experienced by the fifth grade end users when they used version 2 of BCDI-ecole© autonomously: presence of spelling mistakes in the queries formulated; non-use of the online dictionary (called DicoPrim) although it has been launched; persistence to use an unsuccessful query (i.e. “0 references found”): in this case, the user does not change his/her query even he/she has been able to see that no reference was found with this query; non-refinement of the

Information Retrieval: Psycho-Ergonomic Approach

61

query: this type of difficulty was observed in an end user if he/she does not further refine his/her research after proposing more than 100 bibliographic references; nondisplay of the bibliographic entry (the display was necessary to assess the relevance); misunderstanding of the meaning of icons; oversight of “starting the search” (i.e. validate) after formulating the query; and finally, poor relevance evaluation of bibliographical references. For each student, four coders working in a blind manner recorded the presence/absence of these eight difficulties. In addition, they were recorded if the activity was “successful”, notably if the student left the B.C.D. after writing down at least one book title (or the reference). The data show that only 5 students out of 12 have succeeded the activity, whatever the theme (familiar vs. unfamiliar). In other words, searching for bibliographic records with version 2 of BCDI-ecole© is a complex activity since the majority of users observed, which are all regular users of this software, cannot manage to do it. In addition, if certain difficulties are relatively rare among the participants observed (e.g. not refining a query), others on the other hand are quite frequent (e.g. incapacity to display the bibliographic references). 3.2.3. Implications for the (re)design of interfaces Based on the data collected and on the various analyses performed (task analysis and analysis of the actual behaviors of the children), the designers of the software BCDI-ecole© have developed new interfaces. Overall, the new interfaces of version 3 of BCDI-ecole© contain much less information, present a simpler organization and display content in an easier and more understandable manner than version 2, notably for young users (Figure 3.2). For example, the new search interface more clearly distinguishes the fields

62

Information Retrieval in Digital Environments

to which the research relates (a tab by field: themes, title, author, etc.), directly displays the bibliographic references when validating the query and contains much fewer buttons and possible actions. However, the new bibliographic record firstly displays the key information (title, abstract and keywords), clearly indicates if the corresponding document is available, distinctly separate “secondary” information (name of the collection, number, etc.) and does not contain large numbers of buttons and/or actions which until then had not been relevant.

a)

b)

Figure 3.2. a) Old and b) new version of the document query interface

Of course, it would be appropriate to carry out further analyses to test the real impact of these new interfaces on the execution of the activity by young users. Only some feedback and some in situ observations have revealed that the new interfaces appeared to overcome some of the difficulties encountered with the previous version of the software. Moreover, this example aimed to demonstrate that an analysis of the task and an analysis of the actual behavior of end users (i.e., those for whom the software was initially designed) are essentially significant to understand the difficulties and to propose recommendations.

Information Retrieval: Psycho-Ergonomic Approach

63

3.3. Anticipating the needs of users To analyze the actual behaviors of end users is a “compulsory prerequisite” in ergonomic psychology. But the task becomes much more complex when it comes to anticipate the future behavior of public that are not strictly defined. This is the problem encountered by the designers of a priori sociotechnical innovations intended for the general public: how do you know if the system will respond to the expectations of the targeted users? Can we identify a typology from these users? How do we anticipate future needs in order to best meet the expectations? Such are the questions that designers and funders of digital libraries notably ask. A large number of agencies and private or government schools equip themselves and develop digital libraries. Yet, the findings related to the usages of these digital libraries are often bitter [CRE 06]: very little use in relation to the sums invested; too few studies of the actual behaviors of users; a misunderstanding of the actual difficulties of the different end user profiles; a lack of anticipation with respect to future expectations. When the designers, sponsors and professionals in charge of digital libraries are interviewed, the main disadvantages mentioned involve the technical or regulatory dimensions such as technological obsolescence, the cost related to the maintenance and update of the content, the problem of the rights of authors and of access, interoperability between systems, the increase in the flow of information. The users and their possible difficulties to use digital libraries are not so frequently mentioned. If it is true that certain technological and legal barriers persist, the “human barriers” are essentially the barriers that largely explain the absence of usage, the difficulties and abandonment [DIN 08]. Some authors have particularly focused on the divergences in points of view between the sponsors and the designers of digital environments [CHE 05, CHE 06] by showing, assisted

64

Information Retrieval in Digital Environments

by empirical studies, that the use of ergonomics at the source allows avoiding some of the pitfalls that would have been easy to detect and correct. 3.3.1. “If we built it, they will come” This seems to be the leitmotiv of many designers of digital libraries and policy makers: “If we [the designers] build it, they [the users] will come” [MIC 05]. In one of his last articles that looks like a profession of faith, Michel [MIC 05] adds that the role of designers and decision makers is to “build digital libraries as stars guiding end users toward new horizons that they do not even suspect”. Some designers and policy makers seem, on the one hand, more cautious regarding their mission, and on the other hand, more lucid about the actual usages given to the environments they design. Thus, Wilson [WIL 03] defines the mission of the designers of digital libraries in the following manner: “we are building, with the hope that someone will come, without knowing what we build exactly nor know who will come”. Finally, a few designers of digital libraries do not hesitate to speak of the waste, on the one hand, of the intellectual effort and on the other hand, of the public money when they point out with dismay that most of the digital libraries are not or poorly used [JON 99, WAR 06]. Therefore, if some people continue to think that “the offer will determine the usages”, the reality is quite the opposite when one looks more closely to the uses and performances related to digital libraries. Thus, Kani-Zabihi, Ghinea & Chen [KAN 08] show that 76% of adult users who were asked to search for references to works in a “general public” digital library do not manage to complete this simple task. In addition, 80% of the end users who had participated in their study did not understand different features that were

Information Retrieval: Psycho-Ergonomic Approach

65

proposed on screen. In other words, the performance of users often appears “disappointing” in the eyes of the designers and the sponsors of software related to digital libraries. 3.3.2. The analysis of users’ expectations and behaviors Through a series of studies combining interviews and questionnaires conducted with large cohorts of users, KaniZabihi and his collaborators [KAN 06, KAN 07, KAN 08, KAN 10] have determined that the three main user expectations of digital libraries were the following: the content must be easily accessible and not require specific technical skills or knowledge in documentation; a digital library must provide the same services as a traditional library such as the display of the latest releases acquired, the ability to manage the borrowing of books, etc.; a digital library must be designed in a way that the user will find his/her “references” such as obtaining the assistance of employees. However, these three expectations are rarely found in digital libraries. Some specific expectations have been found by Inskip, Butterworth and MacFarlane [INS 08]. These authors have focused on the expectations of users of a digital library specialized in a very particular area: folk music (the “Vaughan Williams Memorial Library”: http://www.efdss.org/ library.htm). The users interviewed were of different levels: professionals from the world of folk music (specialized journalists, music historians, etc.), teachers and students, performers (musicians, dancers, actors, etc.), and enthusiasts. The main regret of the users interviewed was about the absence of contact with professionals and experts of the same domain. Some also regretted the absence of a forum to share opinions and advice between users of this digital library. The idea of a digital documentalist (“digital librarian”) was then advanced by the users themselves who could make this contact possible.

66

Information Retrieval in Digital Environments

Indeed, documentalists are a (re)source of information essential to users. It therefore becomes important that Web services preserve and make room for these documentalists. Several hints are given by [INS 08]: the possibility of contacting a documentalist by electronic mail; the creation of blogs associated with digital libraries; the possibility to create and manage frequently asked questions (FAQs). Moreover, these authors insist on the fact that for very specific areas (such as folk music), the majority of users interviewed insist on the importance of human contact for the exchange of impressions, tastes and emotions. Indeed, the contact with “real” interlocutors seems essential in the case of digital libraries related to the arts (music, literature, painting, etc.), where affect and emotions are prominent. Similar results had already been found previously [MOY 04]. In parallel, several studies have led to better understanding of the actual behavior of end users of digital libraries. For example, Nicholas, Huntingtin and Jamali [NIC 07] have focused on the exact nature of the information accessed by the users of the ScienceDirect database. They used the technique of analysis of transaction files or “log files” to determine the navigation behaviors of readers – researchers of information. The participants in this study were all researchers and/or teachers – researchers from various disciplines, i.e. experts in their field. The study lasted 18 months and has thus enabled “drawing” with precision the actual behaviors of these users. Based on the data collected, Nicholas, Huntingtin and Jamali [NIC 07] have distinguished five strategies used by these users interacting with the ScienceDirect database: (1) reading only the summary of the articles; (2) reading the full article in html version; (3) reading the full article in pdf version; (4) reading the summary before reading the full article and (5) reading the html version before reading the pdf version. The results show that 20% of the users just read the summaries of the articles they cite subsequently in their

Information Retrieval: Psycho-Ergonomic Approach

67

teachings or productions, this percentage reaching 41% for researchers in the humanities and social sciences. Similar results were found during the interaction of experts with other extremely specialized databases such as Nucleic acids Research: the summaries are a source of information critical to end users and sometimes even, the only source of information. Some studies specifically concentrated on the reasons that made individuals who were “major consumers” of libraries reluctant to use digital libraries. Using interviews with fourteen researchers and teachers – university researchers in the human sciences (History department) and arts (English department), Rimmer et al. [RIM 08] sought to understand why these users did not use digital libraries. Based on declarations of researchers and teachers – researchers, Rimmer et al. [RIM 08], have identified several obstacles to the use of digital libraries for these users: – users are unhappy about the uniformity of the presentations in digital libraries. Indeed, the “books” and “documents” objects are of essential importance to these specialists in history and in arts. One of the pleasures of these experts is to manipulate books and to be in contact with these objects that are very different from each other. However, in dematerializing documents, scanning confers an aspect considered “cold” and “remote”; – when they use a digital library, these same experts no longer have the impression of belonging to a community of users and/or of specialists of a specific domain. Indeed, another of their pleasures recalled is that of reuniting, to sit and to exchange next to the bookshelves of “traditional” libraries. In other words, in a digital library, these experts no longer perceive the social space that is precious to them; – related to the two points previously cited (researchers and teachers) researchers in history and say they do not find the emotions that they have in working in a traditional

68

Information Retrieval in Digital Environments

library: pleasure to glean, to browse publications, to exchange with a colleague or a student around the corner of a bookshelf, etc; – gleaning, which is one of the favorite strategies of these users, is more difficult to achieve in a digital library; – finally, these users say they feel a great deal of difficulty in assessing and judging the historical or literary value of works and documents when these are “dematerialized”. In other words, these are essentially reasons linked to affective and emotional dimensions that explain the reluctance of users to use digital libraries. The physical dimension (in the sense of the sensorial) also seems to be an important characteristic for these experts when interviewed, this dimension being lost in digital environments. One of the most recent studies focused on this interaction between the actual behavior of users and their emotions (or affects) has been carried out by Tenopir et al. [TEN 08]. In their study, these authors have analyzed the behaviors of 41 users, all academics (researchers, documentalists, PhD students), during predefined scenarios of literature research in the ScienceDirect database. One of the distinctive features of this work is that several data collection techniques have been jointly used to obtain a maximum of indicators: (1) the concomitant verbalizations of users, during the proceeding of information retrieval; (2) the audio/video recording of behaviors and actions of users (camera placed in threequarters ahead), in order to collect facial expressions and possible “stalls” of the end user; (3) the monitoring of the path followed during logs analysis. Three indicators have particularly attracted the attention of the authors: the time of each of the phases of information retrieval (“session length”), a phase beginning during the formulation of a query and ending when a new query is

Information Retrieval: Psycho-Ergonomic Approach

69

performed; the pace of the phases of information retrieval, corresponding to the ratio calculated by dividing the number of pages consulted by the time of the session considered; the pauses, namely the instants during which the end user performs no action but processes documents (in the sense of reading and understanding). Any time between two actions was regarded as a pause; this time must be equal to at least twice the average time of an action. If many methodological biases prevent any generalization of the results obtained by the authors in this study, three remarks are significant to mention: – the timings of the sessions of information retrieval are relatively short (on average, 25 min). In other words, on average, an expert end user changes his/her research every 25 min when he/she uses ScienceDirect. However, reading and understanding a scientific article stored in this database requires, a priori, a longer time. We can therefore make the assumption that end users are merely conducting a superficial processing of documents. The use of metatextual skills owned by the expert users may in part explain this apparent “overflight” of documents: in effect, these expert users know where and how collect the relevant information in scientific articles in a speedy and efficient manner (e.g. the first sentences of paragraphs, titles and subtitles) and know how to not pay attention to other, less important information (e.g. some methodological details); – the end users are generally satisfied with the ScienceDirect facility: emotions and positive affects (e.g. “I am certain that the system will find it”, “I have always been satisfied with this system”) are cited significantly more often than the emotions and negative affects (e.g. “it is too vague”, “the system is too slow”). – during a phase of information retrieval, the average number of pauses is 4. Two types of pauses could be observed: the majority of pauses correspond to reading, the

70

Information Retrieval in Digital Environments

assessment and the selection of information and article time; other shorter pauses correspond to the time during which end users are not doing any particular action but they say they are thinking about a new documental query and/or a change in strategy (refining research with specific terms? Expanding with generic terms? etc.). In other words, some pauses are related to content processing while others are related to the process management of information retrieval. The main point of this last study [TEN 08] is twofold: on the one hand, it shows the complementarity of different collection techniques by combining, for example, the concurrent verbalizations and the audio–video recordings; on the other hand, it shows that it is important to associate the quantitative data (performance of users, time to complete the activity, etc.) with qualitative data (impression felt, emotions, etc.) 3.3.3. Prospective innovation

ergonomics

and

technological

The designers of digital libraries are confronted with two major difficulties [DIN 09]: on the one hand, their environment is potentially usable by any individual, all the more so since remote access does not allow to know the identity and characteristics of the end user. It is therefore virtually impossible to define a user profile – typical for which the digital library should be designed; on the other hand, even if some measures demonstrate the effectiveness of the usefulness and usability of a digital device, they do not predict in any manner the adoption of this device. According to Rogers [ROG 95] and Tornatzky and Klein [TOR 82], five elements determine the adoption or the dissemination of new technology: – the relative advantage, which is the degree with which an innovation is perceived as being better than those that

Information Retrieval: Psycho-Ergonomic Approach

71

already exist. It is not necessary that an innovation actually has more benefits than others: what is important is that the individual perceives it as being advantageous; – compatibility, which is the degree to which an innovation is perceived as being consistent with existing sociocultural values, past experiences, social practices and standards of end users. An idea that is incompatible with the values and current standards takes longer to be adopted than a compatible innovation. In some cases, the adoption of a compatible innovation requires the prior adoption of a new system of value, which can take a considerable amount of time; – complexity, which is the estimate of the degree to which an innovation is perceived as being difficult to understand and use. The new ideas that are simple to understand are generally adopted much more quickly than those that require developing new skills before they can be understood; – trialability, which consists of the possibility to test an innovation and to modify it prior to its commitment to use. The opportunity to test an innovation allows prospective users to have more confidence in the product, because they have had the opportunity to learn how to use it; – observability, which refers to the possibility of assessing the costs and benefits of an innovation. The more an individual perceives that he can profit from adopting an innovation (performance, profitability, efficiency, etc.) the more quickly this innovation will be adopted. To these five criteria, Moore and Benbasat [MOO 91] have added the following: an individual adopts an innovation more quickly all the more if this innovation improves his social status. These authors have also distinguished two dimensions related to observability: the visibility of the innovation and the ability to demonstrate the results (“demonstrability”). Hayati and Jowkar [HAY 08] have refined each of the five elements proposed by Rogers [ROG 95] by focusing on digital

72

Information Retrieval in Digital Environments

environments related to information. Thus, these authors propose several sub-components that determine the adoption of an innovation in the field of new technologies related to information and documentation (Table 3.1). – The possibility of using Boolean operators – The possibility of doing researches in specific fields (names of authors, keywords, etc.) – The existence of links to authority files – The simplicity of use Relative advantage

– The possibility of searching large bodies – The possibility of storing and consulting the search history – The existence of hypertext links to primary information and documents – The existence of offline media (CD-Rom) that allow consulting the databases – The response time delay of the system (feedback) – The number of results proposed by the system – Access to full-text

Observability

– Access to databases geographically remote – The possibility of printing the results – The possibility of searching multiple databases simultaneously – The reasonable cost of services

Complexity, testability and compatibility

– The simplicity and user-friendliness of the interface – The existence of a context-sensitive help – The possibility of testing the system with fictitious cases

Table 3.1. Criteria for the adoption of a new technology related to information/documentation, according to Hayati and Jowkar [HAY 08]

Information Retrieval: Psycho-Ergonomic Approach

73

Nevertheless, Hayati & Jowkar [HAY 08] point out some discrepancies between what users say to wish for and real usages. Thus, if context-sensitive help is often cited as valuable by users, the examination of actual behaviors demonstrates that such aid is rarely used although available. Nonetheless, [HAY 08] recall the following points for the designers and sponsors of digital libraries: – there are multiple search strategies in a same system. If it is possible to describe the most optimal strategies and therefore to suggest ways of “how to do things right” to end users, many strategies exist depending on the experience level of the user, his knowledge of the area, his language skills, the context, etc; – overall, end users give less credibility to information retrieved from the Internet and digital databases; – general-public users (i.e. non-experts) often encounter great difficulties to achieve information retrieval, even in simple cases; – some users have a very low digital culture. 3.3.4. Anticipating and understanding the needs of users: the method of staff made up of community experts If some experimental techniques allow us to obtain extremely fine-grained behavioral data concerning the interactions between users and technological devices, they are relatively inefficient to collect data related to the social needs of the users, moreover when these users are difficult to define and when the facility does not yet exist [BRA 06]. The methods called participatory, developed in the 1970s in Scandinavian countries, cover a whole range of techniques that seem to be able to achieve this goal. If the participatory techniques are numerous (e.g. brainstorming, delphi, focus groups, workshops on participatory design, storytelling and

74

Information Retrieval in Digital Environments

consensus conferences), they are based on a similar principle: on the basis of the verbalizations of individuals placed in groups and confronted with a system and/or draft of a system; it is about identifying opinions, dislikes and preferences, this information can be used to (re)define new specifications for services or future products. And as SalazarOrvig and Grossen [SAL 04] have demonstrated within these groups, the generation of speech is quite close to daily conversations, which allows the study of social representations. Of course, the animation of the group of users is paramount since it largely determines the quality of the data collected. Since the end of the 2000s, the participatory methods are experiencing a renewed interest, particularly when it comes to collecting data relating to emotions and social representations associated with existing and/or future digital environments. For example, van Velsen and Melenhorst [VAN 09] have led some focus groups with 11 participants divided into two groups depending on their age and according to their level of experience with the Internet. During these sessions, the authors asked the participants to present and discuss their expectations and needs with respect to four different platforms for sharing online videos (including Youtube©). The authors were particularly interested in the reasons that led users to score and note (“tag”) some videos on these sites. The verbalizations collected during the focus groups helped to distinguish three main reasons behind scoring videos: – scoring has a role of socialization, because it allows the user to be identified by other individuals as sharing the same centers of interest; – scoring has a role of a mnemonic aid, because it is a way to quickly find some videos that the user enjoyed; – scoring is a communication tool, because it is often a pretext for the exchange of opinions via email for example.

Information Retrieval: Psycho-Ergonomic Approach

75

Thus, participatory and creative techniques are increasingly being used to rapidly collect data, even if they must be completed later by more traditional techniques, such as questionnaires. Recently, in the context of a study carried out in response to a request of the Center for European Studies (CVCE; http://www.cvce.eu/), a new method of data collection has been developed: the method of expert community staff [BRA 09, DIN 09]. This method is essentially an adaptation of the technique of focus groups in order to overcome certain limitations [KIT 04] that rely on the following protocol: – definition of the communities of practice concerned. Instead of seeking to form groups with people representative of the average general population, the method of expert community staff begins with a mapping of the communities of practices potentially targeted by the project. To know the communities involved in the project, we must first create some kind of map of the communities in order to identify the actors involved directly or indirectly in the use of the product or service. This consists of gathering the maximum amount of information on the people likely to be affected by the development of the project and thus to determine a profile of the future users of the system or product; – identification of community experts. This phase corresponds to the search of experts that can express themselves legitimately for each community identified. An expert from a practice community is a legitimate representative of the practice community to which he/she belongs; – organization and animation of discussion groups. From the expert panel, each targeted community is represented by at least one group comprising of four to six persons filmed over 3 to 3 hr 30 min. The discussion groups are organized in three phases: (i) free expression of all participants on the themes related to the project; (2) free discussion with some

76

Information Retrieval in Digital Environments

specific media (copies of screen, models, etc.) as support; (3) cards sorting performed by all participants in order to organize the ideas presented; – results analysis. This phase corresponds to the analysis of audio–video recordings, the verbalizations produced by the various participants and the outcomes from sorting cards; – consensus conference. This consists of a standard phase in the management of a process of collective reflection in order to discuss controversial issues and to reach an agreement on common recommendations. The set of results is presented to the sponsors, and consensus is sought for, in particular on all controversial, points of disagreement subject to debate. Therefore, the results are discussed and reinterpreted by the sponsors up to their appropriation. 3.3.5. An example of application of the method of staff made up of community experts European NAvigator (ENA, www.ena.lu) is a digital library specializing in the history of European construction and developed by the Center for European Studies (CVCE). It is currently the most important digital library specializing in this area, utilized by users from all over the world, these users covering a wide range of profiles (students from different disciplines, historians, political scientists, teachers, journalists and history enthusiasts, etc.). The initial demand was to develop the ENA digital library and to focus on communities of users with the objective to offer them a richer service, better adapted and customized using the new opportunities of Web 2.0. In agreement with the method of the community experts, fourteen communities of experts have been identified according to the type of expertise recognized a priori. In total, 58 people (Germans, Belgians, Canadians, Spaniards, French, Luxembourgers) have actually

Information Retrieval: Psycho-Ergonomic Approach

77

participated in these 14 expert staffs who were brought together to (1) generate collectively knowledge about their needs (content expertise) or about their representation of the needs of other users (container expertise), (2) express their needs, expectations, requirements of target users and (3) react to the existing services. The same grid of semistructured interviews was used during the meetings with the various discussion groups and the media used was the present site of the ENA digital library. Finally, at the end of each discussion group, the participants were invited to carry out the task of sorting cards thus providing an image of social representation of usages and needs for each community of users. In the framework of the method of staff made up of community experts, the analysis of the current and/or future needs therefore relies on phases of cooperative production where several people negotiate and validate shared representations of what they are doing, what they say they are doing or would like to do. In the study described briefly here and carried out with the 14 communities of users identified, this generation of representations has given birth to a very large number of ideations (n = 134) which have been grouped into 52 themes. Each theme consisted of 1 to 4 ideas about the improvement of the ENA digital library, some being relatively simple to implement while others were very specific and/or more complex to translate into implementation. A detailed analysis of all the ideations produced [BRA 09a, BRA 09b] has helped to show that the wishes of users could be grouped around seven functions directly related to the ENA digital library (Table 3.1).

Information Retrieval in Digital Environments

New functions expressed by the communities experts staff

Functions traditionally expected from a DL

78

Functions

Definitions

1. Archiving

To classify knowledge in a reasoned, reliable and organized manner in order to make it easily accessible to users by clarifying their rights of usage.

2. Accrediting

To officially recognize the digital library by credible institutions and experts, so as to give it the authority on the knowledge it manages.

3. Updating

To update the information and give a character to the current knowledge on Europe so as to respond to the concerns of users. To develop in a regular and continuous manner the knowledge of European construction but also its culture, and its values.

4. Asserting

Its culture in relation to other digital libraries and therefore seek to differentiate themselves to put forward a specific identity.

5. Associating

To make several instances participate (individual or collective, private or public) in the joint development of European knowledge.

6. Analyzing

To foster understanding of a fact by decomposing, comparing, measuring it, by outlining its cultural referents or contextualizing it in order to produce a new meaning. The digital library must therefore propose analysis indices to identify the historical, geographical, cultural, artistic, social, psychological and political constituents of facts and archived events.

7. Animating

Put in motion the actors of European culture by prompting them to act to generate knowledge and exchange.

Table 3.2. The seven new functions of a digital library [BRA 09a]

There are two important elements: some of the functions correspond to the traditional functions that can be expected of all digital libraries (archiving, accrediting, updating); but other functions appear as being specifically associated with the ENA library because of the peculiarities of its theme and its objectives (associating, affirming, analyzing, animating).

Information Retrieval: Psycho-Ergonomic Approach

79

In other words, the users have widely expressed their desire to see the ENA digital library exceed a simple role of storage and of data archiving: according to the users, this digital library must play a role in animation, in connecting users (and between communities of users), in sharing data between users, in the construction of a Europe of citizens, etc. Associated with the method with staff made up of community experts to identify the expectations and future needs, a “traditional” ergonomic inspection conducted on the basis of the criteria of Bastien and Scapin has been carried out. But the objective of this part was to show how an adaptation of participatory conventional methods could allow obtaining a large number of information, which is difficult to collect when it comes to informational devices used and/or to use by users of multicultural, multisite, multilanguages, and with very variable degrees of expertise. 3.4. The motor dimension The involvement of the body in information retrieval cannot be ignored in the same sense that it is unavoidable in all forms of human–machine interaction [MAC 03]: we must at least generally use the keys on the alphanumeric keyboard, click on buttons or icons, manipulate the mouse or a touch pad, read (and therefore collect), etc. And yet, the biomechanical and sensorimotor aspects are significantly under-represented in the studies focusing on the activity of information retrieval in digital environments. 3.4.1. Motor ability and information retrieval in digital environments Before the emergence and the massive deployment of the Web, the difficulties in using the alphanumeric keyboard and mouse were the subject of studies, especially among young

80

Information Retrieval in Digital Environments

users [EDM 90, ERT 85, HOO 89, SOL 93] until junior high school [EAS 86]. These studies have essentially demonstrated that the use of the keyboard and/or mouse posed enormous difficulties for some children and young adolescents, which may explain, in part, the failures in information retrieval. Despite the familiarization of the youth with the Web, several works have shown that these mobility difficulties were always present among a large number of young users [BOR 99, DIN 02, HUT 06, HUT 05]. In addition, these difficulties regarding the handling of media should not be overlooked since they seem to persist among adult users [ROU 03]. However, other studies tend to show that the use of peripherals, such as the mouse, presents no difficulty for young users. For example, Donker and Reitsma [DON 07a, DON 07b] have compared the performance (speed and efficiency) of children aged from 5 to 7 years old in two types of tasks carried out using a mouse: the children had, on the one hand, to click on the objects that appeared on the screen, and on the other hand, to move objects on this same screen. To move objects (here, letters), two procedures were compared: either the children had to click on the object before clicking on the location where the object should be (operation so-called “click-move-click”); or they had to click on the object and move it while holding down the mouse button (operation so-called “drag-and-drop”). Because, the “drag-and-drop” operation requires two activities in parallel (holding down a mouse button while moving the mouse), it was deemed more difficult to achieve especially for young users [JOI 98, MAC 91, STR 93]. However, the results obtained by Donker and Reitsma [DON 07a, DON 07b] show excellent performance for all participants in the two types of tasks, even if the older succeed better and faster than the younger of course. In addition, unlike the results of previous studies, the performance in the task of moving objects on the screen is better when using the “drag-and-drop” operation.

Information Retrieval: Psycho-Ergonomic Approach

81

Figure 3.3. Screen copy of the software used in the study of Donker and Reitsma [DON 07a]

According to Lane and Ziviani [LAN 10], these apparent contradictions in the results of the studies focusing on the motor dimension are in part explained by the great heterogeneity of hardware, instructions and situations used in the different studies. Also, Lane and Ziviani [LAN 10] have compared the performance of 221 children (aged from 5 to 10 years old) in four types of tasks requiring only the mouse (Figure 3.4): (Figure 3.4(a)) point as quickly as possible objects aligned and/or dispersed on the screen; (Figure 3.4(b)) draw a man; (Figure 3.4(c)) move objects by “drag-and-drop” and by “click-move-click” and (Figure 3.4(d)) follow an object moving on the screen. Three major results have been achieved by Lane and Ziviani [LAN 10]: – there are very great intra-individual variabilities: indeed, for the same child, his/her performance is very different depending on the task whatever his/her age. For example, some children perform well in the tasks of pointing (Figure 3.4(a)) but have poor results in the tasks of following objects (Figure 3.4(d));

82

Information Retrieval in Digital Environments

– cognitive maturity and proprioceptive skill (i.e. level of intellectual and kinesthetic development) justify performance more than civil age itself; – the experience of children with the mouse, in terms of frequency of usage at home, is a decisive factor in their performance.

Figure 3.4. Tasks used in the study of Lane and Ziviani [LAN 10]

Another reason explaining the divergences between the results and the findings made by the authors is the confusion between the use of the keyboard and/or the mouse as artifacts on the one hand, and on the other hand, the manipulation of signs and information associated with these artifacts. Indeed, the studies that register difficulties in handling the keyboard suggest above all difficulties in the

Information Retrieval: Psycho-Ergonomic Approach

83

handling of linguistic codes and in the ordering of the alphabet letters on the keyboard (“alphabetizing”) whereas the studies that do not find any major difficulty suggest particularly simple motor gestures related to the typing and clicking. If the main focus of all these studies is to study sensorimotor coordination (and not only one of the two dimensions), they have a limited scope regarding the retrieval of information since the tasks that the participants had to achieve in these studies are extremely simple (e.g. clicking on the squares that appear randomly on the screen, moving an object using the mouse, pointing and drawing). Today, the interest of the community of researchers for these motor dimensions involved in information retrieval seems to have disappeared. More accurately, this interest has moved: – toward teaching content to improve the uses of traditional devices (alphanumeric keyboard and mouse) among young users (orientation of educational psychology [REY 03, SNO 09]); – toward the biomechanical and sensorimotor aspects associated with the use of traditional devices among the elderly [LAU 02, SAN 05, SMI 99]; – toward forms of multimodal interaction to search for information, for example, by combining the written and spoken word [LEB 07, LEB 10] or even of brain – machine interfaces [LEB 06]; – toward new forms of physical interactions allowed with certain digital interfaces that just seem to “free” from the constraints related to the manipulation of traditional devices (multitouch, interactive tablets, etc.). These studies on the new forms of interaction information retrieval are mainly concentrated on the public for whom motor ability may be a problem, such as young users [CIO 12], seniors [LAU 02,

84

Information Retrieval in Digital Environments

PIP 11] or people with disabilities [AST 10, COO 01]. In other cases, it is the nature itself of the information (sound files, video files) and the context (e.g. collaborative work) that force to devise new forms of interaction to search for and share information [CUY 08, TZA 04] (Figure 3.5);

Figure 3.5. Information retrieval through gesture and voice, according to Cuypers et al. [CUY 08]

– toward the impacts of media on the behavior of information retrieval. For example, the ability to search for information on mobile phones significantly changed postural attitudes, gestures and readability (Figure 3.6). To assist users, solutions are currently being tested, such as those that allow zooming on some parts of the screen by simple “click-drag” [CHU 11, KWO 11].

Figure 3.6. Constraints and solutions for information retrieval on mobile phones, according to Kwon, Choi and Chung [KWO 11]

Information Retrieval: Psycho-Ergonomic Approach

85

Although interesting, these studies generally concentrate on techniques and subjects, without carrying out analyses of behaviors and practices. In addition, when tests are performed with users, these are often adults (most of the time, students in computer science or in psychology). In the framework of work combining computing development and ergonomic psychology, we are actually attempting to develop and enrich the lexicon of gestures allowed and usable with new digital media. 3.4.2. Toward a lexicon of intuitive gestures The recent development of tools based on the multitouch interaction mode such as tablets or interactive mobile phones is in large part due to their simplicity of use [OMA 04]. By “simplicity”, it should be understood that most individuals can use and interact with these tools without prior training regarding to the required gestures. This ease of use therefore relies on the implementation of a gesture lexicon that individuals create and understand outside of digital media. For example, to enlarge and/or reduce the size of an object on a multitouch screen, it is generally sufficient to move further apart and/or bring closer the thumb and index finger, this gesture appearing as intuitive (Figure 3.7). If the implementation of a lexicon of gestures in digital environments brings along computing problems [DEL 10, KAL 10], it also raises questions regarding psychological and ergonomic areas [WOB 09]: what is an “intuitive” gesture? How do we quickly indicate to a user(s) the relevant gesture(s) to perform? Are all the gestures transferable to digital media?

86

Information Retrieval in Digital Environments

Figure 3.7. Reducing and/or enlarging an object on a multitouch screen

From a psychological and neuropsychological point of view, performing a gesture results from an interaction between information collected from the environment and mental representations of gestures stored in an internal register [ROT 91, ROT 97]. According to these authors, one of the main input channels for the extraction of information from the environment based on the visual analysis of the stimulus (object). The stimulus is analyzed via structures that compare it to stimuli already known and therefore stored (input lexicons, system of recognition of objects). If the stimulus is recognized, the semantic representation of the action/object is enabled in the semantic system; then, the appropriate action is found again through a lexicon of exit actions; and finally, the motor response is programmed and triggered. There should therefore be activation of the relevant actions related to the vision of an object through an analysis of the permitted actions and/or authorized with this object. This approach is somewhat reminiscent of the theory of affordances of Gibson [GIB 77, REE 82] applied to digital environments [GAV 91] and that finds a direct application in the ergonomic criterion of the significance of codes and denomination of the objects that appear on a screen [BAS 92, BAS 01, SCA 97]. From a psychological point of view, any language is an ability that develops without necessarily requiring any

Information Retrieval: Psycho-Ergonomic Approach

87

explicit learning, voluntary and conscious. Indeed, whether it be verbal or sign language, the human individual acquires and develops language skills by imitation, repetition and a “linguistic bath”. Any language practitioner (verbal or gestural) therefore implicitly acquires the fundamental skill that consists of understanding what a statement and/or a gesture mean: this understanding is achieved by learning the associations that bind a signifier (a word, a gesture, a sound) to a signified. With multitouch screens, the gestures that their users must perform to manipulate objects on the screen correspond to gestures previously learned in other contexts (e.g., scroll through the pages using the index on a screen reminds the same gesture made to flick through a “paper” book); the users of interactive tablets or mobile phones therefore reproduce gestures acquired with other environments and with other media. Because, some of our gestures are culturally determined, the authors seek to establish a grammar of gestures that are “universally intuitive” (e.g. [LEH 10]). Currently, we are only in the design phase and the enrichment of the lexicon of intuitive gestures implementable on multitouch technologies (such as the ipad©); [BER 11, VIV XX]. Nevertheless, the interest is twofold: from an IT perspective, the challenge is to “lift the computing barriers” since some apparently simple gestures are still difficult to model and formalize on multitouch tablets or mobile phones (e.g., to draw a spiral); from a point of view of psychology ergonomics, the challenge is to be able to offer interfaces for which the interactions are intuitive and therefore very inexpensive cognitively. Once the implementation of gestures is complete, further user tests will be carried out notably with children to check the intuitiveness of these gestures. In a later phase, the extension of these modes of multitouch interaction in documental databases is planned in order to examine the

88

Information Retrieval in Digital Environments

relevance and the effectiveness in tasks of information retrieval. Indeed, multitouch interaction is currently permitted for very simple actions related to information retrieval (e.g. browsing through folders, selecting a portion of information). But what is really happening in multitouch interaction during a full information query? 3.5. The social dimension and collaborative More and more often, the activity of information retrieval involves collective and/or working groups. Indeed, for timesaving reasons and increased efficiency, individuals frequently collaborate to search for information at all levels (school, junior high school, secondary school, college, private enterprise, etc.). Therefore, at school, pairs of pupils must most often carry out information retrieval on the Web for pedagogical reasons (i.e. socio-constructivist conflict and coconstruction of knowledge) and for pragmatic reasons related to the lack of material and time [DIN 07]. However, if some empirical studies [BHA 00, COC 01, DIA 04] tend to show that collaborative research is more efficient than information retrieval performed individually (time saving, greater number of visited sites, reduction in the number of revisited sites, better organization of visited sites, etc.), the results are rarely generalizable because of too-restricted numbers of pupils and an imprecise methodology. In addition, these studies usually only focus on populations of students and/or adults. Paradoxically, if most authors involved in collaborative research of information on the Internet recognize the primary role of affective relationships (friendship, enmity, indifference, etc.) between collaborators [DUM 01, HAN 05], no study has sought to truly test the impacts of these relationships on the behavior, performance and interactions between collaborators when searching for information. Also, four issues directly related

Information Retrieval: Psycho-Ergonomic Approach

89

to the collaborative dimension of information retrieval seem crucial to us: – On the one hand, is collaborative information retrieval on the Internet really different, and on the other hand, more efficient than individual research, particularly when it comes to young users? – Does the degree of affinity between the members of a pair affect their behaviors and strategies? – What are the impacts at the interaction level of the degree of affinity between the two collaborators? – Can we, and if yes how, help the collaborators who have to perform collective information retrieval in digital environments such as the Web? 3.5.1. From individual information retrieval

research

to

collaborative

Hansen and Järvelin [HAN 05] therefore define collaborative information retrieval as any task of the same type as problem solving, involving several individuals interacting, synchronously or asynchronously, during a common task of research of Websites or Web pages in contexts more or less defined and in more or less open environments”. This definition appears to be very generalist since it encompasses all of the contexts and devices that may be related to the information retrieval. Currently, one of the most used typologies to report and account for the diversity of situations of collaborative information retrieval is the one created by Twidale, Nichols and Paice [TWI 97]. According to these authors, the activities and situations of collaborative information retrieval can be distinguished according to two orthogonal axes: a spatial axis allows distinguishing the

90

Information Retrieval in Digital Environments

activities according to the spatial distance between the collaborators; a temporal axis distinguishes the activities in which individuals are collaborating synchronously or asynchronously. These two axes are a continua along which a particular situation can be positioned and thus defined. The approach of Twidale, Nichols and Paice [TWI 97] presents two advantages that present all the characteristics of a paradox: – on the one hand, these authors suggest not to confuse the activity of collaborative information retrieval with digital environments (i.e. the tools). Indeed, post-it notes or even a printed dictionary are systems that can withstand a collaborative activity of information; – on the other hand, these authors tend to match tools to situations of collaborative information retrieval. Furthermore, the telephone is “classified” among the devices that support remote synchronous collaborative activities. However, we often physically leave messages destined for our close collaborators (asynchronous situation with physical proximity). Similarly, if electronic mail allows communicating with the whole world in an asynchronous manner, it is mainly with our closest collaborators and of whom we expect a quick response that we exchange the greatest number of messages. 3.5.2. Benefits and limitations information retrieval

of

collaborative

The studies conducted on collaborative information retrieval are especially focused on the research on the Web since this environment has revived the interest in this activity. More specifically, many of these studies have endeavored to demonstrate the superiority of the collaborative dimension on the individual dimension. This supposed superiority is reflected particularly in the title of

Information Retrieval: Psycho-Ergonomic Approach

91

the article published by Lazonder [LAZ 05] entitled “two heads search better than one”. Indeed, a large amount of jobs have effectively demonstrated that collaborative information retrieval on the Web improves the performance of the users [BHA 00, COC 01, DIA 04, DIN 07, DUM 01], particularly with regard to the amount of relevant information found and the time taken to perform the search. More concretely, performing a collaborative research of Web pages has the following advantages: – the total time needed to search for information on the Web decreases; – the volume of information processed and read by these individuals increases significantly; – the information found on the Web looks better organized; – the number of revisited pages decreases significantly when the search for information is collaborative. Indeed, when information retrieval on the Web is carried out by a single individual, the number of visited pages that have already been previously viewed is considerable: according to the studies, the percentage of pages revisited vary from 30% to 61%, with a maxima of 92% for certain individuals! However, revisiting pages inevitably implies loss of time, time that the individual could use to visit Web pages that he/she had not yet viewed. But, the results from previous studies are difficult to compare, and extrapolations are also difficult to achieve because the situations described, the populations involved and the scenarios are very heterogeneous. In addition, a few studies tend to show that collaborative information retrieval on the Web presents three major disadvantages [LIP 99, NUR 99]:

92

Information Retrieval in Digital Environments

– The work of information research is often very unequally distributed between members of the same group. In this case, how to assess the level of personal investment of each collaborator in the collective effort? – The members of a same group sometimes have very different representations of the space-problem; – The interindividual relationships (emotional) seem decisive in the final performance [DIN 07]. 3.6. Impact of emotional ties between collaborators Information retrieval is a human activity in which the social component is extremely strong. Indeed, even when an individual searches for information alone in front of his computer and the Internet, his/her activity is rooted in a social context that pre-exists him/her and that influences him/her (e.g. social origin of the request, school context or professional). In addition, the tools and digital resources that this end user manipulates are also determined from a social point of view. However, in this part, we are interested in the impact of the inter-collaborator social dimension, namely the impact of the relationships that exist between individuals that have to search, together, for information on the Web. Karamuftuoglu [KAR 98] was one of the first authors to emphasize that the search for information in digital environments is rarely a “private and solitaire” activity but that it, on the contrary, involves several individuals in a collaborative process. In school, academic or professional contexts, situations in which individuals must collaborate to search for information in digital environments are increasingly common. The reasons are of an educational nature (e.g. socio-constructivist conflict) or practice (i.e. lack of computer, dispersion of resources).

Information Retrieval: Psycho-Ergonomic Approach

93

For Dillenbourg [DIL 99], the collaboration and interactions that it causes (explanations, disagreements or mutual regulations) trigger additional cognitive mechanisms with respect to the selection of the information and their transformation into knowledge and may allow a reduction in the cognitive load. According [DIL 99] again, four conditions are necessary to be able to speak of collaborative situation: – any collaborative situation, i.e. a common task, must exist; – any interaction of a varying degree of collaboration between the members of the group must also exist; – any process of collaborative learning must be identified. – finally, any effect of the collaborative research must be present for the group as well as for each individual constituting this group. However, if these four conditions are necessary, they are not sufficient [DIN 08]. For example, the fact that a common task exists is not enough to assert that collaborative work exists since there is no indication that there is similarity and homogeneity in the representations of the purpose for each collaborator. Moreover, several works have shown that the same instructions could give rise to multiple mental representations depending on the individuals belonging to the same group of work (for a synthesis: [DIN 02]). Furthermore, researching, or performing the search for information in a collaborative manner requires management and/or group management skills that end users do not likely seem to always possess. However, a new current of multidisciplinary research named CIB for “collaborative information behavior” appeared during the last ten years [HYL 06, RED 08] in order to better understand the behavior, the mental processes underlying the factors and the impacts of the collaborative dimension of information retrieval, particularly in digital environments such as the Internet.

94

Information Retrieval in Digital Environments

3.6.1. Ties between collaborators information retrieval

and

impact

on

Several studies tend to show that the affective ties that exist between collaborators that must retrieve information together on the Web influence their behavior and their performance [CRO 98, CRO 99, LAZ 05, HYU 05, VAS 02]. Indeed, if the two (or more) collaborators maintain friendly ties between them, the research for information tends to be performed more quickly and more effectively. On the basis of these different studies, an author even makes the case for collaborative information retrieval insisting on the fact that “two heads search better than one” [LAZ 05]. Yet, two remarks can be made against these studies: on the one hand, these studies that show the interest of the existence of emotional ties between collaborators do generally only concentrate on adult users (students); on the other hand, the samples of participants are very reduced: these studies, although very rich in teachings, are better seen as empirical case studies than experiments. Also, in a series of experimental studies conducted with young users (enrolled in primary school), we attempted to determine the real impact of emotional ties between individuals who need to collaborate together in order to search for information on the Web [DIN 08, DIN 11, DIN 07]. In all these studies, our protocol was substantially identical. Some students had to search for information on the Web relatively to the themes covered in class (e.g. medieval history) following three successive conditions, each condition being separated by fifteen days: the student must perform the activity alone (condition “ALONE”); the student must perform the activity with another student whom with he/she maintains friendly ties (condition “AFFINITY + ”); the student must again perform the activity with another student, but this time, no friendly tie exists between the two students (condition “AFFINITY - ”).

Information Retrieval: Psycho-Ergonomic Approach

95

The teachers were the ones who have helped to create the pupil pairs for the collaborative sessions. Of course, the sessions, the questions and the items were counterbalanced to avoid possible ordering effects. Regarding the students’ performance (number of good answers given, time to perform the activity, number of Web pages viewed), the results obtained in our studies consistently demonstrate the following two points: – the students individually looking for the information on the Web (condition “ALONE”) are less efficient and less effective than when they work in pairs, regardless of the degree of affinity within this tandem. In addition, the examination of the requests carried out by the participants shows that pupils commit significantly more spelling errors when they search individually for information on the Web. These results thus confirm the interest in making young end users collaborate during information retrieval on the Web; – the pairs are more efficient and more effective when the students do not maintain friendly ties between themselves (“Affinity -”). Similarly, the pairs carry out significantly more queries not regarding the theme of the information retrieval if friendly ties exist between pupils. These two results therefore tend to show that the existence of a friendly link between young collaborators may be detrimental to their effectiveness and performance. A fine analysis of verbal interactions between students during the research for information has been conducted to better understand the reasons that can explain this apparent “negative” effect of the friendship on the performance. The analyses of verbal interactions, conducted on the basis of the grid established by Hmelo-Silver [HME 03], have thus served to highlight two interesting results: when the students are looking for information on the Web with a friend (condition “AFFINITY +” ), there is significantly more

96

Information Retrieval in Digital Environments

interactions reflecting conflicts between the two collaborators; when students are looking for information with another student whom with no friendly link does exist a priori (condition “AFFINITY -”), there is significantly more facilitative interactions being produced. Collaborative information retrieval on the Web seems therefore to present advantages for young end users (enrolled in primary schools): on the one hand, with regard to the selection of the information, working in pairs improves the performance; on the other hand, as regards the phase of documental query (i.e. production of queries), collaborative work resulted in a reduction of spelling errors, as if the “other” played the role of supervisor and corrector. On the other hand, the benefits related to the collaborative situation during a search for information on the Web seem to disappear if the members of a pair who must achieve together the activity maintain friendly ties. Indeed, in the case of two friends collaborating to search for information on the Web, the performance decreases (e.g. decrease in the amount of relevant information found, increase in the number of non-relevant requests) and their verbal interactions are mainly centered on conflicts and on solving conflicts. Conversely, when the pairs are formed by two students not maintaining any particular friendly tie, the performance increases and verbal interactions are especially focused on the task and its facilitation. Even if several methodological biases exist in our experiments, the results obtained lead to the questioning of several points: – at the theoretical level, it is appropriate to further reflect on the way in which we may include emotional aspects in recent models supposed to take into account the behavior and the mental processes involved in collaborative research;

Information Retrieval: Psycho-Ergonomic Approach

97

– at the educational level, it is necessary to ask ourselves about the factors and dimensions that should be necessary and/or possible to take into account to form working groups. Generally, working groups are in practice formed by letting students choose themselves: and very logically, they choose to work with a friend; – at the ergonomic level, it is important to think about the systems and/or devices that can help young users to perform the activity of information retrieval in a collaborative manner. Indeed, as we will see now, the tools generally used to carry out collaborative research are tools created for individual and “solitary” usages. 3.6.2. “RCI-Web”: retrieval

software

to

assist

information

As a result of the previously cited studies, it appeared that it was relevant to design a tool that can facilitate and/or assist information retrieval when it is performed in a collaborative context (e.g. group work). Collaborative information retrieval is typically carried out with tools originally designed to individually achieve this activity. Thus, in France, the search engine Google© is the one predominantly used in schools, junior high schools, secondary schools and colleges to achieve collaborative information retrieval, including in an asynchronous manner. However, this search engine is a tool designed for individual uses. Also, end users who have to search for information along with others on the Web must use other systems to share and/or inform of their lookups (e.g. electronic mail, forum). There are a few tools devolved to the collaborative research of information (e.g. Yoono©, SearchTogether©, TeamSearch©). However, these tools are not intuitive, usually require subscriptions, the installation of additional software, and appear to be more similar to tools to share

98

Information Retrieval in Digital Environments

resources within [VIV 07, VIV 09].

the

previously

defined

communities

In other words, there were no simple and intuitive tools to assist end users who had to search for information on the Web in a group. More specifically, there were no tools that enabled the members of a working group to carry out information retrieval on the Web that knew the sites and/or pages that the other members of his/her group had already consulted. However, this lack of knowledge with regard to the conduct of other members of a work group generates a huge number of revisits (up to 92%), resulting in loss of time and efficiency. This is the reason that motivated the development of an innovative software program named RCI-Web (for “Recherche Collaborative d’Information sur le Web” or “Collaborative Information Retrieval on the Web”), and its source code has been registered with the copyright agency for the protection of programs [VIV 07, VIV 09]. RCI-Web consists of two separate applications: – the first application is based on an agent embedded in the web browser and in the Google© search engine. It has, as main features, noting visited pages, viewing the notes already assigned by the group of collaborators, displaying the list of pages listed in a given research thematic and restituting the notes directly in the pages of the search engine; – the second application is a management and activity tracking tool that allows setting a research project and tracking the evolution, behavior and contributions of all the participants. This second part is therefore intended in particular to the teacher who wishes to follow a posteriori the involvement of each member of the working groups in a collective project.

Information Retrieval: Psycho-Ergonomic Approach

99

Very specifically, after being connected and activating the plug-in (Figure 3.8), the RCI-Web agent is associated with the browser through a task bar that is superimposed on the displayed page. We find on this bar (retractable if necessary) a button of presentation of the history of the research (Figure 3.9), a note area and a menu to manage options. When an collaborator visits a Web page, he/she has the ability to “note” the content simply by clicking on one of the proposed levels (e.g. notes). This action has for consequence to update the history of the research and to change instantly the presentation settings of the page within the search engine. This information is then available to all persons who will be associated with the same research theme (Figure 3.9).

Figure 3.8. RCI-Web home page

Figure 3.9. Examples of the interfaces of RCI-Web

100

Information Retrieval in Digital Environments

An end user may at any time consult the results of previous research (personal or of other members of his/her group) by displaying the data, either in relation to chronological criteria (consultation dates) or in relation to relevance criteria (note assigned), or according to the collaborators of the project. This filter simplifies the representation of results and minimizes the interactions between end user and the application. To test the usefulness and usability of RCI-Web, several experimental studies have been carried out with various different publics concerned with collaborative information retrieval [DIN 05b, VIV 09]. A first study was conducted with 40 students in their 4th year of biology (average age = 23.4 years old). These 40 students were divided into 10 groups of four members. Within each group, each student had to search for relevant Web pages related to three themes pertaining to biology concepts; these themes have been proposed based on the opinion of five experts (teachers) in biology who have distinguished them according to their level of difficulty: – Theme “easy” (easy): “What are the risks of epidemics of deadly flu and what measures are taken to prevent these epidemics?” – Theme moderately solutions and interests”;

easy

(medium):

“Green

roofs:

– Theme difficult (difficult): “The effects of anthropization on Posidonia beds and their ecosystem”. Depending on the conditions, students could or could not use RCI-Web. The results have mainly shown that the use of RCI-Web had had significant and beneficial impacts on (1) the time it took to perform the activity especially when the theme of the information retrieval is difficult, and (2) the percentage of revisits to Web pages, especially, when the theme is difficult. In other words, when RCI-Web is used by

Information Retrieval: Psycho-Ergonomic Approach

101

students to carry out collaborative information retrieval, these end users lose much less time to revisit sites and/or pages already consulted by their collaborators. A second study was conducted with 24 students of the 8th grade (average age = 13.4 years old). Each of these participants was invited to find answers to the factual questions (e.g. “On what date did Julius Caesar die?”, “Which emperor burnt down Rome?”, “What are the seven hills of Rome”), with the help of the Web, during three sessions: during a first session, each student should respond only to predefined questions, using the Google© search engine (situation of individual information retrieval); during a second session, the students had to work in pairs to answer further questions still using the Google© search engine (situation “in pairs without RCI-Web”); finally, during a third session, the same pairs of students had to search again for the answers to new questions using the Google© this time assisted by RCI-Web (situation “in pairs with RCI-Web”). The order of these three sessions was counterbalanced with regard to the participants, each session will drop down to a week of interval during the same time slots. The results obtained with these pairs of 8th grade pupils were similar to those obtained with 4th year students: on the one hand, the time taken to find the answers to the questions was significantly reduced when RCI-Web was being used; on the other hand, the number of revisits has also decreased significantly when RCI-Web was used. From a pedagogical point of view, our results show that the collaborative selection of Web pages can be improved with the help of a very simple technical device that provides information relating to Web pages that each member of the group, to which an individual belongs, consults. Indeed, by knowing which Web pages had already been visited by a member of the group and knowing the opinion he/she expressed in relation to this page (relevant vs. non-relevant), a student or a pupil can devote his/her time to the

102

Information Retrieval in Digital Environments

examination of other Web pages, on the one hand, reducing the total time to complete the task, and on the other hand, the number of unnecessary revisits. 3.7. The cultural dimension The joint development of new technologies, of globalization and of media has allowed the development of the “global village” [MCL 67]. If the Web is a worldwide network, it is subject to the cultural influences of the territories that host it. Since the beginning of the 1990s, many works have concentrated on the influence of culture on the design of Websites [NIE 90, BAR 98, CHA 02, EVE 97]. Generally, these studies have focused on superficial aspects such as the colors presented in the pages, the images used, or even the symbols, usually in relation to the behavior or the preferences of consumers [COL 00, CYR 04, CYR 05]. From the ergonomics and interfaces design point of view, the growing interest in the cultural and ethnic dimensions has thus given rise to a current of design centered on culture [KAM 06, SHE 06]. Among the pages of the Websites, we think the home page is the one that has the greatest importance. 3.7.1. About the importance of the home page If the home page of a Website is so important, it is because it determines the first impression and conditions, in part, the future behavior of users (Figure 3.10). Indeed, it is essentially the home page of the Website that a user first encounters that will encourage this user to enter or not the said site ([CYR 08, KIM 07, KIM 08]; for a summary, [KAR 11]), especially if it is an unknown site. We can then understand why the design and the concept of home pages are of vital interest in e-commerce. Indeed, one of the challenges for companies is that the end user remains long

Information Retrieval: Psycho-Ergonomic Approach

103

enough on the home page of their site to motivate them to enter.

Figure 3.10. The importance of the first impression in the process of information retrieval, according to Kim and Fesenmaier [KIM 08]

A research team at Microsoft Research has recently proposed a mathematical modeling explaining the behaviors of abandonment of users ([LIU 10]). After having analyzed the behaviors of thousands of users on more than 200,000 Websites, this team has determined that the time spent by an individual on a Web page followed the Weibull distribution. The Weibull distribution was initially been designed to account for the reliability of technical systems; more specifically, the distribution of Weibull allows estimating the probability that a component of a technical system is malfunctioning at a moment t then that component has worked correctly until the moment t. Thus, this distribution allows predicting the time at which the component must be replaced. In the case of navigation on the Web, the component failure corresponds to the user abandoning the Web page. According to the Weibull distribution, we can distinguish a “positive aging” and a “negative aging” to explain a failure: – “positive aging” is the most logical since it predicts that the risk of failure of a component is a direct function of its lifetime (t): in other words, the longer a component is functioning (i.e. is being used) the greater the risk of failing;

104

Information Retrieval in Digital Environments

– “negative aging”, on the other hand, predicts that the risk of failure of a component is inversely proportional to its lifetime (t): in other words, the longer a component is being used the less risk there is to malfunction. This result is explained by the fact that, in this case, the component would have proved its robustness. In the case of navigating on the Web, this would be essentially the “negative aging” that would explain the behaviors of users, particularly when these are abandoning a page and/or a Wesite [LIU 10, NIE 12]. More definitively, the probability that an individual leaves a Web page decreases when the time that this individual remains on the said Web page increases (Figure 3.11). And it appears that the first 10 s are critical. In other words, if an individual remains more than 10 s on the home page of a Website, then the probabilities that this individual explores the site increase considerably. In this case, the attention of the end user has been captured and “hooked”. We then realize the challenges related to the home pages of Websites since these are the home pages that will determine the behavior of the individual: enter the site or exit immediately.

Figure 3.11. Weibull distribution applied to Web navigation, according to Liu, White and Dumais [LIU 10]

Information Retrieval: Psycho-Ergonomic Approach

105

3.7.2. Culture and design of Websites home pages: an ergonomic inspection If superficial aspects are undeniably crucial, we have recently been trying to concentrate on fundamental dimensions, such as the nature and the function of the information presented on Web pages. Because, home pages are crucial, we propose to study the cultural differences between home pages of Websites and their impacts on the preferences and behavior of individuals. According to Nielsen [NIE 99, NIE 03], a home page contains the following areas of information by descending order of surface occupied (Figure 3.12): – tools to assist hypertextual links);

navigation

(20%,

e.g.,

frames,

– the contents themselves (20% ); – empty space (20%, i.e. surface not used); – the navigation system and related tools (19%); – self-promotions, generally taking the form of accountings expressing the greatest achievements of the company and/or its Website (9%); – “filling”, typically images or videos aimed to improve the aesthetic (5%); – the company logo, the welcoming text (5%); – advertisements (2%). However, Nielsen [NIE 03] presupposes that the distribution of the different types of information on home pages is universal. Yet, it seems that some home pages are very different, for the same company and/or the same service, depending on the culture of the targeted end users and the culture of the designers.

106

Information Retrieval in Digital Environments

Figure 3.12. Distribution of the different types of information on a home page, according to Nielsen [NIE 03]

To check if home pages really present different designs according to culture, a careful examination of the French and Japanese home pages of 21 large international companies have been carried out [DIN 11]. These 21 companies are: Honda, Lancôme, Mercedes, Bridgestone, Canon, Asahi, Michelin, Fisher Price, Toyota, Fujifilm, Suzuki, Renault, Volvo, Toshiba, Panasonic, Mitsubishi, Mazda, Nintendo, Sony and Peugeot. The choice focused on the Japanese and French versions since it appears that the western and fareastern sites are perceived as the most radically opposed [CYR 04, CYR 05, KAR 11], this perception exists for all types of environments [ISO 11]. On the one hand, this careful consideration shows that the Japanese home pages actually contain more information (essentially pictorial) that the French versions, as well as a

Information Retrieval: Psycho-Ergonomic Approach

107

greater number of hypertextual links. On the other hand, the distribution in terms of occupied surfaces by different types of information defined by Nielsen (2003) is very different depending on the version of the home pages (French vs. Japanese; Figure 3.13): the Japanese pages contain many more ads, self-promotions, navigation tools and content that the French pages; the French pages contain significantly more “aesthetic filling” (filler), generally flash videos; finally, the unused surface (empty or white space) is much more significant in the case of the French versions. In other words, the Japanese home pages are much denser from the informational point of view while the French home pages are much “emptier” and essentially centered on aesthetic animation.

Figure 3.13. Distribution of the different types of information presented on a home page of a Wesite depending on the version: French vs. Japanese [DIN 11]

3.7.3. Information retrieval culture and behavior navigation Therefore, there is an objective difference in the design of French and Japanese home pages. However, what about the impacts of these differences, on the one hand, on the

108

Information Retrieval in Digital Environments

preferences of users, and on the other hand, on their behaviors during visual exploration? A preliminary study has been conducted with six master’s students in psychology [DIN 11]. Each of these students presented five versions of home pages of Websites from six large enterprises (Mercedes-Benz, Samsung, General Motors, Shiseido, Canon and Lancôme): versions in German, Korean, French, Russian and Japanese. For each of the six companies, each participant simply had to classify the home pages according to their preference in descending order. The results show that there is no correlation between the preferences of the participants with regard to the sites of Mercedes-Benz, Samsung and General Motors, these three companies being German, Korean and American, respectively. On the other hand, the preferences of the participants were significantly correlated in the case of the sites of Shiseido, Canon and Lancôme, the first two companies being Japanese and the last being French. In other words, the preferences of the participants were very similar for these three companies. In addition, for these three companies (Shiseido, Canon and Lancôme), the French home pages have systematically been the classified in the first position (i.e. preferred) while the Japanese home pages have systematically been classified in fifth position. In other words, our French participants have always preferred the French versions of home pages and have always rejected the Japanese versions of the same home pages. These first results encourage us to continue our investigations about the relationship between cultural factors and the preferences of end users. More recently, a new research program aims to test the impact of individual factors related to culture on the behaviors during visual exploration of the home pages of Websites. Indeed, since the way that an individual explores

Information Retrieval: Psycho-Ergonomic Approach

109

his/her environment is partly determined by his/her culture [NIS 03, NIS 05], we are currently undertaking experiments to test the impact of the cognitive style of exploration of his/her environment (analytical vs. holistic) on the visual exploration of Websites home pages, the scientific literature tending to show that this cognitive style is related to our culture [MAR 91, MAS 01, NIS 01]. Overall, this intercultural dimension is of interest for a growing number of researchers on the one hand and on the other hand, manufacturers and/or designers of computing environments. 3.8. The visual exploration strategies Because information retrieval primarily requires vision and because the techniques for the collection of accurate data as for visual exploration of its environments have improved considerably (e.g., oculometers called “free head”, mobile oculometers), today there is a large number of studies that are concentrating on the involvement of vision related with “electronic reading” (in the words of the title of a book by [BAC 04]). Regarding the examination of the visual paths of information retrieval on screen, two types of complementary approaches exist [BAS 10, DRU 11]: the bottom-up approach and the top-down approach. The bottom-up approach or ascending represents the majority of the studies focusing on the exploration of visual environments, particularly digital. Since well before the advent of the Internet, many works generally from cognitive ergonomics have studied the impact of visual characteristics of digital environments on the behavior and performance during tasks requiring reading on screen. These studies have essentially focused on the impact of page layouts, typography, fonts or background colors, reading, understanding or even the memorization of materials presented on screen (e.g. [BER 03, DUC 83, LIN 02, PEA 03, SCO 93].

110

Information Retrieval in Digital Environments

The top-down or descending approach brings together the studies focusing on the interindividual differences in visual explorations; these studies began in the early 1970s [NOT 71]. These authors quickly found that the factors related to hardware do not explain by themselves the different strategies of visual exploration; thus, it was appropriate to give more attention to stimulus-independent individual differences. Nevertheless, it is only since the 2000s that several studies have really established that some of the indicators of visual exploration, such as the sizes and frequencies of eye saccades, differed considerably between individuals, regardless of the stimuli used (e.g., [AND 99, CAS 08, RAY 07]). Since then, many studies have focused on studying the impact of individual variables on the exploration of visual materials, these variables either related to culture [BLA 08, CHU 05, MAS 08], pathologies [ELI 12, RIB 08, RIB 09], or even to factors linked to personality [RIS 12]. However, as Ling and van Schaik pointed out [LIN 06], if a very large number of these studies (top-down or bottom-up) usually originate in cognitive ergonomics focused on “electronic reading”, few are concerned with the strategies of visual exploration when searching for information. Indeed, the tasks that the participants must have achieved in the previously cited studies are generally tasks of target detection (objects, words, non-words), or of reading with the objective of memorizing content in recalling tasks, or of reading in order to understand relatively simple content. Among the few studies that concentrated on studying the impact of material characteristics on visual exploration during information retrieval, we can cite those of [LIN 06]. These authors have studied the impact of four widths of texts presented on screen (55, 70, 85 and 100 characters per line) and of two types of fonts (Arial, Times) on performance during information retrieval with 99 participants (average age = 24 years old). The participants had to find, within texts presented on several Websites specifically designed for the

Information Retrieval: Psycho-Ergonomic Approach

111

study, the answers to eight factual questions, these questions always being displayed at the top of the pages. The results show that none of the two manipulated factors (width of the texts and type of fonts) have an impact on the speed with which the answers are found and on the number of pages visited. In other words, the environmental factors that generally affect the detection reading and visual exploration of content in digital environments do not seem to have any impact on the performance in a real information retrieval task. If the studies concentrating on the impact of visual characteristics of the content on the behaviors can be distinguished according to their conceptual approach (bottom-up/top-down), they can also be distinguished according to their methodological approach [BAS 10, DRU 11]: – most works are based on the a priori examination of strategies of visual exploration. In these cases, the content (e.g. Web page, interface, software screen) are segmented into areas corresponding a priori to areas of interest, and the researchers examine how pathways, saccades and eye fixations are divided within these areas and in between these areas; – more recently, a posteriori studies have been focusing on the behavior of visual exploration. In these cases, no zone is predefined and researchers examine the visual pathways with the intent of discovering invariant forms (patterns) that can distinguish categories of visual strategies. One of the main focuses of these studies is to consider that a behavior of visual exploration arose from the interaction between the superficial characteristics present in the environment (e.g. configuration of screens, location of images and location of the menus) and psychological factors (e.g. domain expertise, procedural expertise and motivation). Some studies that we have recently carried out perfectly illustrate these two complementary axes.

112

Information Retrieval in Digital Environments

3.8.1. Impact of the typographical marking (bottom-up approach) In the lists of search engine results (called search engine results page or SERP), words appear with a specific typographical formatting (Figure 3.14, ie. bold and/or capitalization) and this, regardless of the search engine. The keywords appearing typographically formatted correspond to the words used by the end user to query the Web. Typographic formatting is supposed to help the reader to detect more quickly the keywords in the answers offered by the search engine.

Figure 3.14. Example of a results page (SERP)

Information Retrieval: Psycho-Ergonomic Approach

113

During experimental studies focusing on selection behaviors regarding Websites, we have shown that young users were greatly influenced by typographic formatting present in search engines (e.g. [DIN 03, DIN 05a, DIN 05b, ROU 11]). In fact, far from helping young users, typographical formatting encourages them to adopt a simple visual-lexical detection strategy, to the detriment of genuine processing of the content displayed. For example, in one of our first experimental studies [DIN 99], some pupils (5th and 7th grade) were asked to select from lists of document results the most relevant references with regard to familiar themes and unfamiliar topics (e.g. “the French Revolution”). Among these document results, half were relevant (e.g. “Freedom, Equality, Fraternity in France: analysis of the Revolution”) while the other half were totally irrelevant in relation to the predefined themes (e.g. “French cuisine: a true revolution”). In addition, among these references, some had the significant keywords (e.g. “revolution” and “French”) formatted in bold or not. The results showed that the youngest children (5th grade, average age = 9.5 years old) selected above all references containing keywords typographically formatted, regardless of the semantic relevance of such references. The older participants (enrolled in 7th grade, average age = 12.1 years old) selected predominantly the references on the basis of their semantic relevance, independently of the presence or the absence of typographical formatting. In other words, the youngest readers had a tendency to make their selection on the basis of area indices (here, typographical formatting) at the expense of a genuine reading comprehension, especially in the case of unfamiliar themes. Other subsequent studies sometimes involving up to 250 students (for a summary see [DIN 02]) have confirmed these results. Typographic formatting does not help young users in

114

Information Retrieval in Digital Environments

their task of reading selection of documental references and/or Websites, since these young users “merely” detect the keywords typographically formatted. However, if this assumption was the one we followed for a number of years, our studies did not allow testing the hypothesis that young users used to focus on typographically formatted keywords. Through to the use of oculometric techniques, we have been able to analyze with accuracy the visual exploration of the search results pages from search engines (or SERP for search engine results page). Our main result obtained from 89 pupils aged 10 to 16 years old [DIN 10] was to show that there are four visual exploration strategies (Figure 3.15). This result contradicts the assertions of Nielsen and Pernice [NIE 10], or at least moderates them considerably.

Figure 3.15. The four visual exploration strategies of a search engine page [DIN 11]

Information Retrieval: Psycho-Ergonomic Approach

115

According to Nieslen and Pernice [NIE 10], an SERP page is explored visually in the same way as any Web page. In addition, according to these authors, the so-called “F-shaped” strategy is the one that is predominantly being used (not to say exclusively). However, the analysis of visual pathways and visual fixations shows the following results: – there is not only one but four visual exploration strategies for a results page from a search engine; – the so-called “F-shaped” strategy: it corresponds to that described by Nielsen and Pernice [NIE 10] but remains in the minority among our participants; – the “exhaustive” strategy: in this case, the end user explores (reads) all the information contained on the page (title, summary, URL address); – the strategy “by visual jumps”: here, the user explores the SERP page in jumps, by “skipping” keywords displayed with formatting; – the “F-inverse” strategy: in this case, the strategy is similar to that of the “F-shaped” strategy but proceeds upside down (the entry point is at the bottom right, and then the end user goes up gradually, etc.); – the distribution of these strategies is different depending on the school level (and accordingly the age) of users (Table 3.3). More specifically, there seems to be an evolution of the strategies in the sense that the exhaustive strategy becomes the mostly used among the elder children (from 14 years old) while the strategy by visual jumps is the most frequently used among the youngest children. Despite the bias and the methodological limitations of our studies, the oculometric techniques appear to be promising a better understanding of the behavior of visual exploration of search engines. And today, these studies are still very rare (with the exceptions: [AUL 05, CUT 07a, CUT 07b]).

116

Information Retrieval in Digital Environments

Grade 5 (n=26)

Strategy F-shaped strategy Exhaustive strategy Visual jumps F-inverse strategy

7 (n=28)

F-shaped strategy Exhaustive strategy Visual jumps F-inverse strategy

9 (n=19)

F-shaped strategy Exhaustive strategy Visual jumps F-inverse strategy

11 (n=16)

F-shaped strategy Exhaustive strategy Visual jumps F-inverse strategy

Familiar 13% 11% 61% 15% 100% 14% 15% 59% 12% 100% 12% 58% 9% 21% 100% 11% 68% 10% 11% 100%

Unfamiliar 13% 15% 54% 18% 100% 11% 9% 68% 12% 100% 13% 53% 24% 10% 100% 14% 54% 19% 13% 100%

Table 3.3. Distribution of visual strategies according to the school level

The implications of these studies are numerous: – from the theoretical point of view, they provide new insights about the cognitive processes involved in the visual exploration of digital environments. In addition, they add a differentialist vision because of the interindividual differences and developments that seem to exist in these processes; – from the educational point of view, they question the teaching and/or learning practices related to information retrieval through search engines under a new angle, since these engines are the most used tools to access information on the Web [JAN 00]; – from the point of view of ergonomics, they question in a frontal way the relevance of techniques developed by the designers of search engines supposed to help the user (here, typographical formatting).

Information Retrieval: Psycho-Ergonomic Approach

117

With regard to the hypothesis of a possible evolution of the visual exploration strategies with age and/or the level of cognitive development, a longitudinal study is also being undertaken with junior high school pupils. Nonetheless, our results raise the question about the origin of strategic interindividual differences in visual exploration processes. Furthermore and indirectly, these same results also question the possible existence of underlying cognitive styles that can explain these differences. 3.8.2. Impact of the mental model (top-down approach) As it has been outlined in the previous sections, a number of environmental factors (e.g. typographic formatting) influence the visual exploration of search engines, at least among young users. Parallel to the environmental factors, certain individual factors also seem to influence these strategies of visual exploration and of information retrieval in digital environments. Among the individual factors, most likely to greatly influence the behaviors in searching for information, the mental model is often referred to in particular among young users [GRE 06, YAN 06, YAN 09]. In the perspective of Johnson-Laird [JOH 83, NOR 83], the mental model generally referred to corresponds to the mental representation that the end user has of the way in which a system operates, of the parties that make up this system and the interactions between these components [FEI 93]. Overall, studies in this field show that the richer the user model with regard to digital environments, the greater its performance during the information retrieval in a digital environment [LAZ 00, ROO 93, SOL 93, SLO 02, SUH 92, SUT 00, THA 98, ZHA 08]. Even if he/she has not directly studied the mental representation of the end user, Marchionini is one of the

118

Information Retrieval in Digital Environments

authors who has contributed most relatively to the impacts of the expertise (of the domain and procedural) in information retrieval [MAR 92, MAR 95, MAR 87, MAR 90, MAR 91, MAR 93] (for a summary, see [DIN 02]). These jobs are located in the extension of the numerous studies in cognitive psychology that have committed themselves to comparing the behavior and performance of individual experts vs. novices (e.g. in resolution tasks of complex problems such as chess games, algebra or more widely learning). However, because a mental model is by definition internal, unstable and difficult to “achieve”, the collection of its related information is a real problem. In the studies mentioned above, “traditional” techniques such as maintenance, observation or concomitant verbalization have been mainly used. But these techniques are difficult to use with young users: lack of vocabulary, low ability of abstraction, misunderstanding of instructions and disruption of the activity, etc. [DEV 09, DOR 11, LEG 01]. Also, some authors have used drawings to obtain the mental model of the end user, particularly among the youngest [GRA 90, PAP 05, PEJ 98]. Indeed, drawings are one of the modes of expression preferred by youths and their analysis allows obtaining a number of information that are beyond the capabilities of linguistic expression of children. Also, in a recent study [DIN 11], 51 children were invited to perform three activities: fill up a standardized questionnaire [YAN 05, YAN 06, YAN 09] in order to assess their knowledge about information retrieval on the Web; freely draw a picture according to the single following instruction: “draw what is the Internet to you”; carry out two searches for information on the Web in order to respond to two predefined questions (“In what city was Louis XVI arrested?” and “What treaty transforms the EEC into the European Union?”).

Information Retrieval: Psycho-Ergonomic Approach

119

As a result of these tests, four ergonomics experts were invited to classify the drawings made by the participants in our study according to the categories established by Zhang [ZHA 08]. Each expert carried out this classification without knowing the classification of other experts.

Figure 3.16. The six categories of mental models of young users [DIN 11]

The main results of this study is as follows: while the four categories of mental models established by Zhang [ZHA 08] have been successfully found, the four experts further

120

Information Retrieval in Digital Environments

wished to create two new categories, thus bringing the number of mental models pertaining to the Internet among young people to six (Figure 3.16): – the technical view (Figure 3.16(top left)): in this case, the Internet is reduced to its simple medium (i.e. the computer, the screen), thus reflecting a very simplistic vision of the Internet only envisaged from the perspective of the devices; – the centered view on the search engine (Figure 3.16(top right)): here, the end user considers that the search engine is the Internet. The search for information is therefore perceived as the central activity on the Internet; – the view centered on features (Figure 3.16(middle left)): in this case, the end user no longer considers the Internet only from its technical point of view but apprehends it through the features and services to which the Internet gives access; – the view centered on the networks (Figure 3.16(middle right)): the technical dimension of the Internet has completely disappeared, in favor of an approach centered on the dimension of social networks; – the technical-functional view (Figure 3.16(bottom left)): in this new category, either elements relating to media (i.e. screen, computer) as well as relating to accessible services and/or features have been identified; – the functional view connectionist (Figure 3.16(bottom right)): in this second new category, elements translating a vision centered on the functionalities used on the Internet and social networks are mixed together. In addition, it appears that the distribution of these six mental models is different according to the two age classes of our participants. Indeed, the vision centered on the search engine is the most widely found among the youngest (43.4%) while the view centered on functionalities is the most predominant in the older groups (42.8%). Finally, the mental model seems to have an impact on the performance of

Information Retrieval: Psycho-Ergonomic Approach

121

participants. Indeed, even if this is only a tendency, it is interesting to note that the longest execution times concern the participants that have a technical view of the Internet among the two age groups. Similarly, the completion time of the activity is the shortest in cases where the participants have a view centered on the functionalities of the Internet. These first encouraging results should prompt them to continue in this type of studies involving oculometric techniques. Indeed, from a theoretical and methodological point of view, it seems interesting to combine techniques to analyze the strategies of visual exploration and analysis of drawings in order to evaluate the possible relationships between the end user’s mental model, his or her visual exploration strategy and his/her performance during information retrieval. In the medium term, it is of course the interaction between environmental and individual factors that is questioned when searching for information on the Web.

Conclusion

The main objective of this book was to present, in a summary manner, works undertaken for several years relating to the behaviors and cognitive processes involved in information retrieval (IR) in digital environments. More specifically, the aim was to present a few examples of theoretical models and studies to better understand the difficulties, behaviors and strategies of individuals searching for information in digital environments. From a theoretical point of view, it appears undeniable that the ergonomic psychology approach (theoretical and methodological) is able to contribute to enriching the existing theoretical models presented in the first part of this book. In addition, the ergonomic psychology approach is relevant in helping to design new integrative and global models (e.g. [NAH 07]). Indeed, since the activity of IR in digital environments brings together and cuts across a set of human dimensions (cognitive, motor, cultural, social and emotional), it seems appropriate to develop conceptual frameworks capable of apprehending these different interacting dimensions. From a methodological point of view, it appears relevant to combine different techniques of behavioral data collection (e.g. analysis of the activity, drawings, hierarchical analysis

124

Information Retrieval in Digital Environments

of the task, interviews, oculometry and physiological data) even to contribute in developing original methods (e.g. expert staff methods). Indeed, while “traditional” techniques from ergonomic psychology are still relevant in the case of a posteriori analyses with well-targeted audiences, it is important to devise others for the case of a priori analyses with difficult to identify audiences. Again, from a methodological point of view, it becomes absolutely necessary to associate the experimental studies carried out in laboratories (i.e. variables manipulation, a priori control of variables, focus on some behavioral indicators) with more “ecological” studies carried out on the ground. In addition, while the increase of scientific knowledge relating to human behavior during IR motivates a large number of current works, the applications and implications derived from this theoretical knowledge are also very promising and “concrete”. Indeed, through close collaborations with a number of colleagues from several disciplines (especially engineering sciences), it is possible to change or even design the entire or parts of sociotechnical systems therefore allowing us to assist end users during the activity of IR, especially when it comes to publics with specific needs (very young users, seniors, end users with sensory or motor deficits, etc.). As discussed in the first chapter of this book, from the psychological point of view, the activity of IR in digital environments presents several characteristics that make it unique, complex and fascinating, at the same time. Furthermore, it concerns a research thematic that, by nature, is located at the crossroads between psychology, ergonomics, human–machine interaction (HMI) and IR. Also, research about the borrowing of theoretical and/or methodological knowledge from each of these scientific fields must constantly strive in order to succeed in delineating the complexity of the psychological mechanisms involved.

Bibliography

[AFN 93] AFNOR, Vocabulaire de la documentation, AFNOR, Paris, 1993. [AHU 01] AHUJA J.S., WEBSTER J., “Perceived disorientation: an examination of a new measure to assess web design effectiveness”, Interacting with Computers, vol. 14, pp. 15–29, 2001. [AND 99] ANDREWS T.J., COPPOLA D.M., “Idiosyncratic characteristics of saccadic eye movements when viewing different visual environments”, Vision Research, vol. 39, pp. 2947–2953, 1999. [ARA 08] ARAPAKIS I., JOSE J.M., GRAY P.D., “Affective feedback: an investigation into the role of emotions in the information seeking process”, SIGIR ’08: Proceedings of the 31st Annual International ACM SIGIR Conference on Research and Development in Information Retrieval, ACM, New York, NY, pp. 395–402, 2008. [ARA 09] ARAPAKIS I., KONSTAS I., JOSE J.M., “Using facial expressions and peripheral physiological signals as implicit indicators of topical relevance”, Proceedings of the 17th ACM International Conference on Multimedia, Springer, Santa Barbara, CA, USA, pp. 461–470, 2009. [AST 10] ASTELL A.J., ELLIS M.P., BERNARDI L., et al., “Using a touch screen computer to support relationships between people with dementia and caregivers”, Interacting with Computers, vol. 22, no. 4, pp. 267–275, 2010.

126

Information Retrieval in Digital Environments

[AUL 05] AULA A., MAJARANTA P., RÄIHÄ K.J., “Eye-tracking reveals the personal styles for search result evaluation”, Proceedings of INTERACT 2005, LNCS 3585, pp. 1058–106, 16 September 2005. [BAC 04] BACCINO T., La lecture Universitaire de Grenoble, 2004.

électronique,

Presses

[BAC 08] BACCINO T., SALMERON L., CANAS J., “La lecture des hypertextes”, in TRICOT A., CHEVALIER A. (eds.), Ergonomie des documents électroniques, Presses Universitaires de France, Le Travail Humain, Paris, pp. 1–16, 2008. [BAL 99] BALAKRISHNAN K., BOUSQUET O., HONAVAR V., “Spatial learning and localization in rodents: a computation model of the hippocampus and its implications for mobile robots”, Adaptive Behavior, vol. 7, no. 2, pp. 173–216, 1999. [BAL 08] BALMISSE G., “Recherche d’information en entreprise: une question de gouvernance”, in DINET J. (ed.), Usages, usagers et compétences informationnelles au 21ème siècle, Hermès Science Lavoisier, Paris, pp. 71–95, 2008. [BAR 94] BARRY C.L., “User-defined relevance criteria: an exploratory study”, Journal of the American Society for Information Science, vol. 45, pp. 149–159, 1994. [BAR 98a] BARBER W., BADRE A., “Culturability: the merging of culture and usability”, Human Factors and the Web, available at www.research.att.com/conf/hfweb/proceedings/barder/index.htm, 1998. [BAR 98b] BARRY C.L., SCHAMBER L., “Users’ criteria for relevance evaluation: a cross-situational comparison”, Information Processing and Management, vol. 34, pp. 219–236, 1998. [BAS 92] BASTIEN J.M.C., SCAPIN D.L., “A validation of ergonomic criteria for the evaluation of human-computer interfaces”, International Journal of Human–Computer Interaction, vol. 4, pp. 183–196, 1992. [BAS 01] BASTIEN J.M.C., SCAPIN D.L., “Évaluation des systèmes d'information et Critères Ergonomiques”, in KOLSKI C. (ed.), Systèmes d’information et interactions homme-machine, Environnements évolués et évaluation de l’IHM, Interaction homme-machine pour les SI, Hermes, Paris, vol. 2, pp. 53–79, 2001.

Bibliography

127

[BAS 09] BASTIEN J.M.C., BRANGIER E., DINET J., et al., “The expert community staff: an innovative method for capturing end-users’ needs”, in NORROS L., KOSKINEN H., SALO L., et al. (eds.), Designing Beyond the Product: Understanding Activity an User Experience in Ubiquitous Environments, European Conference on Cognitive Ergonomics (ECCE 2009), Springer Verlag, Berlin, pp. 374–379, 2009. [BAS 10] BASTIEN J.M.C., DRUSCH G., DINET J., “Connaître les comportements des internautes: méthodes et outils”, Actes du Séminaire INRIA, L’usager numérique, ADBS, Anglet, Paris, pp. 39–62, 27 September–1 October 2010. [BAU 09] BAUER A., WOLLHERR D., BUSS, M., “Information retrieval system for human-robot communication: asking for direction”, Proceedings of the 2009 IEEE International Conference on Robotics and Automation (ICRA ’09), IEEE Press Piscataway, NJ, pp. 1522–1527, 2009. [BEA 96] BEAULIEU M., ROBERTSON S., RASMUSSEN E., “Evaluating interactive systems in TREC”, Journal of the American Society for Information Science, vol. 47, no. 1, pp. 85–94, 1996. [BER 03] BERNARD M.L., CHAPARRO B.S., MILLS M.M., et al., “Comparing the effects of text size and format on the readability of computer-displayed Times New Roman and Arial text”, International Journal Human–Computer Studies, vol. 59, pp. 823–835, 2003. [BER 11] BERTOLO D., VIVIAN R., DINET J., et al., “Les interfaces tactiles pour des situations collaboratives: vers de nouvelles grammaires de gestes permis par le multi–touché, in SALEH I. (ed.), Hypermédias et Pratiques numériques (H2PTM ’2011), Hermès Science-Lavoisier, Paris, pp. 347–359, 2011. [BHA 00] BHARAT K., “SearchPad: explicit capture of search context to support web search”, Proceedings of the 9th International World Wide Web Conference (WWW9), Amsterdam, The Netherlands, 15–19 May 2000. [BIL 98] BILAL D., “Children’s search processes in using World Wide Web search engines: an exploratory study”, Proceedings of the 6th Annual Meeting of the American Society for Information Science, vol. 34, pp. 45–53, 1998.

128

Information Retrieval in Digital Environments

[BIL 02] BILAL D., KIRBY J., “Differences and similarities in information seeking: children and adults as web users”, Information Processing and Management: An International Journal, vol. 38, no. 5, pp. 649–670, 2002. [BIL 07] BILAL D., BACHIR I., “Children’s interaction with crosscultural and multilingual digital libraries ii: information seeking, success, and affective experience”, Information Processing and Management: An International Journal, vol. 43, no. 1, pp. 65–80, 2007. [BLA 02] BLACKMON M.H., POLSON P.G., KITAJIMA M., et al., “Cognitive walkthrough for the Web”, ACM Conference on Human Factors in Computing Systems (CHI 2002), pp. 463–470, 2002. [BLA 07] BLACKMON M.H., MANDALIA D.R., POLSON P.G., et al., “Automating usability evaluation: cognitive walkthrough for the Web puts LSA to work on real-world HCI design problems”, in LANDAUER T.K., MCNAMARA D.S., DENNIS S., et al. (eds.), Handbook of Latent Semantic Analysis, Lawrence Erlbaum Associates, pp. 345–375, 2007. [BLA 08] BLAIS C., JACK R.E., SCHEEPERS C., et al., “Culture shapes how we look at faces”, PLoS ONE, vol. 3, e3022, 2008. [BOE 00] BOEKHORST A.K., Informatievaardig worden in het onderwijs, eennformatiewetenschappelijk perspectief: Een vergelijkende gevallenstudie in Nederland en Zuid-Afrika [Becoming information literate in education, an information science perspective: a comparative case study in The Netherlands and South Africa], Unpublished dissertation, 2000. Available at www.hum.uva.nl/ albert/public/prom-akb-tot.PDF. [BON 06] BONACCIO S., DALAL R.S., “Advice taking and decisionmaking: an integrative literature review, and implications for the organizational sciences”, Organizational Behavior and Human Decision Processes, vol. 101, no. 2, pp. 127–151, 2006. [BOR 86] BORGMAN C.L., “The user’s mental model of an information retrieval system: an experiment on a prototype online catalog”, International Journal of Man-Machine Studies, vol. 24, pp. 47–64, 1986.

Bibliography

129

[BOR 99a] BORGMAN C.L., “The user’s mental model of an information retrieval system: an experiment on a prototype online catalog”, International Journal of Human–Computer Studies, vol. 51, pp. 435–452, 1999. [BOR 99b] BORGMAN C.L., HIRSCH S.G., WALTER V.A., et al., “Children’s searching behavior on browsing and keyword online catalogs: the science library catalog project”, Journal of the American Society for Information Science, vol. 46, no. 9, pp. 663–684, 1999. [BOR 03] BORLUND P., “The IIR evaluation model: a framework for evaluation of interactive information retrieval systems”, Information Research, vol. 8, no. 3, pp. 152–164, 2003. [BRA 05] BRAND-GRUWEL S., WOPEREIS I., WALVAREN Y., “Information problem solving: analysis of a complex cognitive skill”, Computers in Human Behaviour, vol. 21, pp. 487–508, 2005. [BRA 06a] BRANGIER E., BASTIEN J.-M.-C., “L’analyse de l’activité est-elle suffisante et/ou pertinente pour innover dans le domaine des nouvelles technologies?”, in VALLERY G., AMALBERTI R. (eds.), L’analyse du travail en perspectives: influences et évolutions, Octarès, Collection “Entreprise, Travail, Emploi”, Toulouse, pp. 143–156, 2006. [BRA 06b] BRAND-GRUWEL S., WOPEREIS I., “Integration of the information problem-solving skill in an educational programme: the effects of learning with authentic tasks”, Technology, Instruction, Cognition and Learning, vol. 4, pp. 243–263, 2006. [BRA 08] BRAND-GRUWEL S., GERJETS P., “Instructional support for enhancing students’ information problem solving ability [Special issue]”, Computers in Human Behavior, vol. 24, no. 3, pp. 1–17, 2008. [BRA 09a] BRANGIER E., DINET J., BASTIEN C., “La méthode des staffs d’experts de communauté: orientation théorique, démarche méthodologique et application pratique à la reconception d’une bibliothèque numérique sur la connaissance de la construction européenne”, Document Numérique, vol. 12, pp. 111–132, 2009.

130

Information Retrieval in Digital Environments

[BRA 09b] BRANGIER E., DINET J., EILRICH L., “The 7 basic functions of a digital library: Analysis of 14 focus groups about the usefulness of a digital library on the history of European construction”, in SMITH M.J., SALVENDY G. (eds.), Human Interface, Springer-Verlag, Berlin, pp. 345–354, 2009. [BRA 09c] BRAND-GRUWEL S., WOPEREIS I., WALRAVEN A., “A descriptive model of information problem solving while using Internet”, Computers & Education, vol. 53, no. 4, pp. 1207– 1217, 2009. [BRA 11] BRAND-GRUWEL S., STADTLER M., “Solving informationbased problems: valuating sources and information”, Learning and Instruction, vol. 21, pp. 175–179, 2011. [BRO 00] BROCH E., Children’s search engines from an information search process perspective, available at http://www.ala.org/aasl/aaslpubsandjournals/slmrb/slmrcontent s/volume32000/childrens, 2000. [BRU 94] BRUCE H.W., “A cognitive view of the situational dynamism of user-centered relevance estimation”, Journal of the American Society for Information Science, vol. 45, pp. 142– 148, 1994. [CAS 08] CASTELHANO M.S., HENDERSON J.M., “Stable individual differences across images in human saccadic eye movements”, Canadian Journal of Experimental Psychology, vol. 62, pp. 1– 14, 2008. [CHA 00] CHAN H.C., TAN B.C.Y., WEI K.K., “Three important determinants of user performance for database retrieval”, International Journal of Human-Computer Studies, vol. 51, pp. 895–918, 2000. [CHA 02] CHAU P.Y.K., COLE M., MASSEY A.P., et al., “Cultural differences in the online behavior of consumers”, Communications of the ACM CACM Homepage Archive, vol. 45, no. 10, 2002. [CHE 05] CHEVALIER A., “Evaluer un site Web: les concepteurs et les utilisateurs parviennent-ils à identifier les problèmes d'utilisabilité?”, Revue d'Intelligence Artificielle, vol. 19, pp. 319–338, 2005.

Bibliography

131

[CHE 06] CHEVALIER A., KICKA M., “Web designers and Web users: influence of the ergonomic quality of the Web site on the information search”, International Journal of Human-Computer Studies, vol. 64, no. 10, pp. 1031–1048, 2006. [CHE 07] CHEVALIER A., DOMMES A., MARTINS D., et al., “Searching for information on the Web: role of aging and ergonomic quality of website”, in JACKO J. (ed.), Human-Computer Interaction, Part I, HCII 2007, LNCS 4550, pp. 691–700, 2007. [CHE 11] CHEN J., LIN C., YEN D.C., et al., “The interaction effects of familiarity, breadth and media usage on web browsing experience”, Computers in Human Behaviors, vol. 27, pp. 2141– 2152, 2011. [CHI 85] CHI M.T.H., GLASER R., “Problem solving ability”, in STERNBERG R.J. (ed.), Human Abilities: An Information Processing Approach, Freeman, New York, NY, pp. 227–257, 1985. [CHI 07] CHIARAMELLA Y., MULHEM P., “La recherche d’information. De la documentation automatique à la recherche d’information en contexte”, Document Numérique, vol. 10, pp. 11–38, 2007. [CHO 07] CHOI I., KOO M., CHOI J.A., “Individual differences in analytic versus holistic thinking”, Personality and Social Psychology Bulletin, vol. 33, no. 5, pp. 691–705, 2007. [CHU 05] CHUA H.F., BOLAND J.E., NISBETT R.E., “Cultural variation in eye movements during scene perception”, Proceedings of the National Academy of Sciences of the United States of America, vol. 102, pp. 12629–12633, 2005. [CHU 11] CHUNG M.K., LEE D.H., JEONG C.H., “The effect of zoomable user interfaces and user age in searching for a target with a mouse on a two-dimensional information space”, International Journal of Industrial Ergonomics, vol. 41, no. 2, pp. 191–199, 2011. [CIO 12] CIOCCA G., OLIVO P., SCHETTINI R., “Browsing museum image collections on a multi-touch table”, Information Systems, vol. 37, no. 2, pp. 169–182, 2012.

132

Information Retrieval in Digital Environments

[COC 01] COCKBURN A., MCKENZIE B., “What do web users do? An empirical analysis of web use”, International Journal of Human-Computer Studies, vol. 54, pp. 903–922, 2001. [COL 00] COLE M., O’KEEFE R.M., “Conceptualizing the dynamics of globalisation and culture in electronic commerce”, Journal of Global Information Technology Management, vol. 3, no. 4, pp. 4–17, 2000. [COM 95] COMPEAU D.R., HIGGINS C.A., “Application of social cognitive theory to training for computer skills”, Information Systems Research, vol. 6, no. 2, pp. 118–143, 1995. [COO 01] COOLEY P.S., ROGERS S.M., TURNER C.F., et al., “Using touch screen audio-CASI to obtain data on sensitive topics”, Computers in Human Behavior, vol. 17, no. 3, pp. 285–293, 2001. [COO 06] COOPER D., “Knowledge workers”, Canadian Businesses, vol. 79, no. 20, pp. 59–63, 2006. [CRE 06] CREASER C., HAMBLIN Y., DAVIES J.E., “An assessment of potential efficiency gains through online content use”, Program: Electronic Library and Information Systems, vol. 40, no. 2, pp. 178–189, 2006. [CRO 97] CROOK C., “Computers in the community of classrooms”, in LITTLETON K., LIGHT P. (eds.), Learning with Computers: Analysing Productive Interaction, Routledge, London, pp. 103– 117, 1997. [CRO 98] CROOK C., “Children as computer users: the case of collaborative learning”, Computers & Education, vol. 30, no. 3, pp. 237–247, 1998. [CUT 07a] CUTRELL E., GUAN Z., An eye-tracking study of information usage in Web search: variations in target position and contextual snippet length, Technical report for Microsoft Research, MSR-TR-2007-01, 2007. [CUT 07b] CUTRELL E., GUAN Z., “What are you looking for? An eye-tracking study of information usage in Web Search”, Proceedings of CHI 07, Human Factors in Computing Systems, ACM Press, San Jose, CA, pp. 407–416, April 2007.

Bibliography

133

[CUY 08] CUYPERS T., SCHNEIDER J., TAELMAN J., et al., “Eunomia: toward a framework for multi-touch information displays in public spaces”, Proceedings of BCS-HCI 2008, Australia, pp. 31– 34, 2008. [CYR 04] CYR D., TREVOR-SMITH H., “Localization of Web design: an empirical comparison of German, Japanese, and U.S. website characteristics”, Journal of the American Society for Information Science and Technology, vol. 55, no. 13, pp. 1–10, 2004. [CYR 05] CYR D., BONANNI B., BOWES J., et al., “Beyond trust: website design preferences across cultures”, Journal of Global Information Management, vol. 13, no. 4, pp. 24–52, 2005. [CYR 08] CYR D., KINDRA G.S., DASH S., “Web site design, trust, satisfaction and e-loyalty: the Indian experience”, Online Information Review, vol. 32, no. 6, pp. 773–790, 2008. [CZA 01] CZAJA S.J., SHARIT J., OWNBY R., et al., “Examining age differences in performance of a complex information search and retrieval task”, Psychological Aging, vol. 16, pp. 564–579, 2001. [CZA 06] CZAJA S.J., CHARNESS N., FISK A.D., et al., “Factors predicting the use of technology: findings from the center for research and education on aging and technology enhancement (CREATE)”, Psychological Aging, vol. 21, no. 2, pp. 333–352, 2006. [DAV 05] DAVENPORT T.H., Thinking for a Living: How to get Better Performance and Results from Knowledge Workers, Harvard Business School Press, Boston, MA, 2005. [DAV 07] DAVID P., SONG M., HAYES A., et al., “A cyclic model of information seeking in hyperlinked environments: the role of goals, self-efficacy, and intrinsic motivation”, International Journal of Human-Computer Studies, vol. 65, no. 2, pp. 170– 182, 2007. [DEL 10] DE LA GUÍA E., GALLUD J.A., TESORIERO R., et al., “Co-interactive table: a new facility to improve collaborative meeting mobile”, Proceedings of HCI 2010, Lisbon, Portugal, 7–10 September 2010. ACM 978-1-60558-835-3/10/09, 2010.

134

Information Retrieval in Digital Environments

[DEV 09] DEVINE-WRIGHT H., DEVINE-WRIGHT P., “Social representations of electricity network technologies: exploring processes of anchoring and objectification through the use of visual research methods”, British Journal of Social Psychology, vol. 48, pp. 357–373, 2009. [DIA 04] DIAMADIS E.T., POLYZOS G.C., “Efficient cooperative searching on the Web: system design and evaluation”, International Journal of Human-Computer Studies, vol. 61, pp. 699–724, 2004. [DIL 99] DILLENBOURG P., “What do you mean by collaborative learning?”, Collaborative-Learning: Cognitive and Computational Approaches, Elsevier, Oxford, G.B, pp. 1–19, 1999. [DIN 01] DINET J., PASSERAULT J.-M., ROUET J.-F., “La recherche documentaire informatisée à l’école: vers une modélisation des processus cognitifs liés au jugement de pertinence des références documentaires chez les élèves de CM2”, in DE VRIES E., PERNIN J.-Ph., PEYRIN J.-P. (eds.), Hypermédias et Apprentissages, Actes du cinquième colloque, EPI-INRP, Paris, pp. 135–150, 2001. [DIN 02] DINET J., ROUET J.-F., “La recherche d’information: processus cognitifs, facteurs de difficultés et dimension de l’expertise”, in PAGANELLI C. (ed.), Interaction hommemachineet recherche d’information, Hermès, Paris, pp. 133–161, 2002. [DIN 03] DINET J., “La recherche documentaire informatisée à l’école: désactivation en mémoire et difficultés de sélection de références pertinentes”, Psychologie Française, vol. 3, pp. 3–17, 2003. [DIN 04] DINET J., PASSERAULT J.-M., “La recherche documentaire informatisée à l’école”, in PAUL V., PERRIAULT J. (eds.), Hermès CNRS, vol.39, Paris, 2004. [DIN 05a] DINET J., “Typographie et sélection de sites Internet”, Proceedings of IHM 2005, Toulouse, France, pp. 132–145, 27–30 September 2005.

Bibliography

135

[DIN 05b] DINET J., “La sélection collaborative de pages Web pertinentes”, in TCHOUNIKINE P., JOAB M., TROUCHE L. (eds.), Actes de la conférence sur les Environnements Informatiques pour l’Apprentissage Humain (EIAH), INRP, ATIEF, Montpellier, pp. 347–352, 25–27 May 2005. [DIN 07a] DINET J., “Deux têtes cherchent mieux qu'une? Oui, mais …”, Médialog, vol. 63, pp. 38–41, 2007. [DIN 07b] DINET J., SIMONNOT B., LECLÈRE P., et al., “Collaborative search for information on the Internet by pupils: impacts of friendship”, in TAIT A., GASKELL A., MILLS R. (eds.), Open and Distance Learning: What Do We Know About Using New Technologies for Learning and Teaching? A Ten Year Perspective, Cambridge University Press, pp. 85–91, 2007. [DIN 08a] DINET J., “Collaborative information search in primary grades: impacts of domain knowledge”, SAMPSON K.J.G., SPECTOR J.M., ISAÍAS P., et al. (eds.), Proceedings of the IADIS International Conference on Cognition and Exploratory Learning in Digital Age, Freiburg, Germany, pp. 256–278, 13– 15 October 2008. [DIN 08b] DINET J., TRICOT A., “Recherche d’information dans les documents électroniques”, in CHEVALIER A., TRICOT A. (eds.), Ergonomie cognitive des documents électroniques, Presses Universitaires de France, Paris, pp. 35–69, 2008. [DIN 09a] DINET J., “Pour une conception centrée-utilisateurs des bibliothèques numériques”, Communication & Langages, vol. 161, pp. 59–74, 2009. [DIN 09b] DINET J., PASSERAULT J.-M., “Une approche centrée sur les usagers scolaires de la recherche documentaire informatisée”, in DURAMPART M. (ed.), Sociétés de la connaissance, fractures et evolutions, CNRS Editions, les Editions d'Hermès, Paris, pp. 83–97, 2009. [DIN 10] DINET J., BASTIEN C., KITAJIMA M., “What, where and how are young people looking for in a search engine results page? Impact of typographical cues and prior domain knowledge”, in BERTRAND D., NOIRHOMME-FRAITURE M., TRICOT A. (eds.), Proceedings of IHM 2010, ACM DL, ACM SIGCHI Series, pp. 105–112, 2010.

136

Information Retrieval in Digital Environments

[DIN 11a] DINET J., KITAJIMA M., “Draw me the Web”, Impact of mental model of the Web on information search performance of young users”, in RIVEILL N. (ed.), Proceedings of IHM ’11 23rd French Speaking Conference on Human-Computer Interaction, ACM Press, New York, pp. 35–41, 2011. [DIN 11b] DINET J., VIVIAN R., “The impact of friendship in synchronous collaborative retrieval in primary school”, British Journal of Educational Technology, vol. 43, no. 3, pp. 439–447, 2011. [DIN 11c] DINET J., VIVIAN R., SIMONNOT B., “La recherche collaborative d'information sur Internet: impact de l’affinité entre les jeunes collaborateurs”, Journal d’Interaction Personne - Système, vol. 2, no. 1, pp. 1–19, 2011. [DIN 12a] DINET J., CHEVALIER A., “Information searching in digital environment”, European Review of Applied Psychology, Special Issue, pp. 49–62, 2012. [DIN 12b] DINET J., SHIBATA T., “Des robots et des Hommes”, in CARRÉ M. (ed.), Innover pour + d'autonomie, Médialis, Paris, pp. 111–134, 2012. [DIN 13] DINET J., VIVIAN R., LA MANTIA K., et al., “Understanding the cultural dimension on the Web homepage preferences and visual exploration”, Proceedings of ICDS 2013, The 7th International Conference on Digital Society, Nice, France, pp. 81–86, 24 February–1 March 2013. [DON 07b] DONKER A., REITSMA P., “Drag-and-drop errors in young children’s use of the mouse”, Interacting with Computers, vol. 19, pp. 257–266, 2007. [DON 07a] DONKER A., REITSMA P., “Young children’s ability to use a computer mouse”, Computers & Education, vol. 48, no. 4, pp. 602–617, 2007. [DOR 11] DORUM K., GARLAND K., “Efficient electronic navigation: a metaphorical question?”, Interacting with Computers, vol. 23, pp. 129–136, 2011. [DRE 04] DRESANG E.T., GROSS M., HOLT L.E., “Children’s inlibrary use of computers in an urban public library”, Library and Information Science Research, vol. 26, no. 3, pp. 311–337, 2004.

Bibliography

137

[DRU 64] DRUCKER P., Managing for Results, Collins, London, 1964. [DRU 11] DRUSCH G., BASTIEN C., DINET J., “L’exploration visuelle de pages web”, in DINET J., BASTIEN C. (eds.), L’ergonomie des objets et des environnements physiques et numériques, HermesLavoisier, Paris, pp. 219–244, 2011. [DUC 83] DUCHNICKY R.L., KOLERS P.A., “Readability of text scrolled on visual display terminals as a function of window size”, Human Factors, vol. 25, pp. 683–692, 1983. [DUF 10a] DUFRESNE A., COURTEMANCHE F., PROM TEP S., “Analyse des interactions en utilisant le suivi oculaire, le suivi physiologique et les structures d’actions”, 22ème conférence francophone sur l’Interaction Homme-Machine (22nd French Conference on Computer-Human Interaction), Luxembourg, 20– 23 September 2010. [DUF 10b] DUFRESNE A., PROM TEP S., SÉNÉCAL S., et al., “Physiological measures, eye-tracking and task analysis to track user reactions in UGC”, 7th International Conference on Methods and Techniques in Behavioral Research: Measuring Behavior, Eindhoven, Netherlands, 24–27 August 2010. [DUM 01] DUMAIS S., CUTRELL E., CHEN H., “Optimizing search by showing results in context”, Proceedings of the ACM Conference on Human-Computer Interaction (CHI ’01), Seattle, WA, ACM Press, New York, pp. 277–284, 31 March–5 April, 2001. [EAS 86] EASTMAN S.T., AGOSTINO D.E., “Commanding the computer: functions and concepts of videotex technology for height-grade students”, Journal of Research and Development in Education, vol. 19, no. 2, pp. 49–57, 1986. [EDM 90] EDMONDS D.S, MOORE P., BALCOM K.M., “The effectiveness of an online catalog”, School Library Journal, vol. 10, pp. 28–32, 1990. [EDW 89] EDWARDS D.M., HARDMAN L., “Cognitive mapping and navigation in a hypertext environment”, in MCALEESE R. (ed.), Lost in Hyperspace, Ablex Publishing, NJ, pp. 90–105, 1989. [EIS 90] EISENBERG M.B., BERKOWITZ R.E., Information Problemsolving: The Big Six Skills Approach to Library and Information Skills Instruction, Ablex, Norwood, NJ, 1990.

138

Information Retrieval in Digital Environments

[EIS 02] EISENBERG M.B., JOHNSON D., “Learning and teaching information technology”, ERIC Clearinghouse on Information & Technology, Syracuse, NY, No. ED465377, 2002. Available at http://searcheric.org/digests/ed465377.html, 2002. [ELI 89] ELLIS D., “A behavioral approach to information retrieval system design”, Journal of Documentation, vol. 45, no. 3, pp. 171–212, 1989. [ELI 12] ELISON J.T., SASSON N.J., TURNER-BROWN L.M., et al., “Age trends in visual exploration of social and nonsocial information in children with autism”, Research in Autism Spectrum Disorders, vol. 6, no. 2, pp. 842–851, 2012. [ERT 85] ERTHAL M.J., “The status of keyboarding”, The Journal of Business Education, vol. 60, pp. 192–193, 1985. [ETH 08] ÉTHIER J., HADAYA P., TALBOT J., et al., “Interface design and emotions felt on B2C Web sites: empirical testing of a research model”, Computers in Human Behavior, vol. 24, no. 6, pp. 2771–2791, 2008. [EVE 97] EVERS V., DAY D., “The role of culture in interface acceptance”, In HOWARD S., HAMMOND J., LINDEGAARD G. (eds.), Human Computer Interaction, INTERACT ’97, Chapman and Hall, London, pp. 260–267, 1997. [FED 03] FEDDES R., VERMETTEN Y., BRAND-GRUWEL S., et al., “Strategische kennis over het oplossen van informatieproblemen: een exploratief onderzoek” [Strategic knowledge about information problem solving: an explorative study], Pedagogische Studiën, vol. 80, pp. 210–225, 2003. [FEI 93] FEIN R.M., OLSON G.M., OLSON J.S., “A mental model can help with learning to operate a complex device”, INTERACT ’93 and CHI ’93 Conference Companion on Human Factors in Computing Systems, pp. 157–158, 1993. [FID 99] FIDEL R., DAVIES R.K., DOUGLASS M.H., et al., “A visit to the information mall: Web searching behavior of high school students”, Journal of the American Society of Information Science, vol. 50, no. 1, pp. 24–37, 1999.

Bibliography

139

[FID 01] FIDEL R., BRUCE H.W., PEJTERSEN A.M., et al., “Collaborative Information Retrieval (CIR)”, The New Review of Information Behaviour Research: Studies of Information Seeking in Context, vol. 1, pp. 235–247, 2001. [FID 04] FIDEL R., PEJTERSEN A.M., CLEAL B., et al., “A multidimensional approach to the study of human-information interaction: a case study of collaborative information retrieval”, Journal of the American Society for Information Science and Technology, vol. 55, no. 11, pp. 939–953, 2004. [GAB 99] GABRIELLI S., MIZZARO S., “Negotiating a multidimensional framework for relevance space”, Proceedings of Mira ’99: Evaluating Interactive Information Retrieval, Glasgow, Ecosse, April 1999. [GAV 91] GAVER W., “Technology affordances”, Proceedings of the CHI ’1991, ACM Press, New York, pp. 79–84, 1991. [GIB 77] GIBSON J.J., “The theory of affordances”, in SHAW R.E., BRANSFORD J. (eds.), Perceiving, Acting, and Knowing, Lawrence Erlbaum Associates, Hillsdale, NJ, pp. 67–82, 1997. [GOL 99] GOLDMAN S.R., ZECH L.K., BISWAS G., et al., “Computer technology and complex problem solving: issues in the study of complex cognitive activity”, Instructional Science, vol. 27, pp. 235–268, 1999. [GRA 90] GRAY S.H., “Using protocol analyses and drawings to study mental models construction during hypertext navigation”, International Journal of Human–Computer Interaction, vol. 2, pp. 359–377, 1990. [GRE 06] GREENFIELD P., YAN Z., “Children, adolescents, and the Internet: a new field of enquiry in developmental psychology”, Developmental Psychology, vol. 42, no. 3, pp. 391−394, 2006. [GRO 95] GROSS M., “The imposed query”, Research Quaterly, vol. 35, no. 2, p. 236, 1995. [GRO 99] GROSS M., “Imposed versus self-generated questions: implications for reference practice”, Reference & User Services Quarterly, vol. 39, no. 1, p. 53, 1999.

140

Information Retrieval in Digital Environments

[GRO 00] GROSS M., “The imposed query and information services for children”, Journal of Youth Services in Libraries, vol. 13, no. 2, p. 10, 2000. [GRO 01] GROSS M., SAXTON M.L., “Who wants to know? Imposed queries in the public library”, Public Libraries, vol. 40, no. 3, p. 170, 2001. [GRO 02] GROSS M., SAXTON M.L., “Integrating the imposed query into the evaluation of reference service: a dichotomous analysis of user ratings”, Library & Information Science Research, vol. 24, no. 3, p. 251, 2002. [GUT 87] GUTHRIE J.T., MOSENTHAL P., “Literacy as multidimensional: locating information and reading comprehension”, Educational Psychologist, vol. 22, pp. 279–297, 1987. [GUT 88] GUTHRIE J.T., “Locating information in documents: examination of a cognitive model”, Reading Research Quarterly, vol. 23, pp. 178–199, 1988. [GUT 90] GUTHRIE J.T., DREHER M.J., “Literary as search: exploration via computer”, in NIX D., SPIRO R. (eds.), Cognition, Education, and Multimedia, Erlbaum, Hillsdale, NJ, pp. 65– 113, 1990. [GUT 91] GUTHRIE J.T., BRITEN T., BARKER K.G., “Roles of document structure, cognitive strategy and awareness in searching for information”, Reading Research Quarterly, vol. 26, pp. 300–321, 1991. [HAN 05] HANSEN P., JARVELIN K., “Collaborative information retrieval in an information-intensive domain”, Information Processing & Management, vol. 41, no. 5, pp. 1101–1119, 2005. [HAY 08] HAYATI Z., JOWKAR T., “Adoption of electronic reference materials in academic libraries of Iran”, The International Information & Library Review, vol. 40, pp. 52–63, 2008. [HIR 96] HIRSH S.G., “Complexity of search tasks and children’s information retrieval”, Proceedings of the 59th American Society for Information Science Annual Meeting, vol. 33, pp. 47–51, 1996.

Bibliography

141

[HIR 99] HIRSH S., “Children’s relevance criteria and information seeking on electronic resources”, Journal of the American Society for Information Science, vol. 50, no. 14, pp. 1265–1283, 1999. [HME 03] HMELO-SILVER C.E., “Analyzing collaborative knowledge construction: multiple methods for integrated understanding”, Computers & Education, vol. 41, no. 4, pp. 397–420, 2003. [HOL 91] HOLLNAGEL E., “Cognitive ergonomics and the reliability of cognition”, Le Travail Humain, vol. 54, no. 4, pp. 305–321, 1991. [HOO 89] HOOTEN P.A., “Online catalogs: will they improve children’s access?”, Journal of Youth Services in Libraries, vol. 2, pp. 267–272, 1989. [HUT 05] HUTCHINSON H., DRUIN A., BEDERSON B.B., et al., “How do I find blue books about dogs? The errors and frustrations of young digital library users”, Proceedings of HCII 2005, Las Vegas, NV, pp. 22–27 July 2005. [HUT 06] HUTCHINSON H., BEDERSON B.B., DRUIN A., “The evolution of the international children’s digital library searching and browsing interface”, Proceedings of IDC ’06, Tampere, Finland, pp. 105–112, 7–9 June 2006. [HYU 05] HYUN E., “A study of 5- to 6-year-old children’ peer dynamics and dialectical learning in a computer-based technology-rich classroom environment”, Computers & Education, vol. 44, pp. 69–91, 2005. [HYL 06] HYLDEGARD J., “Collaborative information behaviorexploring Kuhlthau's information search process model in a group-based educational setting”, Information Processing and Management, vol. 42, no. 1, pp. 276–298, 2006. [HYL 09a] HYLDEGARD J., “Beyond the search process: exploring group members’ information behaviour in context”, Information Processing & Management, vol. 45, no. 1, pp. 142–158, 2009. [HYL 09b] HYLDEGARD J., “Uncertainty dimensions of information behaviour in a group based problem solving context”, Nordic Journal of Information Literacy in Higher Education, vol. 1, no. 1, pp. 4–24, 2009.

142

Information Retrieval in Digital Environments

[ING 92] INGWERSEN P., Information Retrieval Interaction, Taylor Graham, London, 1992. [ING 96] INGWERSEN P., “Cognitive perspectives of information retrieval interaction: elements of a cognitive IR theory”, Journal of Documentation, vol. 52, no. 1, pp. 3–50, 1996. [ING 05] INGWERSEN P., JÄRVELIN K., The Turn: Integration of Information Seeking and retrieval in Context, Springer Verlag, Berlin, 2005. [INS 08] INSKIP C., BUTTERWORTH R., MACFARLANE A., “A study of the information needs of the users of a folk music library and the implications for the design of a digital library system”, Information Processing and Management, vol. 44, pp. 647–662, 2008. [ISO 11] ISOZAKI A., Japan-ness in Architecture, MIT Press, London, 2011. [JAN 00] JANSEN B.J., POOCH U., “A review of web searching studies and a framework for future research”, Journal of the American Society of Information Science and Technology, vol. 52, pp. 235–246, 2000. [JOC 08] JOCHMANN-MANNAK H., HUIBERS T., SANDERS T., “Children’s information retrieval: beyond examining search strategies and interfaces”, Paper presented at The 2nd BCS IRSG Symposium: Future Directions in Information Access 2008 (FDIA 2008), 2008. [JOH 83] JOHNSON-LAIRD P.N., Mental Models. Towards a Cognitive Science of Language, Inference, and Consciousness, Cambridge University Press, Cambridge, 1983. [JOI 98] JOINER R., MESSER D., LIGHT P., et al., “It is best to point for young children: a comparison of children’s pointing and dragging”, Computers in Human Behavior, vol. 14, pp. 513–529, 1998. [JON 99] JONES M.L.W., GAY G.K., RIEGER R.H., “Comparing evaluations of digital collection efforts”, DLib Magazine, vol. 5, no. 11, 1999. Available at http://www.dlib.org/dlib/november99/ 11jones.html.

Bibliography

143

[JUL 05] JULIEN H., MCKECHNIE L.E., HART S., “Affective issues in library and information science systems work: a content analysis”, Library & Information Science Research, vol. 27, no. 4, pp. 453–466, 2005. [KAI 91] KAISH S., GILAD B., “Characteristics of opportunities search of entrepreneurs versus executives: sources, interests, general alertness”, Journal of Business Venturing, vol. 6, no. 1, pp. 45–61, 1991. [KAL 10] KALTENBRUNNER M., “TUIO 2.0 Protocol Specification” (Draft), 18 January 2010. Available at http://www.tuio.org/?tuio20. [KAM 06] KAMPPURI M., BEDNARIK R., TUKIAINEN M., “The expanding focus of HCI: case culture”, Proceedings of NordiCHI 2006, Changing Roles, Oslo, Norway, pp. 405–408, 14–18 October 2006. [KAN 06] KANI-ZABIHI E., GHINEA G., CHEN S.Y., “Digital libraries: what do users want?”, Online Information Review, vol. 30, no. 4, pp. 395–412, 2006. [KAN 07] KANI-ZABIHI E., GHINEA G., CHEN S.Y., “Involving users in OPAC interface design: perspective from a UK study”, Lecture Notes in Computer Science, Springer Berlin/Heidelberg, vol. 4558, pp. 374–383, 2007. [KAN 08] KANI-ZABIHI E., GHINEA G., CHEN S.Y., “User perceptions of online public library catalogues”, International Journal of Information Management, vol. 28, no. 6, pp. 492–502, 2008. [KAN 10] KANI-ZABIHI E., COLES-KEMP L., “Service users’ requirements for tools to support effective on-line privacy and consent practices”, NordSec 2010, Epsoo, Finland, 2010. [KAR 98] KARAMUFTUOGLU M., “Collaborative information retrieval: toward a social informatics view of IR interaction”, Journal of the American Society for Information Science, vol. 49, no. 12, pp. 1070–1080, 1998.

144

Information Retrieval in Digital Environments

[KAR 11] KARANAM S., A cognitive model of Web-navigation based on semantic information from pictures, Thesis submitted in partial fulfillment of the requirements for the award of Degree of Doctor of Philosophy in Computer Science, Cognitive Science Lab, International Institute of Information Technology, Hyderabad, April 2011. [KEL 07] KELLY D., “Methods for evaluating interactive information retrieval systems with users”, Foundations and Trends in Information Retrieval, vol. 3, nos. 1–2, pp. 12–24, 2007. [KIM 07] KIM H., FESENMAIER D.R., “The persuasive architecture of destination websites: the effect on first impressions”, Information and Communication Technologies in Tourism, vol. 6, pp. 25–266, 2007. [KIM 08] KIM H., FESENMAIER D.R., “Persuasive design of destination Web sites: an analysis of first impression”, Journal of Travel Research, vol. 47, no. 1, pp. 3–13, 2008. [KIN 88] KINSTCH W., “The role of knowledge in discourse comprehension: a construction-integration model”, Psychological Review, vol. 95, no. 2, pp. 163–182, 1988. [KIN 98] KINTSCH W., Comprehension: A Paradigm for Cognition, University Press, Cambridge, 1998. [KIT 95] KITAJIMA M., POLSON P.G., “A comprehension-based model of correct performance and errors in skilled display-based human–computer interaction”, International Journal of Human–Computer Interaction, vol. 43, no. 1, pp. 65–99, 1995. [KIT 97] KITAJIMA M., POLSON P.G., “A comprehension-based model of exploration”, Human–Computer Interaction, vol. 12, no. 4, pp. 345–89, 1997. [KIT 00] KITAJIMA M., BLACKMON M.H., POLSON P.G., “A comprehension-based model of Web navigation and its application to Web usability analysis”, in MCDONALD S., WAERN Y., COCKTON G. (eds.), People and Computers XIV - Usability or Else! (Proceedings of HCI 2000), Springer, pp. 357–373, 2000.

Bibliography

145

[KIT 03] KITAJIMA M., “Comprehension-based approach to HCI for designing interaction in information space”, Proceedings of HCI International, Human-Centred Computing: Cognitive, Social and Ergonomic Aspects, vol. 3, pp. 1031–1035, 2003. [KIT 04] KITZINGER J., “Le sable dans l’huître: analyser des discussions de focus group”, Bulletin de psychologie, vol. 57, no. 3, pp. 299–307, 2004. [KIT 05] KITAJIMA M., BLACKMON M.H., POLSON P.G., “Cognitive architecture for website design and usability evaluation: comprehension and information scent in performing by exploration”, Oral communication at HCI International, 2005. [KIT 08] KITAJIMA M., “Vers un modèle de recherche d'information basé sur la compréhension”, in DINET J. (ed.), Usages, usagers et compétences informationnelles au 21e siècle, Hermes-Lavoisier, Paris, pp. 285–310, 2008. [KNI 08] KNIGHT S.A., SPINK A., “Toward a Web search information behaviour model”, Information Science and Knowledge Management, vol. 14, pp. 209–234, 2008. [KRA 02a] KRACKER J., “Research anxiety and students’ perceptions of research: an experiment. Part I: effect of teaching kuhlthau’s ISP model”, Journal of the American Society for Information Science and Technology, vol. 53, no. 4, pp. 282–294, 2002. [KRA 02b] KRACKER J., WANG P., “Research anxiety and students’ perceptions of research: an experiment. Part II: content analysis of their writings on two experiences”, Journal of the American Society for Information Science and Technology, vol. 53, no. 4, pp. 295–307, 2002. [KUB 99] KUBECK J.E., “Finding information on the world wide web: exploring older adults exploration”, Educational Gerontology, vol. 25, no. 2, p. 167, 1999. [KUH 85] KUHLTHAU C., Teaching the Library Research Process, Centre for Applied Research in Education, New York, NY, 1985. [KUH 88] KUHLTHAU C.C., “Perceptions of the information search process in libraries: a study of changes from high school through college”, Information Processing & Management, vol. 24, no. 4, pp. 419–427, 1988.

146

Information Retrieval in Digital Environments

[KUH 91] KUHLTHAU C.C., “Information search process: a summary of research and implications for school library media programsé”, in WOOLS B. (ed.), The Research of School Library Media Centers Papers of the Treasure Mountain Research Retreat, Park City, Utah, pp. 11–120, 17–18 October 1989, Hi Willow Research and Publishing, 1991. [KUH 99a] KUHKTHAU C.C., “The role of experience in the information search process of an early career information worker: perceptions of uncertainty, complexity, construction and sources”, Journal of the American Society for Information Science, vol. 50, no. 5, pp. 399–412, 1999. [KUH 99b] KUHKTHAU C.C., “Accommodating the user’s information search process: challenges for information retrieval system designers”, Bulletin of the American Society for Information Science 50th Anniversary Special Issue on Information Seeking and Finding, vol. 25, no. 3, pp. 12–16, 1999. [KUH 01] KUHLTHAU C.C., TAMA S., “Information search process of lawyers: a call for ‘just for me’ information services”, Journal of Documentation, vol. 57, no. 1, pp. 25–43, 2001. [KUH 04] KUHLTHAU C.C., Seeking Meaning: A Process Approach to Library and Information Services, 2nd ed., Libraries Unlimited, Westport, CT, 2004. [KUH 07] KUHLTHAU C.C., MANIOTES L., CASPARI A., Guided Inquiry: Learning in the 21st Century, Libraries Unlimited, Westport, CT, 2007. D., Modelle zur Repräsentation [KUR 04] KUROPKA natürlichsprachlicher Dokumente, Logos Berlin, Verlag, 2004. [KWO 11] KWON S., CHOI E., CHUNG M.K., “Effect of control-todisplay gain and movement direction of information spaces on the usability of navigation on small touch-screen interfaces using tap-n-drag”, International Journal of Industrial Ergonomics, vol. 41, no. 3, pp. 322–330, 2011. [LAB 05] LABERGE J.C., SCIALFA C.T., “Predictors of web navigation performance in a life span sample of adults”, Human Factors, vol. 47, no. 2, pp. 289–302, 2005.

Bibliography

147

[LAN 84] LANGER J.A., “Examining background knowledge and text comprehension”, Reading Research Quaterly, vol. 19, no. 4, pp. 468–481, 1984. [LAN 10] LANE A.E., ZIVIANI J.M., “Factors influencing skilled use of the computer mouse by school-aged children”, Computers & Education, vol. 55, no. 3, pp. 1112–1122, 2010. [LAU 02] LAURSEN B., JENSEN B.R., GARDE A.H., et al., “Effect of mental and physical demands on muscular activity during the use of a computer mouse and a keyboard”, Scandinavian Journal of Work Environmental Health, vol. 28, no. 4, pp. 215– 221, 2002. [LAZ 00] LAZONDER A.., BIEMAS J.A., WOPEREIS G.J.H., “Differences between novice and experienced users in searching information on the world wide web”, Journal of the American Society for Information Science, vol. 51, no. 6, pp. 576–581, 2000. [LAZ 05] LAZONDER A., “Do two heads search better than one? Effects of student collaboration on web search behaviour and search outcomes”, British Journal of Educational Technology, vol. 36, no. 3, pp. 465–475, 2005. [LEB 06] LEBEDEV M.A., NICOLELIS A.L., “Brain–machine interfaces: past, present and future”, Trends in Neurosciences, vol. 29, no. 9, pp. 536–547, 2006. [LEB 07] LE BIGOT L., ROUET J.-F., “The impact of presentation format, task assignment, and prior knowledge on students’ comprehension of multiple online documents”, Journal of Literacy Research, vol. 39, pp. 445–470, 2007. [LEB 10] LE BIGOT L., TERRIER P., JAMET E., et al., “Does textual feedback hinder spoken interaction in natural language?”, Ergonomics, vol. 53, pp. 43–55, 2010. [LEG 01] LEGGETT M., FINLAY M., “Science, story, and image: a new approach to crossing the communication barrier posed by scientific jargon”, Public Understanding of Science, vol. 10, no. 2, pp. 157–171, 2001. [LEH 10] LE HONG S., BIESTERFELDT J., “Les gestes en fonction des cultures”, available at http://www.karizmatic.fr/les-gestes-enfonction-des-cultures/, 2010.

148

Information Retrieval in Digital Environments

[LEV 90] LEVITT T.S., LAWTON D.T., “Qualitative navigation for mobile robots”, Artificial Intelligence, vol. 44, pp. 305–360, 1990. [LIN 00] LINDERHOLM T., GADDY M., VAN DEN BROEK P., et al., “Effects of causal text revisions on more and less-skilled readers’ comprehension of easy and difficult texts”, Cognition and Instruction, vol. 18, pp. 525–556, 2000. [LIN 02a] LINDERHOLM T., VAN DEN BROEK P., “The effects of reading purpose and working memory capacity on the processing of expository text”, Journal of Education Psychology, vol. 94, no. 4, pp. 778–784, 2002. [LIN 02b] LING J., VAN SCHAIK P., “Effect of text and background colour on visual search of Web pages”, Displays, vol. 23, pp. 223–230, 2002. [LIN 04] LINDERHOLM T., VIRTUE S., TZENG Y., et al., “Flunctuations in the availability of information during reading: capturing cognitive processes using the landscape model”, Discourse Processes, vol. 37, pp. 165–186, 2004. [LIN 06] LING J., VAN SCHAIK P., The influence of font type and line length on visual search and information retrieval in web pages, International Journal of Man-Machine Studies, vol. 64, no. 5, pp. 395–404, 2006. [LIN 11] LIN P., ABNEY K., BECKEY G., “Robot ethics: mapping the issues for a mechanized world”, Artificial Intelligence, vol. 175, nos. 5–6, pp. 942–949, 2011. [LIP 99] LIPPONEN L., “Challenges for computer-supported collaborative learning in elementary and secondary level: Finnish perspective”, Proceedings of CSCL ’99, The 3rd International Conference on Computer Support for Collaborative Learning, Erlbaum, Mahwah, NJ, pp. 368–375, 1999. [LIU 10] LIU C., WHITE R.W., DUMAIS S., “Understanding Web Browsing Behaviors through Weibull Analysis of Dwell Time”, Oral communication in SIGIR’10, Geneva, Switzerland, July 19–23, 2010,. [LOC 02] LOCKE E.A., LATHAM G., “Building a practically useful theory of goal setting and task motivation”, American Psychologist, vol. 57, no. 9, pp. 705–717, 2002.

Bibliography

149

[LOP 10] LOPATOVSKA I., ARAPAKIS I., “Theories, methods and current research on emotions in library and information science, information retrieval and human–computer interaction”, Information Processing & Management, vol. 47, no. 4, pp. 575– 592, 2010. [MAC 91] MACKENZIE I.S., SELLEN A., BUXTON W., “A comparison of input devices in elemental pointing and dragging tasks”, in ROBERTSON S.P., OLSON G.M., OLSON J.S. (eds.), Proceedings of CHI ’91 Human Factors in Computing Systems: Reaching through Technology, ACM, New York, NY, pp. 161–166, 1991. [MAC 94] MACKENZIE J., “Grazing the Net : raising a generation of free range students – part one”, From Now On: The Educational Technology Journal, available at www.fno.org/grazing1.html, 1994. [MAC 03] MACKENZIE S., “Motor behavior models for humancomputer interaction”, in CARROLL J.M. (ed.), HCI Models, Theories, and Frameworks: Toward a Multidisciplinary Science, Morgan Kaufmann, New York, pp. 27–54, 2003. [MAR 87] MARCHIONINI G., TEAGUE J., “Elementary students’ use of electronic information services: an exploratory study”, Journal of Research on Computing in Education, vol. 20, pp. 139–155, 1987. [MAR 88] MARCHIONINI G., SCHNEIDERMAN B., “Finding facts versus browsing knowledge in hypertext systems”, IEEE Computers, vol. 21, pp. 70–88, 1988. [MAR 91a] MARKUS H.R., KITAYAMA S., “Culture and the self: implications for cognition, emotion, and motivation”, Psychological Review, vol. 20, pp. 568–579, 1991. [MAR 91b] MARCHIONINI G., MEADOW C., DWIGGINS S., et al., “A study of user interaction with information retrieval interfaces: progress report”, Canadian Journal of Information Science, vol. 16, pp. 42–59, 1991. [MAR 92] MARCHIONINI G., “Interfaces for end-user information seeking”, Journal of the American Society for Information Science, vol. 43, pp. 156–163, 1992.

150

Information Retrieval in Digital Environments

[MAR 93] MARCHIONINI G., DWIGGINS S., KATZ A., et al., “Information seeking in full-text end-user-oriented search systems: the role of domain and search expertise”, Library and Information Science Research, vol. 15, pp. 35–69, 1993. [MAR 95] MARCHIONINI G., Information Seeking in Electronic Environments”, Cambridge University Press, Cambridge, 1995. [MAR 04a] MARCHIONINI G., Funding opportunities for research in HIB, Report on the 2004 Annual Research Symposium of ASIS&T’s Special Interest Group on Information Needs, Seeking and Use (SIG USE), available at http://www.asis.org/ SIG/SIGUSE/ASIST04_SIG-USE_Symposium_Report.doc, 2004. [MAR 04b] MARCHIONINI G., LIN X., DWIGGINS S., “Effects of search and subject expertise on information seeking in a hypertext environment”, In HENDERSON D. (ed.), Proceedings of the American Society for Information Science, Annual Meeting, Learned Information, Toronto, pp. 129–142, 1990. [MAR 06] MARCHIONINI G., “Toward human-computer information retrieval”, Bulletin of the American Society for Information Science and Technology, vol. 32, no. 5, pp. 20–22, 2006. [MAR 09] MARANGUNIC N., GRANIC A., “The influence of cognitive and personality characteristics on user navigation: an empirical study”, in STEPHANIDIS C. (ed.), Universal Access in HCI, Part III, HCII 2009, LNCS 5616, Springer-Verlag, Berlin/Heidelberg, pp. 216–225, 2009. [MAS 01] MASUDA T., NISBETT R.E., “Attending holistically vs. analytically: comparing the context sensitivity of Japanese and Americans”, Journal of Personality and Social Psychology, vol. 81, pp. 922–934, 2001. [MAS 08] MASUDA T., ELLSWORTH P.C., MESQUITA B., et al., “Placing the face in context: cultural differences in the perception of facial emotion”, Journal of Personality and Social Psychology, vol. 94, pp. 365–381, 2008. [MCD 98] MCDONALD S., STEVENSON R.J., “Effects of text structure and prior knowledge of the learner on navigation in hypertext”, Human Factors, vol. 40, no. 1, pp. 18–27, 1998.

Bibliography

151

[MCD 05] MCDERMOTT M., “Knowledge workers: you can gauge their effectiveness”, Leadership Excellence, vol. 22, no. 10, pp. 15–17, 2005. [MCL 67] MCLUHAN M., The Gutenberg Galaxy: The Making of Typographic Man, University of Toronto Press, Toronto, 1967. [MEY 03] MEYER J.-A., FILLIAT D., “Map-based navigation in mobile robots. A review of map-learning and path-planning strategies”, Cognitive Systems Research, vol. 4, no. 4, pp. 283–317, 2003. [MEY 04] MEYER T., RODON C., “Trouver sur Internet une réponse à une question”, Hermès – CNRS, vol. 39, pp. 27–34, 2004. [MIC 05] MICHEL P., “Digitizing special collections: to boldly go where we have been before”, Library Hi Tech, vol. 23, no. 3, pp. 379–395, 2005. [MIT 80] MITTELSTAEDT M.-L., MITTELSTAEDT H., “Homing by path integration in a mammal”, Naturwissenschaften, vol. 67, pp. 566–567, 1980. [MIZ 98] MIZZARO S., “Relevance: the whole history”, Journal of the American Society for Information Science, vol. 48, no. 9, pp. 810–832, 1998. [MOO 91] MOORE G.C., BENBASAT I., “Development of an instrument to measure the perceptions of adopting an information technology innovation”, Information Systems Research, vol. 2, pp. 192–222, 1991. [MOY 04] MOYO L.M., “Electronic libraries and the emergence of new service paradigms”, The Electronic Library, vol. 22, no. 3, pp. 220–230, 2004. [NAH 04a] NAHL D., “Affective and cognitive information behavior: interaction effects in internet use”, Proceedings of the 68th ASIST Annual Meeting, Information Today, Medford, NJ, vol. 42, pp. 104–116, 2004. [NAH 04b] NAHL D., “Measuring the affective information environment of web searchers”, Proceedings of the 68th of the American Society for Information Science and Technology, Information Today, Medford, NJ, vol. 41, pp. 191–197, 2004.

152

Information Retrieval in Digital Environments

[NAH 07] NAHL D., “Social–biological information technology: an integrated conceptual framework”, Journal of the American Society for Information Science, vol. 58, no. 13, pp. 2021–2046, 2007. [NIC 07] NICHOLAS D., HUNTINGTON P., JAMALI H.R., “The use, users, and role of abstract in the digital scholarly environment”, The Journal of Academic Librarianship, vol. 33, no. 4, pp. 446– 453, 2007. [NIE 90] NIELSEN J., (ed.), Designing User Interfaces for International Use, Elsevier, Amsterdam, 1990. [NIE 99] NIELSEN J., Designing Web Usability: The Practice of Simplicity, New Riders Publishing, Indianapolis, IN, 1999. [NIE 03] NIELSEN J., Homepage real estate allocation. On-line resource, available at http://www.useit.com/alertbox/ 20030210.html, 2003. [NIE 10] NIELSEN J., PERNICE K., Eyetracking Web Usability, New Riders, London, 2010. [NIE 11] NIE J.Y., Le domaine de recherche d’information – Un survol d’une longue histoire, available at http:// www.iro.umontreal.ca/~nie/IFT6255/historique-RI.html, 2011. [NIE 12] NIELSEN J., “How long do users stay on Web pages?”, available at http://www.useit.com/alertbox/page-abandonmenttime.html, 2012. [NIS 01] NISBETT R.E., PENG K., INCHEOL C., et al., “Culture and systems of thought: holistic versus analytic cognition”, Psychological Review, vol. 108, pp. 291–310, 2001. [NIS 03a] NISBETT R.E., The Geography of Thought: How Asians and Westerners Think Differently and Why, The Free Press, 2003. [NIS 03b] NISBETT R.E., MASUDA T., “Culture and point of view”, Proceedings of National Academic Science USA, vol. 100, pp. 11163–11175, 2003. [NIS 05] NISBETT R.E., MIYAMOTO Y., “The influence of culture: holistic versus analytic perception”, TRENDS in Cognitive Sciences, vol. 9, no. 10, pp. 467–473, 2005.

Bibliography

153

[NOR 83] NORMAN D.A., “Some observations on mental models”, in GENTNER D.R., STEVENS A.L. (eds.), Mental Models, Erlbaum, Hillsdale, NJ, pp. 7–14, 1983. [NOT 71] NOTON D., STARK L., “Scanpaths in saccadic eye movements while viewing and recognizing patterns”, Vision Research, vol. 11, pp. 929–942, 1971. [NUR 99] NURMELA E., LEHTINEN E., PALONEN T., “Evaluating CSCL log files by social network analysis”, Proceedings of CSCL ’99: The 3rd International Conference on Computer Support for Collaborative Learning, Erlbaum, Mahwah, NJ, pp. 434–444, 2004. [OMA 04] O’MALLEY C., FRASER D.S., Literature review in learning with tangible technologies, available at http://hal.archivesouvertes.fr/hal-00190328/en/, 2004. [ONW 04] ONWUEGBUZIE A.J., JIAO Q.G., “Information search performance and research achievement: an empirical test of the anxiety-expectation médiation model of library anxiety”, Journal of the American Society for Information Science and Technology, vol. 55, no. 1, pp. 41–54, 2004. [OVE 01] OVER P., “The TREC interactive track: an annotated bibliography”, Information Processing & Management, vol. 37, pp. 369–381, 2001. [PAK 08] PAK R., PRICE M.M., “Designing an information search interface for younger and older adults”, Human Factors, vol. 50, no. 4, pp. 614–628, 2008. [PAP 05] PAPASTERGIOU M., “Students’ mental models of the Internet and their didactical exploitation in informatics education”, Education and Information Technology, vol. 10, no. 4, pp. 341–360, 2005. [PEA 03] PEARSON R., VAN SCHAIK P., “The effect of spatial layout and link colour of web pages on performance in a visual search task and an interactive search task”, International Journal of Human-Computer Studies, vol. 59, pp. 327–353, 2003. [PEI 98] PEILING W., SORGEL D., “A cognitive model of document use during a research project. Study 1. Document selection”, Journal of the American Society for Information Science, vol. 49, no. 2, pp. 115–125, 1998.

154

Information Retrieval in Digital Environments

[PEJ 98] PEJTERSEN A.M., FIDEL R., A framework for work centered evaluation and design: a case study of information retrieval on the Web, Working Paper for MIRA workshop, 1998. [PET 01] PETTIGREW K.E., FIDEL R., BRUCE H., “Conceptual frameworks in information behavior”, Annual Review of Information Science and Technology, vol. 35, pp. 43–78, 2001. [PIC 02] PICARD R.W., KLEIN J., “Computers that recognise and respond to user emotion: theoretical and practical implication”, Interacting with Computers, vol. 14, no. 2, pp. 141–169, 2002. [PIC 03] PICARD R.W., “Affective computing: challenges”, International Journal of Human–Computer Studies, vol. 59, nos. 1–2, pp. 55–64, 2003. [PIP 11] PIPER A.M., HOLLAN J.D., “Supporting medical communication for older patients with a shared touch-screen computer”, International Journal of Medical Informatics, vol. 82, no. 11, pp. 242–250, 2013. [QIA 10] QIAN K., MA X., DAI X., et al., “Robotic etiquette: socially acceptable navigation of service robots with human motion pattern learning and prediction”, Journal of Bionic Engineering, vol. 7, no. 2, pp. 150–160, 2010. [QUA 11] QUAN W., NIWA H., ISHIKAWA N., et al., “Assisted-care robot based on sociological interaction analysis”, Computers in Human Behavior, vol. 27, no. 5, pp. 1527–1534, 2011. [RAY 07] RAYNER K., LI X., WILLIAMS C.C., et al., “Eye movements during information processing tasks: individual differences and cultural effects”, Vision Research, vol. 47, pp. 2714–2726, 2007. [RED 08] REDDY M., JANSEN B.J., “A model for understanding collaborative information behavior in context: a study of two healthcare teams”, Information Processing & Management, vol. 44, no. 1, pp. 256–273, 2008. [REE 82] REED E., JONES R., Reasons for Realism: Selected Essays of James J. Gibson, Erlbaum, Hillsdale, NJ, 1982.

Bibliography

155

[REY 03] REYNOLDS W.M., MILLER G.E., “Current perspectives in educational psychology”, in REYNOLDS W.M., MILLER G.E. (eds.), Handbook of Psychology, John Wiley & Sons, Hoboken, NJ, pp. 3–20, 2003. [RIB 08] RIBY D.M., HANCOCK P.J.B., “Viewing it differently: social scene perception in Williams syndrome and Autism”, Neuropsychologia, vol. 46, pp. 2855–2860, 2008. [RIB 09] RIBY D.M., HANCOCK P.J.B., “Looking at movies and cartoons: eye-tracking evidence from Williams syndrome and autism”, Journal of Intellectual Disability Research, vol. 53, pp. 169–181, 2009. [RIM 08] RIMMER R.J., WARWICK C., BLANDFORD A., et al., “An examination of the physical and the digital qualities of humanities research”, Information Processing and Management, vol. 44, pp. 1374–1392, 2008. [RIS 12] RISKO E.F., ANDERSON N.C., LANTHIER S., et al., “Curious eyes: individual differences in personality predict eye movement behavior in scene-viewing”, Cognition, vol. 122, no. 1, pp. 86–90, 2012. [ROG 95] ROGERS E., Diffusion of Innovation, 4th ed., Free Press, New York, 1995. [ROO 93] ROOK F.W., DONNELL M.L., “Human cognition and the expert system interface: mental models and inference explanations”, IEEE Transactions on Systems, Man, and Cybernetics, vol. 23, pp. 1649–1661, 1993. [ROT 91] ROTHI L., OCHIPA C., HEILMAN K., “A cognitive neuropsychological model of limb praxis”, Cognitive Neuropsychology, vol. 8, no. 6, pp. 443–458, 1991. [ROT 97] ROTHI L.J.G., OCHIPA C., HEILMAN K.M., “A cognitive neuropsychological model of limb praxis”, in ROTHI L.J.G., HEILMAN K.M. (eds.), Apraxia. The Neuropsychology of Action, Taylor & Francis, Erlbaum, UK, pp. 29–49, 1997. [ROU 03] ROUET J.-F., Cent fenêtres sur Internet. Représentations et usages des nouvelles technologies dans le grand public, Presses du CNRS, Paris, 2003.

156

Information Retrieval in Digital Environments

[ROU 11] ROUET J.-F., ROS C., GOUMI A., et al., “The influence of surface and deep cues on grade school students’ assessment of relevance in Web menus”, Learning and Instruction, vol. 21, no. 2, pp. 205–219, 2011. [ROW 98] ROWE D.W., SIBERT J., IRWIN D., “Heart rate variability: indicator of user state as an aid to human-computer interaction”, Proceedings of the CHI ’98 Human Factors in Computing Systems Conference, Los Angeles, CA, 1998. [SAL 04] SALAZAR O., GROSSEN M., “Représentations sociales et analyse de discours produit dans des focusgroups: un point de vue dialogique”, Bulletin de Psychologie, vol. 57, no. 3, pp. 263– 272, 2004. [SAN 05] SANDFELD J., JENSEN B.R., “Effect of computer mouse gain and visual demand on mouse clicking performance and muscle activation in a young and elderly group of experienced computer users”, Applied Ergonomics, vol. 36, no. 5, pp. 547– 555, 2005. [SAR 96] SARACEVIC T., “Modeling interaction in information retrieval (IR): a review and proposal”, Journal of the American Society for Information Science, vol. 33, pp. 3–9, 1996. [SAR 97] SARACEVIC T., “The stratified model of information retrieval interaction: extension and applications”, Journal of the American Society for Information Science, vol. 34, pp. 313–327, 1997. [SCA 97] SCAPIN D.L., BASTIEN J.M.C., “Ergonomic criteria for evaluating the ergonomic quality of interactive systems”, Behaviour and Information Technology, vol. 6, nos. 4–5, pp. 220–231, 1997. [SCH 91] SCHAMBER L., “User’s criteria for evaluation in a multimedia environment”, Proceedings of the American Society for Information Science, Washington, DC, Learned Information, Medford, NJ, pp. 126–133, 1991. [SCH 94] SCHAMBER L., “Relevance and information behavior”, Annual Review of Information Science and Technology, vol. 29, pp. 3–48, 1994.

Bibliography

157

[SCH 98] SCHACTER J., CHUNG G., DORR A., “Children’s internet searching on complex problems: performance and process analysis”, Journal of the American Society for Information Science, vol. 49, pp. 840–849, 1998. [SCO 93] SCOTT D., “Visual search in modern human-computer interfaces”, Behaviour and Information Technology, vol. 12, pp. 174–189, 1993. [SHA 02] SHANNON D., “Kuhlthau’s Information Search Processé”, School Library Media Activities Monthly, vol. 19, no. 1, pp. 19– 23, 2002. [SHA 08] SHARIT J., HERNANDEZ M.A., CZAJA S.J., et al., “Investigating the roles of knowledge and cognitive abilities in older adult information seeking on the Web”, ACM Transactions on Computer-Human Interaction, vol. 15, no. 1, pp. 3–28, 2008. [SHE 06] SHEN S., WOOLLEY M., PRIOR S., “Towards culturecentred designé”, Interacting with Computers, vol. 18, no. 4, pp. 820–852, 2006. [SIM 08] SIMONNOT B., “La pertinence en sciences de l'information: des modèles, une théorie ?”, in PAPY F. (ed.), Problématiques émergentes dans les Sciences de l'Information, Hermes/Lavoisier, Paris, pp. 161–182, 2008. [SLE 09] SLEGERS K., VAN BOXTELL M., JOLES J., “Effects of computer training and internet usage on cognitive abilities in older adults: a randomized controlled study”, Aging Clinical and Experimental Research, vol. 21, no. 1, pp. 43–54, 2009. [SLO 02] SLONE D.J., “The influence of mental models and goals on search patterns during Web interaction”, Journal of the American Society for Information Science and Technology, vol. 53, no. 13, pp. 1152–1169, 2002. [SMI 96] SMITH P.A., “Towards a practical measure of hypertext usability”, Interacting with Computers, vol. 4, pp. 365–381, 1996. [SMI 99] SMITH M.W., SHARIT J., CZAJA S.J., “Aging, motor control, and the performance of computer mouse tasks”, Human Factors, vol. 41, no. 3, pp. 389–396, 1999.

158

Information Retrieval in Digital Environments

[SNO 09] SNOWMAN J., MCCOWN R., BIEHLER R., Psychology Applied to Teaching, 12th ed., Houghton Mifflin Company, Boston, MA, 2009. [SOL 93] SOLOMON P., “Children's information retrieval behavior: a case analysis of an OPAC”, Journal of the American Society for Information Science, vol. 44, pp. 245–264, 1993. [SOL 99] SOLOMON P., “Information mosaics: patterns of action that structure”, in WILSON T.D., ALLEN D.K. (eds.), Exploring the Contexts of Information Behaviour: Proceedings of the 2nd International Conference on Research in Information Needs, Seeking and Use in Different Contexts, Taylor Graham, London, pp. 150–175, 1999. [SOL 03] SOLOMON P., “Children's information retrieval behavior: a case analysis of an OPAC”, Journal of the American Society for Information Science, vol. 44, pp. 245–264, 2003. [STR 93] STROMMEN E.F., “Is it easier to hop or walk. Development issues in interface design”, Human–Computer Interaction, vol. 8, pp. 337–352, 1993. [STR 06] STRONGE A.J., ROGERS W.A., FISK A.D., “Web-based information search and retrieval: effects of strategy use and age on search success”, Human Factors, vol. 48, no. 3, pp. 434–446, 2006. [SUH 92] SUH K.S., JENKINS M., “A comparison of linear keyword language and restricted natural language data base interfaces for novices users”, Information Systems Research, vol. 3, pp. 252–272, 1992. [SUT 98] SUTCLIFFE A.G., ENNIS M., “Towards a cognitive theory of information retrieval”, Interacting with Computers, vol. 10, no. 3, pp. 321–351, 1998. [SUT 00] SUTCLIFFE A.G., ENNIS M., WATKINSON S.J., “Empirical studies of end-user information searching”, Journal of the American Society for Information Science, vol. 51, no. 3, pp. 1211–1231, 2000. [TAR 99] TARDIF J., Pour un enseignement stratégique: l’apport de la psychologie cognitive, Logiques, Montréal, 1999.

Bibliography

159

[TEN 08] TENOPIR C., WANG P., ZHANG Y., et al., “Academic users’ interactions with ScienceDirect in search tasks: affective and cognitive behaviors”, Information Processing and Management, vol. 44, pp. 105–121, 2008. [THA 98] THATCHER A., GREYLING M., “Mental models of the Internet”, International Journal of Industrial Ergonomics, vol. 22, pp. 299–305, 1998. [TIS 99] TISSEYRE R.-C., Knowledge management, théorie et pratique de la gestion des connaissances, Hermès, Paris, 1999. [TOR 82] TORNATZKY L.G., KLEIN K.J., “Innovation characteristics and adoption-implementation: a meta-analysis of findings”, IEEE Transactions on Engineering Management (EM), vol. 29, no. 1, pp. 28–45, 1982. [TU 08] TU Y.-W., SHIH M., TSAI C.-C., “Eighth graders’ web searching strategies and outcomes: the role of task types, web experiences and epistemological beliefs”, Computers & Education, vol. 51, no. 3, pp. 1142–1153, 2008. [TWI 97] TWIDALE M.B., NICHOLS D.M., PAICE C.D., “Browsing is a collaborative process”, Information Processing & Management, vol. 33, no. 6, pp. 761–783, 1997. [TZA 04] TZANETAKIS G., COOK P., “Music analysis and retrieval systems”, Journal of American Society for Information Science and Technology, vol. 55, no. 12, pp. 1077–1083, 2004. [TZE 05] TZENG Y., VAN DEN BROEK P., KENDEOU P., et al., “The computational implementation of the Landscape Model: modeling inferential processes and memory representations of text comprehension”, Behavioral Research Methods, Instruments & Computers, vol. 37, pp. 277–286, 2005. [VAN 93] VAN DEN BROEK P., FLETCHER C.R., RISDEN K., “Investigations of inferential processes in reading: a theoretical and methodological integration”, Discourse Processes, vol. 16, pp. 169–180, 1993. [VAN 95] VAN DEN BROEK P., RISDEN K., HUSEBYE-HARTMANN E., “The role of reader’s standards of coherence in the generation of inferences during reading”, in LORCH E.P., O’BRIEN E.J. (eds.), Sources of Coherence in Reading, LEA, Hillsdale, NJ, pp. 353– 374, 1995.

160

Information Retrieval in Digital Environments

[VAN 96] VAN DEN BROEK P., RISDEN K., FLETCHER C.R., et al., “A ‘landscape’ view of reading: fluctuating patterns of activation and the construction of a stable memory representation, in BRITTON B.K., GRAESSER A.C. (eds.), Models of Understanding Text, Erlbaum, Hillsdale, NJ, pp. 165–187, 1996. [VAN 99] VAN DEN BROEK P., YOUNG M., TZENG Y., et al., “The landscape model of reading: inferences and on-line construction of a memory representation”, in VAN OOSTENDORP H., GOLDMAN S.R., (eds.), The Construction of Mental Representations During Reading, LEA, Mahwah, NJ, pp. 71–98, 1999 [VAN 01] VAN DEN BROEK P., TZENG Y., VIRTUE S., et al., “Inference making and memory for text: a computational model”, Paper presented at The 42nd Annual Meeting of the Psychonomic Society, Orlando, Florida, FL, 2001. [VAN 02] VAN DEN BROEK P., VIRTUE S., GADDY M., et al., “Comprehension and memory of science texts: inferential processes and the construction of a mental representation”, in OTERO J., LEON J.A., GRAESSER A.C. (eds.), The Psychology of Science Text Comprehension, LEA, Mahwah, NJ, pp. 131–154, 2002. [VAN 05] VAN DEN BROEK P., RAPP D.N., KENDEOU P., “Integrating memory-based and constructionist approaches in accounts of reading comprehension”, Discourse Processes, vol. 39, pp. 299– 316, 2005. [VAN 08] VAN DEN BROEK P., KENDEOU P., “Cognitive processes in comprehension of science texts: the role of co-activation in confronting misconceptions”, Applied Cognitive Psychology, vol. 22, pp. 335–351, 2008. [VAN 09] VAN VELSEN L., MELENHORST M., “Incorporating user motivations to design for video tagging”, Interacting with Computers, vol. 21, no. 3, pp. 221–232, 2009. [VAN 10] VAN DER WARDT V., BANDELOW S., HOGERVORST E., “The relationship between cognitive abilities, well-being and use of new technologies in older people”, The Proceedings of ECCE 2010 Conference, Delft, The Netherlands, pp. 333–334, 2010.

Bibliography

161

[VAS 02] VASS E., “Friendship and collaborative creative writing in the primary classroom”, Journal of Computer Assisted Learning, vol. 18, no. 1, pp. 102–110, 2002. [VIV 07] VIVIAN R., DINET J., “La recherche collaborative d'information: vers un système centré utilisateur”, Document numérique, vol. 10, no. 3/4, pp. 25–4, 2007. [VIV 09] VIVIAN R., DINET J., “RCI WEB: Présentation des premiers résultats de l’utilisation d’un système collaboratif de recherche d’information centré utilisateur”, Revue sur les Interactions Homme – Machine, vol. 9, no. 2, pp. 85–110, 2009. [VIV 12] VIVIAN, R., DINET, J., BERTOLO, D., “Spring: a Solution for Managing the Third DOF with Tactile Interface”, APCHI ‘12: Asia Pacific Conference on Computer Human Interaction, Matsue, Japan, pp. 625–635, August 28–31, 2012. [WAL 94] WALTER V.A., “The information needs of children”, Advances in Librarianship, vol. 18, pp. 111–129, 1994. [WAL 97] WALLACE R., KUPPERMANN J., “On-line search in the science classroom: Benefits and possibilities”, AERA, Chicago, available at http://mydl.soe.umich.edu/papers/online_search.pdf, 1997. [WAL 00] WALLACE R., KUPPERMAN J., KRAJCIK J., et al., “Science on the Web: students on-line in a sixth grade classroom”, Journal of the Learning Sciences, vol. 9, no. 1, pp. 75–104, 2000. [WAL 08] WALRAVEN A., BRAND-GRUWEL S., BOSHUIZEN H.P.A., “Information-problem solving: a review of problems students encounter and instructional solutions”, Computers in Human Behavior, vol. 24, no. 3, pp. 623–648, 2008. [WAL 10] WALRAVEN A., BRAND-GRUWEL S., BOSHUIZEN H.P.A., “Fostering transfer of web searchers’ evaluation skills: a field test of two transfer theories”, Computers in Human Behavior, vol. 26, pp. 716–728, 2010. [WAL 09b] WALRAVEN A., BRAND-GRUWEL S., BOSHUIZEN H.P.A., “How students evaluate information and sources when searching the World Wide Web for information”, Computers & Education, vol. 52, no. 1, pp. 234–246, 2009.

162

Information Retrieval in Digital Environments

[WAN 06] WANIA C.E., ATWOOD M.E., MCCAIN K.W., “How do design and evaluation interrelate in HCI research?”, in Proceedings of Designing Interactive Systems, New York: ACM, pp. 90–98, 2006,. [WAR 01] WARD R., MARSDEN P., CAHILL B., et al., “Using skin conductivity to detect emotionally significant events in humancomputer interaction”, in 13th Annual Conference of the Association Francophone d’Interaction Homme-Machine (AFIHM) and 15th Annual Conference of the Human-Computer Interaction Group of the British Computer Society (IHM-HCI 2001), Lille, 10–14 September 2001. [WAR 03] WARD R., MARSDEN P., “Physiological responses to different web-page designs”, International Journal of HumanComputer Studies, vol. 59, no. 1/2, pp. 199–212, 2003. [WAR 06] WARWICK C., TERRAS M., HUNTINGTON P., et al., “If you build it will they come? The Lairah study: quantifying the use of online resources in the arts and humanities through statistical analysis of user log data”, The Proceedings of Digital Humanities, pp. 225–228, 2006. [WIL 99] WILSON T.D., “Models in information behaviour research”, The Journal of Documentation, vol. 55, no. 3, pp. 249–270, 1999. [WIL 00] WILSON T.D., “Human information behavior”, Special Issue on Information Science Rsearch, vol. 3, no. 2, pp. 49–55, 2000. [WIL 03] WILSON L.A., “If we build it, will they come? Library users in a digital world”, in LEE S.H. (ed.), Improved Access to Information: Portals, Content Selection, and Digital Information, Haworth Information Press, Binghamton, New York, pp. 19–28, 2003. [WOB 09] WOBBROCK J.O., MORRIS M.R., WILSON A.D., “Userdefined gestures for surface computing”, Proceedings of CHI 2009, Boston, MA, pp. 1083–1092, 4–9 April 2009. [YAN 05] YAN Z., “Age differences in children's understanding of complexity of the Internet”, Journal of Applied Developmental Psychology, vol. 26, pp. 385−396, 2005.

Bibliography

163

[YAN 06] YAN Z., “What influences childrens and adolescents understanding of the complexity of the Internet?”, Developmental Psychology, vol. 42, pp. 418−428, 2006. [YAN 09] YAN Z., “Limited knowledge and limited resources: children’s and adolescents’ understanding of the Internet”, Journal of Applied Developmental Psychology, vol. 30, pp. 103– 115, 2009. [ZHA 06] ZHAI C.X., LAFFERTY J., “A risk minimization framework for information retrieval”, Information Processing & Management, vol. 42, no. 1, pp. 31–55, 2006. [ZHA 08] ZHANG Y., “The influence of mental models on undergraduate students’ searching behavior on the Web”, Information Processing and Management, vol. 44, pp. 1330– 1345, 2008.

Index

A, C

I, M, N

activation of concepts, 44, 46 activity, 1–3, 7, 9–14, 17, 19–24, 26, 31–33, 35, 40– 42, 45, 48, 49, 52, 55, 60– 62, 70, 79, 88, 90, 92, 94– 98, 100, 118, 120, 121 affect, 10, 40, 66, 69, 89, 111 cognitive space, 27 computational model, 43 consumers, 67, 102

inferences, 38, 40, 46 information behavior, 4, 93 retrieval, 1–4, 6–15, 17, 19–36, 38–42, 46–52, 68, 69, 73, 79, 80, 83, 84, 88–98, 100, 101, 103, 107, 109–111, 116–118, 121 intuitive, 85, 87, 97, 98 iterative model, 21 marketing, 10 navigation, 7, 8, 9, 15, 39, 66, 103, 104, 105, 107 needs of users, 63, 73

D, E, H declarative knowledge, 40, 43, 46 design, 6, 8, 9, 12–14, 17, 23, 26, 51, 61, 64, 73, 87, 97, 102, 105, 107 economy, 10 human factors, 7

P, R, S, polyrepresentation, 28 procedural knowledge, 38, 40, 41 prospective ergonomics, 70 robotics, 9, 14, 15, 17 search engine, 2, 39–42, 47, 97–99, 101, 112–117, 120

166

Information Retrieval in Digital Environments

searching behavior, 4 seeking behavior, 4 social need, 73 T, U, V, W task analysis, 7, 55, 61 user experience, 32

visual exploration, 7, 108, 109, 110, 111, 114, 115, 116, 117, 121 working memory, 40, 42, 44, 46

Other titles from

in Information Systems, Web and Pervasive Computing

2013 BERNIK Igor Cybercrime and Cyberwarfare CAPET Philippe, DELAVALLADE Thomas Information Evaluation LEBRATY Jean-Fabrice, LOBRE-LEBRATY Katia Crowdsourcing: One Step Beyond

2012 GAUSSIER Eric, YVON François Textual Information Access STOCKINGER Peter Audiovisual Archives: Digital Text and Discourse Analysis VENTRE Daniel Cyber Conflict

2011 LEMBERGER Pirmin, MOREL Mederic managing Complexity of Information Systems

STOCKINGER Peter Introduction to Audiovisual Archives STOCKINGER Peter Digital Audiovisual Archives VENTRE Daniel Cyberwar and Information Warfare

2010 BONNET Pierre Enterprise Data Governance

2009 BONNET Pierre, DETAVERNIER Jean-Michel, VAUQUIER Dominique Sustainable IT Architecture: the Progressive Way of Overhauling Information Systems with SOA PAPY Fabrice Information Science RIVARD François, ABOU HARB Georges, MERET Philippe The Transverse Information System VENTRE Daniel Information Warfare

2008 MANOUVRIER Bernard, LAURENT Ménard Application Integration : EAI, B2B, BPM and SOA PAPY Fabrice Digital Libraries

2006 CORNIOU Jean-Pierre Looking Back and Going Forward in IT