Chapter 6

Multimedia Information Seeking Through Competitive Intelligence Process¹

6.1. Introduction

Value-added information is widely considered a strategic asset for any decision maker or company able to find, use, and reuse it. Advances in information and communication technology have led to improvement in the quality and increases in the quantity of mutimedia information.

Given this increase in the volume of multimedia information, and the complexity attached to its structure, information seekers may need assistance in searching and efficaciously accessing multimedia information. Rapidly obtaining relevant information may be difficult, but nevertheless this remains indispensable in responding to user information needs and, consequently, in resolving decision problems.

The connotative nature of multimedia information does not in any way facilitate its search and retrieval. The task of multimedia information retrieval is rendered delicate given the fact that multimedia information supports a wide variety of viewpoints, readings, and interpretations. Multimedia information representation and indexing is complex, and this complexity also extends to the expression of a user’s need for multimedia information. We have observed [MAG 08] that a semantic gap may emerge between the user, their information needs, and the information solution proposed by the information system. This semantic gap is, in our opinion, caused by problems associated with the notion of user needs and by the specificities of multimedia information.

Competitive intelligence (CI) can be used to provide approaches and tools for solving these problems. In this chapter, we will present our methodology, based on CI, for facilitating research and access to multimedia information.

6.2. The two dimensions of CI: decisions and information

For [LEV 01], “Competitive intelligence can be regarded as a mode of thinking and acting in the new economy”; this concerns not only the material economy but also the “immaterial” economy: information and knowledge. The report produced by the French General Planning Commission, led by Henri Martre, in 1994 [MAR 94] defines CI as “the set of coordinated research, processing and distribution actions involving useful information and economic actors susceptible to use this information. These actions are carried out legally with all the necessary protection for the safeguard of the company’s patrimony and with the best quality, delay and cost”. This definition, therefore, describes not only those operations in which CI may be of assistance but also the way in which these operations should be carried out.

We will consider CI under two main aspects: first, the decision aspect and second, the information that is at the root of, and which justifies, the decision [MAG 07].

CI is a form of “decision intelligence”. It may be defined not only as “the action of judgment in relation to a point of contention” but also as “the judgment that produces a solution”. The decision process is therefore complex, but produces solutions.

A decision problem may be seen as a gap established between a given situation and a desired situation, considered to be stable. To reduce this gap, a decision problem solving process must be put in place. This process elicits different possibilities for problem resolution, measuring the points for and against each possible solution. The adoption of a solution proposed by the decision-making process allows us to resolve a decision problem with little or no risk. Consequently, a “good” decision is unenvisageable without intelligent information use.

CI is also, by and large, a form of “informational intelligence” in that it professes mastery of an insightful approach to research and information processing and promotes intelligent usage of value-added information.

Many authors [DAV 03, DOU 95, MAR 94, REV 98, SAL 00, CAR 03, LES 06] consider CI as an information process that exists with the aim of assisting the decision process.

In our research, we have adopted a fairly explicit approach put forward by David [DAV 06] who considers CI as an eight-step process:

1) identification and definition of a decision problem;

2) transformation of the decision problem into an information problem;

3) identification of relevant sources of information;

4) collection of relevant information;

5) analysis of collected information to extract indicators for decision;

6) interpretation of indicators;

7) decision;

8) capitalization and protection of assets.

Capitalization and protection of patrimony are transversal steps that should be applied to all other stages of the process.

The CI process presented above emphasizes the link that exists between a decision problem and an information problem (Figure 6.1). It represents a set of decision and information issues that are interconnected by logical relationships and exchanges between actors (decision maker, watcher, and infomediary) aiming to resolve these problems.

The explicit expression of a decision problem allows us to deduce the information need(s) of users/actors (decision makers). Based on this statement, we will look at the representation of multimedia information not only from the information process view but also as an element of a decision problem.

6.3. Multimedia information: between complexity and accessibility

As explained above, multimedia information is a rich type of information that may provide elements of solution to any information problem and thus constitute a useful basis for decision making. However, the richness of multimedia information — the combination of a set of elements rich in meaning — may complicate or restrain user access to this information.

For instance, the polysemic nature of an image and its ambiguity can pose problems for information seekers. The image may have many meanings and can be interpreted in various ways. Yet, “a sign may express not only several meaning but a quantity of meanings as the image only takes on its quality of signification in relation to the context and implications it assumes”. The interpretation of visual signs — that is, an image — is therefore a delicate matter as it varies from one individual to another.

Information seeking, indexing, and representation are also difficult activities. Questions may be raised concerning which elements to take into account in the representation of this kind of information.

6.4. The information seeking process: an overview of paradigmatic evolution

We will tackle the question of information seeking as a process. The expression “information seeking process” refers to all stages and methodology that provide access to information.

Little after the invention of computers, models and tools were proposed with the aim of facilitating the progress of information seeking processes. In 1985, following the first international Conference on Scientific Information, dedicated to “information seeking”, and Peter Luhn’s demonstration of his indexing system (KWIC). The first conference on scientific information held in 1985 and dedicated to “Information Seeking” (which with Peter Lunhn’s demonstration of his indexing system, from this time information seeking) are standing tradition. Since the first conference, the chain, or documentary process, which consists of collection, storage, processing, and diffusion of documents has passed from a simple design to a more systemic approach.

The notion of “System Relevance” emerged at during this period. Indeed, the first models and studies of the information seeking process focused on systems as technical equipment based on the general theory of systems¹ developed by Von Bertalanffy (1937) and on the systemics of these models.

The process of information retrieval in computer systems is based on three fundamental steps: the representation of the request, the representation of information, and the establishment of a relationship between the two. Over time and with technological developments, work has focused on perfecting these steps.

Since then, many studies concerning information seeking systems and processes have been published. The first studies on the subject of the information seeking process looked at the technical aspects of information retrieval systems. In the 1990s, a second approach emerged, involving the user in the information seeking process. This new focus, known as “the user-oriented cognitive paradigm”, led to modifications in the perception of performance of an information seeking process and/or an information seeking system by taking into consideration the user, with his/her specific cognitive characteristics and information needs.

6.5. Actors involved in information seeking processes and problem solving

To solve an information problem or decision-making problem, a person is given the responsibility and the charge of carrying out the necessary research activities. In a CI process, this may be the watcher [BOU 04, KIS 07], the “infomediary” [KNA 07], the archivist, the librarian, a consultant, or the person requesting the information [GOR 06].

The watcher is responsible for information seeking and for monitoring of information sources. The “infomediary” plays a role of coordination and mediation between actors involved in the CI process. The watcher, “infomediary”, information officer, or consultant are all actors involved in the information seeking process with the aim of providing necessary relevant information for solving a decision problem.

For Goria [GOR 06], the person who is responsible for solving a decision problem or making the decision must create “a sequence for the transformation of the problem, enabling the transformation of problem data, in relation to the initial state envisaged, which helps define a path towards a solution to the problem”.

This definition is interesting and attractive because the actor seeking information must possess some knowledge, not only in relation to the decision problem but also in connection with the information problem. Although knowledge relating to the decision problem is mainly communicated by the decision maker — who identifies and formulates the decision problem — the person in charge of information seeking will transform the decision problem into an information problem. This transition from decision to information problem is carried out by combining information sources that may contain relevant information that can solve a decision problem.

The information problem may also be approached in the form of a communication problem, where CI actors are not the final users of the information. In this case, a request is made to a human intermediary — a search specialist. In the context of CI, this may be the watcher responsible for information retrieval. The clarification of the decision or information problem by the decision maker and the watcher is, therefore, essential before any use of a computerized information seeking system. The CI actors must have a common understanding of the terms used, their meaning, and their interpretation. This is absolutely critical to avoid difficulties linked to the expression of the problem. Work may be needed to reformulate the problem with the participation of both the decision maker and the watcher.

Before presenting our approach of multimedia information seeking through CI processes, we should clarify our choices in terms of terminology. Researchers working on information retrieval issues and access use generally varied vocabulary; the variety of terms used is probably due to reference points from different contexts. Thus, the terms used in information and communication sciences in connection with information seeking are not necessarily the same as those used by a computer scientist. It is therefore important to clearly explain our choices and our intended meanings before going any further.

6.5.1. Terminology: the notion of the user

A user is a person who uses a product or a service. The elements used, in this context, subsist after use. A person becomes a user after a single use of a service, product, or fragment of information, but their relationship with these items may change with repeated and/or frequent use.

Depending on the context of research and the type of study, the “user” may be defined using different situations. In the context of our work, information seeking is seen as the use of an information system with the aim of accessing multimedia information, and also the use of the retrieved information. The user may, therefore, be defined as an information seeker interacting with the information system, an interaction that may itself be presented as a communication situation.

6.5.2. Terminology: the notion of use

In terms of terminology, “use” and the less frequent variant “usage” are presented as more or less synonymous.

In the Petit Robert dictionary [PRO 05] (in French, but the definitions and distinctions made remain relevant in English), two main meanings are given to the term “usage”. The first refers to “the social practices which age (or frequency) renders normal in a given culture”. The second refers to “the use of an object, natural or symbolic, to particular ends”.

According to [LEC 97], “use is an action and a means of putting something into service in order to attain a specific result”.

The term “usage” is more widely encountered in the fields of information and communication sciences and sociology. From a sociological standpoint, Breton and Pro [BRE 06] and Ihadadene and Chaudiron [IHA 08] describe the insertion of new technologies into society. The observation of the way in which individuals interact with technological innovations was at the core of research on users and gratification in the American functionalist approaches of the 1960s and 1970s. “What people do with media” was the main slant of this research, based on the assumption that people actively use media to obtain specific elements of satisfaction that correspond to their psychological or psychosociological needs.

The notion of “usage” is, therefore, found in the context of study of interactions between humans and new technologies.

By focusing on the notion of “usage”, the individual ceases to be considered first and foremost as an epistemic subject, that is, as a learner faced with knowledge constituted as mental representations. Attention is instead focused on the socio-technical knowledge the user puts into action to access this knowledge. This somewhat changes the perspective of the user. Although no such distinction exists in English, in French the words “utilisateur” and “usager” are used to differentiate between the former and latter cases, as stated in [IHA 08]. There is also a shift in perspective in that work on the user not only no longer concentrates exclusively on the cognitive dimension of person/system interactions but also on social and symbolic dimensions. “Usage” is thus found in a socio-technical framework as “a stabilized use of an object, a tool to obtain an effect”.

Cognitive approaches have been proposed for the study of usages of objects and techniques and their appropriation by individuals. However, if we focus our research on the sphere of mediation studies, we must place these approaches at social rather than technical mediation level. We must remember that, before belonging to collective units, users are individuals in their own right.

In this chapter, we will place an emphasis on the interaction of the individual (user) with the information seeking system, aiming to go beyond the technical aspects of interaction, by treating the information retrieval system as a “partner” in the cognitive activity of individuals, giving it the status of a cognitive artifact.²

When considering the use of multimedia information, we subscribe to the point of view set out in [IHA 08], where “use” refers to the interaction between the human individual and the way in which an individual “uses” equipment based on their own competences, cognitive style, and habits. This does not mean that the user is free from the influence of their environment (private, social, symbolic, etc.) but simply that emphasis is placed on the interaction aspect.

Three main principles stand out in the definitions presented above: precision, repetition of the use action, and the socio-psychological and technical contextualization of the action of use. In this work, we will consider the user to be any person who interacts with an object; this object may be an “information system”, where the information is the object sought by a person for a particular use.

6.6. Applying a user-centered approach to facilitate multimedia information seeking

The representation of information is a key step in ensuring the success of an information seeking process. An examination of literature on the subject of standards in the representation of multimedia information and a study of the needs of users on the empirical approach³ have shown us that the user is barely present, if not absent, in considerations of representation of this information; there is a clear lack of balance between the representation of information on the one hand and the effective needs of users on the other.

In recent times, however, new technologies have reached a degree of maturity which, in our opinion, allows us to place the user and their information needs at the center of considerations. We, therefore, offer a new approach dedicated to the representation of multimedia information, centered on the user and on the use of the information. Our approach allows us to re-establish balance by integrating the information needs of users. An approach of this kind is “open” as it brings together a group of information elements concerning not only the information itself but also elements that characterize its production and use. The aim of this approach is to adapt the representation of multimedia information to the user’s information needs and expectations to personalize the seeking process for each user.

Our methodology for a multimedia information seeking process highlights two elements which, from our point of view, are essential in developing a multimedia information representation in that they best ensure access to this type of information. The first of these elements is the multimedia information itself and its characteristics; the second is the user and their information needs.

We suggest that user parameters and those parameters linked to their information needs be used to enrich the representation of standardized information (e.g. by Dublin Core initiation). The user’s information needs are determined by the user, by the multimedia information available, and by the context of use of the information.

Contextualization involves the interaction of three elements: multimedia information, the user, and the information need (shown in the form of attributes representing the context of use of the information sought — see Figure 6.2).

The “multimedia information” element represents the object of the information needs of the user. The representation of this element is standardized.

The “user” element corresponds to the subject interacting with the object (the information). The user also interacts with the information system, which facilitates access to multimedia information. As we intend to adapt the representation of information to the user and their information needs, a representation of the user must be established as a vital phase of our suggestions.

The “context of use” element refers to the aims of the subject (the user) concerning the object (the information). This element also reflects the information needs of the user. The context of use of the information represents the end results of an information seeking process from the user’s point of view. We consider that every usage is contextualized.

Contextualization, which brings together these three interactive elements, is a “package” that characterizes the information needs of the user and can provide context for different adaptation processes linked to the user’s information needs. Each of these three elements possesses a group of attributes for illustration and specification. These attributes are based on the CI approach as they consider the notion of the user’s information and decision problems. We will provide specific details of this later on.

6.6.1. Multimedia information granulation to support multimedia information seeking processes

Multimedia information presented following different approaches in published literature on the subject: we encounter low-level, high-level, and structural approaches [MAG 06]. The use of standard description forms set up by the community of information officers and computer scientists aims to provide tools for the description of the contents and structure of a document. It would seem, however, that these approaches do not give sufficient consideration to the information needs of the user or to the context of use of research results.

In specialist proposals for indexing and representing information, the usual approach consists of describing a given document using a standard bibliographic notice or description form. These forms, also known as “metadata” — that is, structured data about other data (http://dublincore.org/resources/faq/#whatismetadata) — give access to documents by representational attributes (date, authors, abstract, etc.).

Two standards have emerged which are useful for the representation of multimedia information: the Dublin Core and the International Association of Sound and Audiovisual Archives (IASA) standard.

The Dublin Core is an international standard for the definition of bibliographical elements. Hillman [HIL 01] defines the metadata standard of the Dublin Core as “a set of simple but effective elements for the description of a wide variety of network resources”. This standard is made up of 15 fields or elements: cover, description, type, relationship, source, subject, title, collaborator, creator, editor, rights, date, format, identifier, and language.

According to Weibel [WEI 99], the Dublin Core defines a list of fundamental properties capable of providing basic descriptive information for any and all types of resources, independently of format or type. It is essential that the model should remain independent from the platform on which the resources are found. Unfortunately, this standard does not include a “user” field or a field referring to the context of use.

The IASA standard, created by the International Association of Sound and Audiovisual Archives, was set up in 1969 in Amsterdam as a standard for international cooperation between archives holding sound and audiovisual recordings. As with the Dublin Core, the IASA standard contains the following specific descriptive elements:

1) title and responsibility relationships,

2) edition, issue,

3) publication, production, distribution, transmission and date of creation,

4) copyright,

5) physical description,

6) series,

7) notes,

8) note and limits of availability.

An analytical or multilevel description is then used for the first five elements. As with the Dublin Core, use and the user are absent from this standard.

The work of Weibel [WEI 99], Lelong [LEL 00], Hillman [HIL 01], Roussi [ROU 06], and Auffret [AUF 00], in continuing reflections on standards (such as those of the AFNOR or the ISO (International Standard Organization)), attempts to demonstrate and explain the positive and negative points of various standards. We are forced to recognize that approaches to the representation of multimedia information, whether the result of academic research or consensus-based institutional considerations (such as the AFNOR⁴ or the ISO (www.iso.ch)), the Dublin Core (http://dublincore.org/documents/usageguide/) or the IASA), barely consider the user and/or the user’s information needs. This is due to the contents of the standards themselves. Despite occasional claims to the contrary (in terms of operability, reusability, etc.), these standards are not used in relation to the user; we must conclude that their missions are the object of incomplete and often implicit definitions on the part of the user.

6.6.2. Integration of the representation of the user into the multimedia information retrieval process

The notion of representing the user is fairly recent. The interest of the notion lies in the information seeking process and in information retrieval systems devoted to the user, which then allow users to find required multimedia information.

The representation of a user is an explicit and simplified model of their characteristics. The use of a representation leads to a strategy for integration of the user in the information seeking process and allows exploitation of necessary knowledge concerning the user and their behaviors. User representation is a fairly complex method that requires us to “borrow” from different branches of science.

The representation of the user in an information retrieval system is generally undertaken with the aim of improving the operation of the system. A more specific aim is to personalize the responses of the information system and reduce the level of complexity, which may “shut in” a user in their interactions with an information seeking system during the information seeking process. In the context of our study, the user (the information seeker) aims to find a particular fragment of multimedia information. Therefore, the user has an information need that is at the root of their engagement in the research process.

Representation of the user allows us to personalize the information retrieval process and develop a flexible information system, one capable of behavioral modification and personalization of responses in relation to the information needs of specific users.

The architecture of an information retrieval system that adapts the representation of the user is based on the cognitive development of users. The abilities and knowledge of users in relation to specialist fields and to the types of problems they can solve are recorded to refine the information seeking process and functions of the information system. The collection of information concerning users allows us to apply an information filtering procedure, whether thematic or functional. After representation, thematic filtering is used to store information that may be used to solve decision problems in the system. Functional filtering is based on the profile of the user and the information proposed corresponds to the user’s preferences.

Representation of the user may be carried out using several criteria. User characteristics may be broken down into two categories: static characteristics (representing static parameters) and dynamic characteristics (which are changeable). In the context of our work, we will predetermine characteristics as described in the following sections.

6.6.2.1. Representation of user characteristics

This parameter, used to identify a user, contains elements that distinguish the user (a decision maker or watcher in the context of CI). It is presented as a list of attributes concerning the identity and training of the user: for example <surname, first name, postal address, email address, telephone number, status, training>. The <training> attribute contains the name of the course followed by the user and the highest level attained in this training.

The identity parameter may be shared by different users. We include references to contexts of use in this parameter, that is, references to decision and information problems posed by the user and processed by the information system, along with associated annotations.

6.6.2.2. Representation of user knowledge

We draw a distinction between theoretical knowledge and practical knowledge, or “know-how”. A user may have practical abilities without necessarily possessing theoretical knowledge, or, in the opposite case, have theoretical knowledge, but very little idea of practical applications.

A second distinction should be made between knowledge of the information seeking system and knowledge of the domain concerned by research. The user knowledge parameter allows us to determine what the user does or does not know concerning the operation of the information seeking system and concerning the domain of research.

6.6.2.2.1. Representation of user knowledge of the system used

Knowledge of the system has an impact on the information seeking system; it influences the formulation of the intention of the user, the expression of an information need, and the execution of tasks by the user. The integration of user knowledge within a representation of the user is a particularity often found in the domain of adaptive hypermedia.

Two types of representation are used. The first is a group model, based on the classification of a user among a group of users using a fuzzy matrix associating the user with a group. User preferences are deduced directly from the preferences of other members of the same group. The second type of representation is an individual model, based either on annotation of the user or on the detection and identification of the activities of the user during the information seeking process.

We will look at both types of representation. From our perspective, we consider that it is possible to deduce a user’s preferences from their requests (level of granularity of the information concerned by the research, vocabulary used in expressing the request, etc.), from definition of the context of use of information gathered by the user, and from the annotations made by the user.

In our representation of this parameter, knowledge of the system is represented by an attribute using value pairing. A first attribute concerns the precision of user knowledge of operators used in formulating information seeking equations (e.g. Boolean operators). A second attribute concerns the success of an information seeking task in an information seeking system or (Internet) search engine over a series of 10 cases. The third of these attributes concerns the success of an information seeking task. This presumes mastery of information seeking operators and of the conversion of an information requirement into a search request.

6.6.2.2.2. Representation of user knowledge in the research domain

The characteristics of this parameter concern the knowledge of patterns of reading multimedia information (including knowledge of the structure of audiovisual elements, technical characteristics of audio and video elements, fixed images, etc.). Following the example of the representation of knowledge of the system, knowledge of the domain of research is represented as an attribute-value pairing. A first attribute describes the name of the research domain, and a second gives detail as to the knowledge the user has of the domain. With this aim, we have chosen to study the number of information seeking tasks successfully carried out by a user in a particular domain over a total of 10 cases.

6.6.2.3. Representation of user competences

Competence may be considered an “internalization” of a set of knowledge and practical abilities or as a judgment of this internalization. There is always a form of “ability” and knowledge above and beyond competence. Judgment of competence is either a judgment of our own abilities or of those of others.

The user competence parameter refers both to competence in a domain and to competence in using a system. System competence reflects the ability of the user to formulate a request, clarify a need in the form of a context of user, and successfully carry out an information seeking task. We consider that a user has information seeking competences in a specific domain if they have both the necessary knowledge (practical and theoretical knowledge of the specificities of the domain of research) and the ability to mobilize this knowledge successfully to use the information seeking system.

Competences are built up through learning. In our case, the user acquires competences (system competences and competence in the domain of multimedia information) by carrying out information seeking tasks using information systems. Competence is, therefore, acquired by working on real information seeking tasks. To build competence, therefore, we must valorize the set of meaningful research situations.

6.6.2.4. Representation of user preferences

The user preference parameter allows us to determine the preferences and perceptions of the user concerning:

– The representation of multimedia information (the nature of attributes used by the user during the information seeking process, using attributes concerning denotation, connotation, symbolism, etc.).

– The way in which research results are presented by the system. A user might, for example, prefer the most recent documents. Other considerations include the language in which results are displayed, the number of results shown by the system, and appreciation of the way in which the user gathers information (in this case, we aim to know whether the user prefers the PULL or PUSH method).

– In gathering information, does the user take an intuitive, analytical, discursive, or sensory approach? Does the user treat the information in a reflective, methodical, or affective way? Finally, are the user’s preferred sources formal or informal in nature?

The representation of user characteristics is a key aspect of our proposals. The following section looks at the representation of user information needs concerning multimedia information. Clarification of the notion of information need is therefore necessary.

6.6.2.5. Representation of information needs

The notion of information needs is crucial in information seeking. An information need is, on the one hand, a transformation of the information and decision problem and, on the other, the cornerstone of information seeking systems, as the information need determines the relevance of items of information.

The information need is part of the cognitive process whereby the user becomes aware of a need and expresses this need not only in a particular language, but also when the user engages in an information seeking process and interacts with an information system. According to [LEC 08a], “the information need corresponds to a lack of knowledge by an individual in a situation, where this lack of knowledge prevents the individual from understanding or from acting in an optimal manner in the situation”.

The information need “is born of a cognitive impulse” [LEC 08] but also from a psychological impetus [TRI 04a]. Since the 1980s, cognitive psychologists have been interested in the notion of information need, relating it to a state of gaps in knowledge. An individual responsible for information seeking begins an information seeking process to fill this knowledge gap, thus creating the ability to solve an information problem. By defining the information need, it becomes easier to identify sources and find relevant information later on.

6.6.2.5.1. Awareness of the information need

The user may become aware of their information need by setting it out in two ways: either through interaction with an entity responsible for information seeking (which may be a human intermediary) or by interaction with an information seeking system, through the transformation of an information need (in natural language) into an information request (language used by the information system) that takes the form of a research equation. Tricot [TRI 04b] states that the dawning of awareness of an information need is not always as obvious as it might seem, but it is necessary for the definition of a research goal.

Each user encounters his/her own difficulties in expressing an information need. Each expresses a need using their own vocabulary and in connection with a set of parameters including culture, language, professional experience, and so on. A single word may have many meanings and connotations depending on context and on the person using the word.

Given the complex nature of the question of information need, interview techniques have been developed to evaluate the process by which a user becomes aware of a need.

6.6.2.5.2. Information needs and knowledge: paradoxal or complementary?

Researchers in the domain of psychology consider that awareness of an information need presumes knowledge of the object of research (connected subject and themes) and the ability to find and use information. In our opinion, the user should also possess knowledge concerning the formulation of information needs and of methods used to obtain information. Knowledge of the operation of information systems is also essential.

The fact of launching an information seeking process is positively influenced by the fact of having previously sought information. Thus, a study carried out by Joo and Grable [JOO 01] on information demands concerning retirement showed that those retired individuals who sought the most information on financing their retirement were the richest and the best informed.

Information that is suitable to respond to an information need is said to be relevant. However, a certain level of knowledge of the information language is required to make use of this information. Le Coadic [LEC 08b] highlights the paradox of information needs and the knowledge concerning these needs. According to the author, “an information need shows the state of knowledge in which an information seeker finds himself when confronted with a demand for information he does not have, a piece of information which is necessary for him to continue his research work. It is thus born of a cognitive impulse”.

The information need, an important factor in the information seeking process, may be perceived from the perspective of cognitive style. According to Legendre [LEG 93], “cognitive style is a personal, global and relatively stable approach which characterizes the distinct manner in which one person prefers to think, learn, organize their experiences and knowledge, perceive and process information, comprehend perceptual elements or solve a problem in a wide variety of situations”. For this reason, the question of information need takes on various dimensions according to the information seeking situation of the user.

Information needs may be seen as falling into the following three categories:

– Information need/acquisition of new knowledge: This first category is the broadest of the three. The user seeks information that he/she needs but does not possess. Tricot [TRI 03a] [TRI 03b] assumes that “we only seek if we know what we don’t know, and we know that we may find”.

– Information need/confirmation of preexisting knowledge: the user identifies a need, links it to what they already know, and determines what knowledge must be acquired to enrich their knowledge with additional and precise information.

– Information need/exploration of knowledge: in this case, the user takes an exploratory approach. The information need is absolute and does not relate to the first two aspects.

We support the idea expressed by Le Coadic [LEC 08a] that an information need is “extensible”, as one information requirement can produce another. Thus, a need to acquire new knowledge may lead to a need to confirm this knowledge.

In our opinion, an information need presupposes two things:

– A hope, a perspective in that each user has their own interpretation of the information they require.

– A use or usage for this information. The user has certain intentions that direct their information need. The gathered information will be used for a precise aim in a particular context. For these reasons, we can say that the information need is a connection between a user and an item of information with the intention to use. To find this information, each user carries out at least one contextualization operation, with the possibility of modifications.

The question of information need only has meaning in relation to information, that is, as an interaction between an individual requiring certain information and a source that does or does not provide a response to this need.

6.6.2.5.3. Transformation of the information need in the context of information use

Multimedia information is viable in any decision situation on the condition that it is able to respond to information needs. The information system should not only facilitate but also personalize access to information. The final use to which information will be put has early implications; we must be able to bring out information through its representation, and this information must be useful in the creation and transfer of knowledge.

In our opinion, the value and relevance of information may be measured by its effective use by final users. The value of information is, therefore, determined late in the process by its use, and earlier in the process by a set of specific characteristics (the content of the information, its author, the sources of the information, and so on).

The user needs a piece of information in relation to their specific situation. The achievement of this aim necessitates a very precise approach from each user. At this point, the CI process that aims to clarify the information problem to improve perception and delivery of relevant information is useful.

The notion of information use is fairly prevalent in definitions of CI (e.g. [DAV 02, MAG 07, MAR 94]. To assist the user in the information seeking process, we propose that the user should clearly set out their information needs in the context of use parameter.

In accordance with our definition of the information need, the satisfaction of this need must involve the establishment of a relationship between a user and a piece of information in the context of use, with each user providing the necessary context as required.

In our proposals, the context of use is represented as a dynamic interaction between different elements (user, information, and use), representing the user’s information needs as more than just an external characteristic of the user. The specification of an information need is, by its very nature, dynamic, as it may evolve over the course of the information seeking process or through the interactions of a user with the information. This being the case, the nature of the context demands a precise goal, and the user is the person best placed to specify their needs in terms of representation of information.

The information need of a user is the product of an information problem. The information problem is a transformation of a decision problem, following the CI process suggested by David [DAV 06]. We will represent the information need as an attribute stemming from the <information_problem> attribute, which itself flows from the <decision_problem> attribute.

The <decision_problem> attribute is modeled by the name of the decision problem, the date of formulation of the decision problem, the domain to which the problem pertains, the aim of the decision problem, the name and identifier of the decision problem, and the identifier of the user who formulated the decision problem.

The <information_problem> attribute is represented by the name of the information problem, the date of formulation of the information problem, the domain to which the problem pertains, the aim of the information problem, and the identifier of the user who formulated the information problem.

The <information_need> attribute is modeled by the following elements:

– the cultural context of the information need;

– the spatial context of the information need;

– the temporal context of the information need;

– the physical context of the information need;

– the name of the context of use of the information sought by the user;

– the subname of the context of use;

– the data defining the context of use;

– the identifier of the user who defined the context of use;

– the domain relating to the context of use and the specificities of this domain;

– the aim of the user in using the sought information and the stake element;

– stakes are represented by three subelements: object, signal, and hypothesis.

We have taken inspiration from the proposals set out in [BOU 04] in the context of research into modeling decision problems. This inspiration is most clear in what follows; in particular, we borrowed Bouaka’s “stakes” element. This corresponds to the aim of the user. The “stakes” element is made up of three subelements that characterize it: “object”, “signal”, and “hypothesis”. The “object” subelement corresponds to the object of the problem. The “signal” subelement corresponds to the decision environment triggered by observation of the object. The “hypothesis” subelement is linked to the level of signal of the object. The signal corresponds to that which leads the user to seek information. The “stakes” element may be resumed as follows: “if I do not act on “object” in spite of “signal”, I run the risk of “hypotheses””.

The context of use element and its subelements aim to render the information needs of the user distinct. Once the context of use has been clarified, it becomes easier to save this information for future reuse. The representation of the information problem in this form enriches the information seeking phase as it allows us to find out if other users have formulated the same context of use, and what information was used.

The integration of representations of the user and their information needs on the same level as the representation of the characteristics of multimedia information is done with the aim of reducing the gap between user information needs and the representation of information, thus facilitating access to multimedia information.

6.7. Conclusion

The goal of an information seeking approach is to satisfy the expectations of an information seeker by fulfilling their information needs. Key notions of information seeking, such as representation, recording, and information needs are taken into consideration in our approach. The integration of user information needs via the context of use of research results in the representation of multimedia information seems, to us, to be the cornerstone of a successful information seeking process and is central to any “intelligent” information seeking system.

6.8. Bibliography

[AUF 00] AUFFRET G., “Structuration de documents audiovisuels et publication électronique. Constitution d’une chaîne éditoriale pour la mise en ligne de collections audiovisuelles”, Computer Science Thesis, Compiègne University of Technology, Compiègne 2000.

[BOU 04] BOUAKA N., “Développement d’un modèle pour l’explicitation d’un problème décisionnel: un outil d’aide à la décision dans un contexte d’intelligence économique”, Thesis, Information and Communication Sciences, University of Nancy 2, Nancy, 2004.

[BRE 06] BRETON P., PRO S., L’Explosion de la Communication. Introduction aux Théories et aux Pratiques de la Communication, La Découverte, Paris, 2006.

[CAR 03] CARAYON B., Intelligence Économique, Compétitivité et Cohésion Sociale, La Documentation Française, Paris, 2003.

[DAV 02] DAVID A., THIERY O., “Application of EQuA²te architecture in economic intelligence ICTEI”, IERA, Ibadan, Nigeria, 2002.

[DAV 03] DAVID A., THIERY O., “L’architecture EQuA²te et son application à l’intelligence économique”, IERA, Nancy, France, 2003.

[DAV 06] DAVID A., “La recherche collaborative d’information dans un contexte d’intelligence économique”, Le Système d’Information de l’Entreprise, Algérie-Télécom, Alger, Algeria, 2006.

[DOU 95] DOU H., Veille Technologique et Compétitivité, Dunod, Paris, 1995.

[GOR 06] GORIA S., Proposition d’une démarche d’aide à l’expression des problèmes de recherche d’informations dans un contexte d’intelligence territoriale, Doctoral Thesis, University of Nancy 2, Nancy, 2006.

[HIL 01] HILLMAN D., Guide d’utilisation du Dublin Core, 2001, available online at: www.bibl.ulaval.ca/dublincore/usageguide-20000716fr.htm.

[IHA 08] IHADADENE M., CHAUDIRON S., L’étude des Dispositifs d’Accès à l’Information Électronique: Approches Croisées, Hermès, Paris, 2008.

[JOO 01] JOO S., GRABLE J.E., “Factors associated with seeking and using professional retirement-planning help”, Family and Consumer Sciences Research Journal, vol. 30, pp. 37–63, 2001.

[KIS 07] KISLIN P., Modélisation du problème informationnel du veilleur dans la démarche d’intelligence économique, Doctoral Thesis, University of Nancy 2, Nancy, 2007.

[KNA 07] KNAUF A., Caractérisation des rôles du coordinateur-animateur: émergence d’un acteur nécessaire à la mise en pratique d’un dispositif régional d’intelligence économique, Doctoral Thesis, University of Nancy 2, Nancy, 2007.

[LEC 97] Le COADIC Y.F., Usages et Usagers de l’Information, Nathan, Paris,1997.

[LEC 08a] Le COADIC Y.F., La Science de l’Information, 3rd edition, Que sais-je?, no. 2873, PUF, Paris, 2004.

[LEC 08b] Le COADIC Y.F., Le Besoin d’Information: Formulation Négociation, Diagnostic, ADBS, Paris, 2008.

[LEG 93] LEGENDRE R., Dictionnaire Actuel de l’Éducation, 3rd edition, Guérin, Montréal, 1993.

[LEL 00] LELONG B., MALLARD A., “La Fabrication des Normes”, Réseaux, vol. 18, no. 102, 2000.

[LES 06] LESCA H., “Veille stratégique pour le management stratégique de l’entreprise”. Economie et sociétés, Séries sciences de gestion, vol. l5, no. 20, pp. 31–50, 1994.

[LEV 01] LEVET J.L., L’Intelligence Économique Mode de Pensée, Mode d’Action, Economica, Paris, 2001.

[MAG 06] MAGHREBI H., DAVID A., “Toward a model for the representation of multimedia information based on users’ needs: economic intelligence approach”, IV International Conference on Multimedia and Information and Communication Technologies in Education (m-ICTE), Spain, 2006.

[MAG 07] MAGHREBI H., DAVID A., “Integrating users’ needs into multimedia information retrieval system”, 3rd International Conference on Computer Science: ATINER, Athens, Greece, 2007.

[MAG 08] MAGHREBI H., DAVID A., “Open system for indexing and retrieving multimedia information”, 36th Canadian Association for Information Science Conference (CAIS/ACSI), Vancouver, Canada, 2008.

[MAR 94] MARTRE H., CLERC P., HARBULOT C., Intelligence Économique et Stratégie des Entreprises, La Documentation Française, Paris, 1994.

[PRO 05] PROULX S., “Penser la conception et l’usage des objets communicationnels”, in SAINT-CHARLES J., MONGEAU O., (eds), Communication. Horizon de Recherches et Depratiques, Presses de l’Université du Québec, Québec, pp. 297–318, 2005.

[REV 98] REVELLI C., Intelligence Stratégique sur Internet, Dunod, Paris, 1998.

[ROU 06] ROUSSI S., “Standardisation de l’indexation de documents numériques: le cas des questionnaires pour l’évaluation des connaissances”, Indice, Index, Indexation, ADBS 2006.

[SAL 00] SALLES M., “Problématique de la conception de méthode pour la définition de système d’intelligence économique”, Revue d’intelligence économique, no. 6–7, Association française pour le développement de l’intelligence économique, October 2000.

[TRI 03a] TRICOT A., “IHM, cognition et environnements d’apprentissage”, in BOY G., (ed.), L’Ingénierie Cognitive: IHM et Cognition, Hermès, Paris, pp. 411–447, 2003.

[TRI 03b] TRICOT A., “Pour une approche ergonomique de la conception d’un dispositif de formation à distance utilisant les TIC”, Revue STICEF, vol. 10, pp. 1–27, 2003.

[TRI 04] TRICOT A., “Besoin d’information”, Argos, no. 36, pp. 36–39, 2004.

[WEI 99] WEIBEL S., “The state of the Dublin core metadata initiative”, D-Lib magazine, April 1999, available online at: www.dlib.org/dlib/april99/04weibel.html.

---

¹ Chapter written by Hanène MAGHREBI.

1 Systems theory is a principle according to which everything is a system, or everything can be conceptualized following system-based logic. This principle led to the establishment of systemics as a scientific method. Durand’s 1979 work, La systémique, constitutes an essential reference for understanding systems theory.

2 For Norman, an artifact is a tool designed to conserve, expose, and process information with the aim of satisfying a representation function. The artifact is used to compensate for the lack of necessary objects.

3 We carried out a study among students at the Institut Européen du Cinéma et de l’Audiovisuel to study their needs in terms of audiovisual information.

4 Afnor: Association française de normalisation, French standardization organization (www.boutique.afnor.org/NEL1Accueil) NormeEnLigne.aspx).