site modeling and development of economic intelligence systems

TEAMSITE

Modeling and Development of EconomicIntelligence Systems

Table of contents

1. Team . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12. Overall Objectives . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1

2.1. Introduction 12.2. Brief presentation of our issues as related to an Environment for the Realization of an EI project

(EREIP) 22.3. Highlights 4

3. Scientific Foundations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .43.1. User modeling in information systems 43.2. Modeling of the interaction between the user and a mediator 63.3. Decision-maker profiling, decision-problem understanding 63.4. Design and exploitation of data warehouse 7

3.4.1. From the design of IS to the design of S-IS 73.4.2. Data warehouses and strategic information systems 8

4. Application Domains . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .84.1. Information retrieval 84.2. User modelling and Contextualisation 94.3. Decision support system 94.4. Risk management 9

5. Software . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 95.1. METIORE 9

5.1.1. Functional characteristics 95.1.2. Data model 10

5.2. METIORE-WISP 105.3. RUBICUBE platform 105.4. MECOCIR 115.5. FuzzyMatch Tool 11

6. New Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 116.1. Collaborative information retrieval 116.2. Document annotation 126.3. Multimedia document modeling 126.4. Knowledge capitalization 136.5. Knowledge and strategy elicitation 136.6. Risk management in Economic Intelligence 13

6.6.1. Information risk management 136.6.2. Risk among decision actors 14

6.7. Economic Intelligence organizations and systems 156.7.1. The coordinator: the manager of EI process 156.7.2. Territorial intelligence: practices and conceptions 15

7. Contracts and Grants with Industry . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 157.1. Institutional partnerships 157.2. Contracts, research agreements and industrial actions 167.3. Visit, Invitation 16

8. Dissemination . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 168.1. Animation of scientific community 168.2. Teaching 178.3. Administrative loads 17

9. Bibliography . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .17

Table of contents

1. TeamFaculty Member

Amos David [ (Team Leader) Professor, Information and Communication Sciences (PhD Computer Science)Université Nancy 2, HdR ]Odile Thiéry [ Professor, Computer Science, Université Nancy 2, France, HdR ]Fabrice Papy [ Professor, Information and Communication Sciences, Université Nancy 2, France, HdR ]Gérald Duffing [ Associate professor, Computer Science, ICN ]Stéphane Goria [ Associate professor, Information and Communication Sciences, IUT, Université Nancy 2,France ]Audrey Knauf [ Associate professor, Information and Communication Sciences,Université Nancy 2, France ]

PhD StudentVictor Odumuyiwa [ Université Nancy 2, (thesis defended in 2010) ]Olufade Onifade [ French governement scholarship since 2007(co-supervised thesis defended in 2010) ]Olusoji Okunoye [ French governement scholarship since 2007 (co-supervised thesis to be defended in 2011) ]Fausat Oladejo [ French governement scholarship since 2007(co-supervised thesis to be defended in 2011) ]

Administrative AssistantDelphine Hubert

OtherHanene Maghrebi [ Associate lecturer IUT, Information and Communication Sciences, Université HenryPoincaré, France ]Babajide Afolabi [ Associate Professor, Computer Science, Obafemi Awolowo University, Ile-Ife, Nigeria ]Charles Robert [ Associate Professor, Computer Science, University of Ibadan, Nigeria ]David Bueno [ Associate Professor, Computer Science, University of Malaga, Spain ]Philippe Kislin [ Associate Professor, Information and Communication Sciences, Paris VIII University,France ]Chedia Dhaoui [ Research Assistant, Faculty of Economics and Business, University of Sydney, Australia ]Najoua Bouaka [ Université Nancy 2, (Thesis defended in December 2004) ]Frédérique Peguiron [ Chief librarian, Université Nancy 2 ]Patrick Nourrissier [ NetLorConcept ]

2. Overall Objectives

2.1. IntroductionThe focus of this team is on the design and development of an environment for the realization of EconomicIntelligence (EI)1 projects in which strategic information system plays an important role.

Definitions of EI

1. It is a set of coordinated actions of search, processing and distribution for exploitation, of usefulinformation for economic actors. These actions are carried out legally with all the necessaryprotection for the safeguard of the company’s patrimony, and with the best quality, delay and cost2.

2. It is the process of collection, processing and distribution of information with the goal of reducinguncertainty in taking strategic decisions3.

1The closest term in English is "Competitive Intelligence"2Martre, H., "Intelligence Économique et stratégie des entreprises", Rapport du Commissariat Général au Plan, Paris, La Documenta-

tion Française, 1994, pp. 17,183Revelli, C., "Intelligence stratégique sur Internet", Paris, Dunod, 1998, pp. 18,19

Team SITE 1

2 Activity Report INRIA 2010

For us, these two definitions characterize our research objective. In fact, the increasing predominant role ofinformation in socio-economic sectors or in organization in general is evident.

Communication and information technologies, particularly data processing and Internet, make it possible tomanage information of different nature: primary information, secondary information, tertiary information,information with added value. Whereas primary information is the direct product of authors, secondary andtertiary information are transformations into reduced models of the primary information to feed databases.Information with added value is the product of analysis and synthesis of these various types of information.Information is used more and more as object of reference and aid tool for strategic decision making. Theconcept of EI becomes indispensable for the production of interpretable indicators for decision making basedon the use of the company’s internal and external information.

The process of EI is based on the process of watch. We distinguish here two types of watch: tactical andstrategic watch. The tactical watch feeds the company’s actors with information, and the studied temporalhorizon relates to the present and the very short term (information on the economic situation). The strategicwatch is characterized by the distribution of information to the entities of management of the company(Directorate-General, Direction of Plan, Direction of Strategy...). The studied temporal horizon is the present,the very short term, medium term and long term.

In order to consider EI as a research object, we believe that it should be considered as a process. This processcan be defined as follows:

a) Identification of the decision problems to solve in terms of threat, risk or danger.

b) Transformation of decision-problem into information search problem.

c) Identification of relevant information sources.

d) Validation of the information sources.

e) Collection and validation of collected information.

f) Processing of the collected information for the calculation of indicators.

g) Interpretation of the indicators.

h) Decision making for the resolution of the decision problem.

The phases (a), (g) and (h) are of particular importance because they determine the success of an EI project.Indeed the EI process is a global process where the orientation chosen in each phase will determine the type ofthe final result. For example, if a company decides to make decisions in offensive manner towards a competitor,the information to use, the sources of this information, their processing and the interpretation of the final resultwill determine the effect of the company’s decision. It should be noted that information used is not onlyfactual as in corporate databases but heterogeneous by nature as well by their source and by their validity ortheir range. It is also informal because it might not have been published.

2.2. Brief presentation of our issues as related to an Environment for theRealization of an EI project (EREIP)Based on the EI process adopted by the research team as presented above and as illustrated in figure 1, anenvironment for the realization of an economic intelligence project should allow the implementation of all thephases in the process or assist in their implementation. Therefore, the environment is based on a "strategicinformation" system that will integrate all the functional characteristics necessary in all the phases.

We have identified four invariants in all economic intelligence projects, which are, a decision problem, adecision maker, required information for solving the decision problem, and the protection mechanismsfor the organization’s patrimony. Our work is based on the hypothesis that the efficiency of the environmentis highly dependent on the adequacy of the collected information as a result of the EI process in solving thedecision problem.

2 LORIA activity report 2010

Project-Team SITE 3

Figure 1. Architecture of an Environment for the Realization of an EI Project

Team SITE 3


Therefore, as illustrated in figure 1, there should be a module for the specification of the decision problemand this step should be the starting point in the realization of an EI project. The main issue here concerns theproblem of specifying the decision problem by the decision-maker and making it comprehensible for the otheractors, particularly the person that will be charged with collecting relevant information. Apart from the studyon information representation, the study here embodies studies in psychology, cognition, sociology and otherfields concerned with the study of "needs".

The second module concerns the collection of the relevant information for solving the decision-problem.One of the issues related to this phase concerns the transformation of the decision-problem into informationretrieval problems. The success of this phase depends on how well the decision-problem has been understoodby the person charged with the information collection. This explains why the first phase which concerns thedecision-problem specification is primordial for the success of all the succeeding phases in the environment.Another issue in this phase concerns the structures and the forms of the collected information. Often, theinformation collected is of various forms (text, images, sound or combination) and of diverse structures.Some information are not structured or semi-structured. So the issue of structuring information, integrationof heterogeneous information and representation of multimedia information in preparation for their use indecision-problem solving constitute some study areas in the research team.

Another module in the environment is that which will allow for the exploitation of the collected andpreprocessed information. The research team has come up with a functional model for this module whichis called EQuA2te. This model resulted from the study on personalization of information retrieval systemsas presented in section 3.1. EQuA2te stands for Explore, Query, Analyze and Annotate. In fact we believethat the end-users of the information base in the EREIP should be able to perform any of the functions inEQuA2te.

The information base in the information system for managing the collected information is just a part of theinformation needed for solving the decision-problem. There is also need to acquire and store all the generatedknowledge by the various actors during the process of solving the decision-problem, from the specificationof the decision problem to the interpretation of the indicators generated from the collected information. Thematerialization of the knowledge will constitute the knowledge base.

Finally, two transversal problems concern the management of risks in the process and how to coordinate thewhole process.

2.3. Highlights1. Three Phd thesis defended of which one co-supervised.

2. In terms of capacity building, the theme of the research team has enjoyed tremendous recognitions.Professor Odile THIERY was promoted to the category of "Exceptional class" and Professor AmosDAVID was promoted "First class".

3. In terms of research work exchanges, we received two invited lecturers from Nigerian universitieswith which we are in collaboration.

4. Fabrice PAPY joined the team in september as Professor and Audrey KNAUF as Associate Profes-sor.

5. Audrey KNAUF won the Economic Intelligence special jury price for his book : "Les dispositifsd’intelligence économique : Compétences et fonctions utiles à leur pilotage", proposed by CED-3AF(two french associations : Culture, Economie, Defense - Association Aéronautique et Astronautiquede France).

3. Scientific Foundations

3.1. User modeling in information systems


Project-Team SITE 5

The objective of this research topic is to allow the taking into account of the characteristics of the actorsin the process of EI by user modeling. Our goal is to propose models and methods making it possible toproduce results as relevant as possible by the system, in response to the user’s information need. This shouldbe applicable to traditional information retrieval and also in a context of strategic information systems wherethe concept of data mart corresponds to the modeling of the end-user of the system, for example the directorfor whom such a system is essential for good decision making.

The foundation of personalization techniques of the system’s response is based on the concept of relevance.The relevance of a solution is often measured compared to the user’s query. As the query does not necessarilyrepresent the user’s information need, the user judges the relevance of the response compared to his need,which does not correspond to the measure of relevance by the system. The technique which consists of theevaluation of the proposals of the system to indicate their degree of relevance is integrated into certain IRS.Thus, the system has knowledge of the adequacy of the system’s response to the user’s need. Nevertheless, thesystem obtains this need only by an estimate based on the queries. Instead of calculating this need, we proposeto integrate its representation in the user model. This constitutes the originality of our proposal. This amountsto storing in the Strategic Information System, among the meta data of the system, an explicit representation ofthe structure of various data marts. These orientations of taking into account the actor during the developmentand then during the exploitation of the data warehouses is a major field of our research.

The objective of modeling the user is to be able to personalize the system’s response. User modeling is theway of representing a user and his behaviours. That also relates to the way of exploiting knowledge availableon the user. Three categories of model are proposed:

(a) User profile: where to a user is associated his query that represents his need. In this context, theuser’s need is relatively stable. The profile is applied to new information in order to propose themost relevant information to him.

(b) The implicit model: where the user’s behaviour and his preferences are represented implicitly. Forexample the visualization of a document by the user can be interpreted as adequacy of the documentcompared to his request.

(c) The explicit model: where user’s behaviour and his preferences are also represented but accordingto user’s specifications. For example, even if the user visualizes a document, it is necessary that heindicates his opinion on the degree of relevance of the document compared to his request.

The exploitation of a user profile (a) is generally individualized. The implicit or explicit model, (b) and (c),can be individualized or processed using the method of stereotype. By the technique of stereotype, the usersare grouped into classes and an interpretation applies to all the users of a class.

The representation of the cognitive parameters on the users, for example the parameters necessary to knowtheir levels of knowledge for a better interpretation of their request, requires the safeguard of the user modelthrough sessions and individualized.

Our work on personalization of response in IRS began within the framework of our work on Computer-Assisted Learning (CAL), by student modeling, in a context of training using images. We proposed anexplicit model represented by cognitive parameters for each learner. The cognitive representation is basedon the cognitive phases identified in a process of human learning. Four phases, which correspond to levels ofevocative habits, were integrated into the model:

• The phase of observation: the learner takes note of his environment by the process of observation;

• The phase of elementary abstraction: the learner designates the objects observed by names, whichalso corresponds to a phase of acquisition of vocabulary;

• The phase of symbolization and reasoning: the learner employs specialized vocabularies whichconcern a level of abstraction of high concepts. For example, somebody on a low level of abstractioncan say "I see a bird", but cannot say "I see a piscivorous bird" (i.e. a bird that feeds on fish, implyinga form of abstract classification and specialized vocabulary);

Team SITE 5


• The phase of creativity: the learner discovers and adapts knowledge which is not presented in anexplicit way in the system. For example, during the experimentation using our prototype BIRDS,a pupil identified the fact that all the carnivorous birds have short legs, which was not explicitlypresented in the system.

As a result of these studies, we developed a functional model that we believe should be implemented in an IRSin the context of economic intelligence. This model is called EQuA2te, meaning Explore, Query, Analyzeand Annotate. The four functional characteristics represent the four levels of evocative habits.

3.2. Modeling of the interaction between the user and a mediatorThe objective of this research topic is to propose models of interaction between a user and a mediator, orbetween two users. In relation to user modeling, the issues in this are relate to the determination of the statusand the knowledge on the collaborating actors. In a context of EI, the taking into account of the status of acollaborator is essential because it is necessary to check his competence and the degree of confidence thatone can have in him. This last point shows the relation of the studies in user modeling and collaborativework. We hope to refer on the one hand to the results of the studies on protection of patrimony, in particularthe concept of misinformation, and on the other hand to the results of studies on interactions between usersof information resource centers. The specificity of this topic relates to the introduction of the concept ofinterpersonal communication which implies two users in an information retrieval system. The results of thisaxis will allow the sharing of domain knowledge as well as human competencies. The scientific community ofinformation and communication sciences is very interested in the studies of this axis because the results willmake it possible to try out interpersonal communication models in a precise context of information retrieval.There are currently tools for forums of discussion and chats but these tools do not allow the coordination ofthe dialogue and the process of information problem solving.

Collaborative work is already initiated in the field of groupware, where the objective is on designing modelsfor collective management of tasks. Systems’ architectures available make it possible to share data and towork in virtual spaces. This approach is becoming increasingly important for maximizing the capitalization ofknowledge in organizations in order not to lose it with time. This is currently called knowledge management.In the field of information retrieval, collaborative work does not relate to only automated knowledge manage-ment but also to the interaction between the human actors in the process of information retrieval. The mostfrequent situation is that of interaction between a user and an expert in the methodology of information re-trieval, for example a researcher and a person in charge of a library. Knowledge to be exploited in this contextrelates to knowledge on the researcher, his current needs and the history of the current need. Studies in infor-mation retrieval show that the knowledge of the experts in methodology of information retrieval on the usersaccelerates obtaining the solutions to the user’s needs.

Amos David proposed for prototypes STREEMS and METIORE three modes of use:

• Autonomous mode where only the user uses the system.

• Observation mode where one user observes another user while carrying out a process of informationretrieval without being able to intervene.

• Collaborative mode where two users intervene together to carry out the process of informationretrieval.

3.3. Decision-maker profiling, decision-problem understandingEI is mainly an activity of decision-problem solving. This activity is carried out by the decision-maker whomust formulate exact description of the decision-problem in collaboration with the watcher (the person chargedwith information collection) who must locate, supervise, validate and emphasize the strategic informationneeded for solving the problem. In order to optimize the sharing of knowledge between the decision-makerand the watcher, we tried to establish a bridge between two models: the model of the decision-maker and themodel of the watchers information search problem. These linked models can help to increase the precision


Project-Team SITE 7

of the representation of the various parameters of the problem and to allow a greater performance and anoptimization of the resolution to the decision-problem.

One of the issues in this topic is Information Retrieval Problem expression activity which is realized betweenthe person requesting for information and an expert charged to resolve the problem. For this, we defined aset good expression principles (notably inspired from Tauli and Grices pragmatic conversation principles).These are three principles: Adhesion principle; Reformulation principle; and Memorization principle. Thesecond issue concerns a model called WISP (Watcher’s Information Search Problem). It is an extension andan application of a EI metamodel to describe and help the user formulate his information needs during theprocess of Information Retrieval. It is tridimensional and multifaceted integrating: 1) analytical dimensionview; 2) methodological dimension view; and 3) operational dimension view.

3.4. Design and exploitation of data warehouse3.4.1. From the design of IS to the design of S-IS

Current evolution of information systems (IS) in companies towards client-server architectures requiresadapting the traditional methods of IS design. Indeed, in these new architectures, the client-server applicationsappear as components of various types: storage components, processing components, dynamic components ofexecution. The basic idea of our proposals is that it is possible to use the same model throughout the successivephases of this design, to obtain the whole software components constituting the automated information system.This is the model OOE, standing for Object Operation Event. We introduced it starting from our research onREMORA method and of the study of commercial MERISE method in France.

Presently, the standard in company’s data processing remains the MERISE method even if the tendency istowards object method design such as OMT or UML or at least based on more advanced models such as that ofMERISE/2. Without occulting the merits and achievements that MERISE method allowed in company’s dataprocessing, it should be recognized that multiplicity of levels of design and realization of IS, the multiplicityof models at various phases and maybe especially the lack of rules for passing from one level to another, orfrom a model to another seriously handicap the use of the method. This is why we are trying to define not onlyone single model but also the set of rules for passing from conceptual modeling to logical-physical modelingwhich will lead to the list of software components to be programmed. We present here briefly this model andits bases:

• The OBJECT category represents the concrete or abstract system elements and its environment i.e.the organization. For example an order.

• The OPERATION category represents the actions within the system or its subsystems. For example,analysis of the order following its entry which will modify the quantity in stock of the orderedproduct.

• The EVENT category represents the events occurring in the system in the course of time. Forexample, the arrival of an order which starts the analysis of an order, out-of-stock conditioncharacterized by the fact that the quantity in stock became lower than a certain stock, etc.

In this approach, the state of the system is defined at a given time by the state of the objects which belong toit at that moment. The system evolves in the course of time following the execution of operations which aretriggered by the system’s internal or external events. The operations act on the objects and cause changes ofstate which in turn can be events.

This proposal is called causal dynamics because the same cause always produces the same effects. The origincomes from the method REMORA which was popular world wide in the domain of IS design methods. This isbecause, well before becoming obvious since 1975, we proposed a modeling method based on simple conceptswhich then not only make it possible to build data bases almost automatically but also to build transactions onthe data base by transformation rules using the conceptual solution (starting from the events and operations).

The idea that we retain is that the strategic information systems are particular information systems and that themodeling technique that we proposed previously can adapt to these new IS.

Team SITE 7


3.4.2. Data warehouses and strategic information systemsData warehouses have become not a phenomenon of mode but an essential instrument for a good managementof an organization. They are at the base of any strategy and decision making of a company. The followings aresome definitions of the basic concepts.

A data warehouse is a data base organized to meet the specific needs for decision making. This base containshistorical information on the company, its operation and its environment. It is fed from production data basesand from external information of the company. It is thematic, related to a field that interests the decision-maker, having a temporal reference, sure i.e. whose quality was checked, easy to access, non volatile andregularly updated. In fact, a data warehouse is an integrated view of the organization. It is the core of StrategicInformation Systems (SIS).

Information systems (IS) can be strategic from two angles. On the one hand all the current IS of organizationscomprise strategic information and allow the automation of the organization as well as making it possibleto satisfy the strategic objectives of the direction (example: an IS for improving the inventory control,development using accounting incomes of summary tables in a Speadsheet). This is called S-IS (Strategic"information system"). In addition, more and more of IS are dedicated only to decision making (example: ISas aid for market choice). This type of IS is called SI-S ("Strategic information" system). In this case, it isthe IS in its entirety that is devoted to strategic decisions and comprises only information of strategic nature(example: the results of sales turnovers by country over several years). We are interested here in SI-S i.e. IS ofthe second type, those which are directly of concern of researchers in EI field. A data warehouse serves as a linkbetween the two types of IS. The company’s IS are the first built. They are various and divers, and comprisestrategic information. It is necessary to extract from it information necessary for decision making and alsotheir structure (called meta data). This constitutes the relational warehouse (so called because it is currentlymanaged by a relational DBMS). From this data warehouse are extracted multidimensional data bases, socalled because they make it possible to view the organization from various angles or dimensions (example:from axis of time, sold quantity of products or sales turnover). These multidimensional data bases constitutewhat is called strategic information system. Indeed they are made up only of data suitable for decision.

This data warehouse gives rise to, by filtering in terms of user profiles (or finally a user model), to datamarts. These are the smaller dimensions of the data warehouse designated to a department or a function of theorganisation: Marketing, Finance ... They are updated periodically, based on a multidimensional view of data,and are non modifiable by the users.

We are well aware that the design of SIS requires a particular design steps and a complex modelingmethodology. However the related idea of modeling based on the minimum of reusable concepts at each phaseappears completely realistic to us. This research orientation has already been started within the framework ofa previous DRT project with SNVB Bank.

4. Application Domains4.1. Information retrieval

Participants: Amos David, Stéphane Goria, Victor Odumuyiwa, Fausat Oladejo.

Evolutions in Information and Communication Technologies (ICT) have made it possible to extend thecontexts and domains of application of information retrieval systems (IRS). IRS that initially focused on singleend-users become tools for collaborative activities in solving information problems. In our research team, wehave focused our work on collaborative information retrieval in which two or more end-users can collaboratesimultaneously in solving an information need problem to achieve various objectives [11], [15], [18], [26].The results of this work are applicable to any socio-economic organization in which the sharing of knowledgeis highly important. The areas in which we are investigating include

• medical structures where the end-users, the information needed and the objectives are divers andheterogeneous;


Project-Team SITE 9

• education and research activities in which knowledge sharing is indispensable, involving the trans-formation of knowledge into manageable information;

• management and commercial activities in which knowledge sharing and knowledge capitalization inthe process of decision-making and process of innovation are indispensable

4.2. User modelling and ContextualisationParticipants: Amos David, Hanene Maghrebi, Olusoji Okunoye.

The ICT have revealed the polysemous nature of information elements provided through the numerousinformation sources. This explains why the semantic interpretation of an information element can no longerbe considered as universal but highly contextual depending on the author, the end-user and the applicationdomain. The efforts in this area include the work on semantic web. Our interest is focused not only on webdocuments but on any type of information from any type of information (text, fixed image, video) fromany type of information source (web pages, databases, etc). The results of this work are applicable in anydecision-making environment in which the information provider is not the user of the information. Exampleof application areas include the activities in using information as backbone of decision process such as in anytype of competitive intelligence project, and in training where there is need for adapting existing informationor knowledge [9], [16], [25].

4.3. Decision support systemParticipants: Amos David, Stéphane Goria, Audrey Knauf, Fausat Oladejo.

Decision support systems through the use of informatics or human organization systems is becoming more andmore predominant. The issues here include knowledge acquisition and capitalization, human organization foran efficient management of knowledge so as to reduce uncertainty resulting from a decision to take, etc. Theapplication areas include government policy initiatives to coordinate economic strategies, like in competitivepoles or clusters. There is need to study the roles and responsibilities of the actors involved [8], [11], [14],[18].

4.4. Risk managementParticipants: Odile Thiéry, Gérald Duffing, Olufade Onifade.

A very strong hypothesis in computer applications is that information is objective and that all the end-users areof good intentions. Our work on risk management is based on the fact that one can not have absolute confidenceon information to use nor on the end-users. This implies that there is need to propose a system that will takeinto consideration the various risk factors. The application domains range from quality of information, themeasure of trust on an end-user, the degree of confidence on an information source, etc. The applicationsinclude information retrieval systems [19], [20], decision process [28].

5. Software

5.1. METIOREThe system METIORE is the core system that we are using to experiment the various models and methods ofour team. Two versions of METIORE are presently under experimentation: METIORE and METIORE-Wisp

5.1.1. Functional characteristicsMETIORE is a system used for the management of bibliographic references: bibliographic references ofLORIA since 1980. It is used as test base to implement our proposals on user modeling and the functionalmodel EQuA2te. It is now implemented using Java programming language (we are presently testing theJAVA version) so that the scientific community (LORIA and outside) can try out the system.

Team SITE 9


The functional model EQuA2te implemented in METIORE is an acronym:

• Explore: Allow the exploration of the data base to discover the content of the information base;

• Query: Allow information retrieval by query;

• Analyze: Allow a global analysis of the data of the system;

• Annotate: Allow the user to annotate the proposed results either to express his evaluation of theproposed solution or to attach some other personal information on the proposed result.

In addition to these functions, the system:

• allows collaborative information retrieval by two users, for example between a person in charge ofinformation resource center and a researcher;

• permits the submission of documents by individual users. These submitted documents can beanalyzed along with existing documents;

• alerts users of updates on the information base.

5.1.2. Data modelThe behaviour of the user is represented by his activities: exploration, query, synthesis, and evaluation ofthe answers of the system. Each activity is regarded as a document. The objects managed in the system areregarded as documents. Each bibliographical reference is represented as a document.

XML format is used to represent the objects: users’ behaviour and bibliographic references. The documentrepresentation and the search engine in the current version of METIORE is entirely in XML.

5.2. METIORE-WISPThis prototype implements the wisp model and takes up fundamentally, original characteristics of the ealierMETIORE. The goal of this software consists in helping the watcher in his information retrieval process byrecording search activities for supervising, understanding and future reuse. The new architecture is composedof different modules surrounded by a data warehouse:

• A GUI that includes in one hand, a full web browser to navigate between web pages and documentsstored in the database and in the other hand spaces for annotation and metadatas layers.

• A full text index network for retrieving documents with any information elements.

• Additional modules for visualisation, information mapping, cross-analyzing and data mining.

METIORE-Wisp is created with Windev IDE for RAD and its 5GL. Some modules, like the cross-analyzeris developed in C language to optimise the processing speed of co-occurrence analysis. Others are DLL orActiveX objects. Our approach considers that all objects are treated as documents so the prototype uses entirelyXML and DOM.

5.3. RUBICUBE platformWe developed a decisional paltform called "RUBI3" User and user’s need representation during queryingbuilt on a decisional schema. This open source application offers some innovative perspectives to informationcontent processing because it uses XMLA schema for data analysis. We adopted this new analysis model forour research work in which we used a programming language that allows MDX querying of the database. Thisfree decisional tool relies on Mondrian OLAP engine. We elaborated some analysis schemas in XMLA, whichallows by using MDX queries to carry out multidimensional analysis via a web interface according to differentactors profiles. Our application described above explains data retrieval for dynamic analysis. The retrieved andanalyzed data result into data mart views based on actors.


Project-Team SITE 11

5.4. MECOCIRMECOCIR is a collaborative information retrieval system which implements the COCIR model. The nameMECOCIR is formed from COCIR model and METIORE (which is the mother system in our researchteam). From our study of the Collaborative Information Behaviour (CIB) of users, we discovered that usersexhibit a CIB in their quest to resolve their information problem. The system’s objective is to allow users toexpress their information problem, to choose the information system from which to carry out their informationretrieval activities, to interact with other users synchronously and asynchronously, to view past users activities,to exchange synchronous and asynchronous annotation during information retrieval activities. This systemfacilitates knowledge sharing among users in:

• the definition and clarification of information problem

• the choice of information system

• the formulation and clarification of queries based on the chosen information system

• the evaluation of results

MECOCIR is developed with java, php and Ajax. The system relies on the morzilla XULrunner for renderingstatic and dynamic web pages. The interprocess communication between our system and the morzilla engineis achieved through XPCOM.

5.5. FuzzyMatch ToolFuzzyMatch can be considered as a search engine which helps to reduce risk during decision making. Weemployed a database of a prominent international institute in Nigeria. This database has close to five thousand(5000) entries which are cleaned to represent extracted, transformed, and loaded data example. We also forconvenience and example purpose include other entries that were not in the original database to depict therobust nature of the system. The system was rigorously tested with various databases both as stand alone or anetwork environment. The results are similar. However, we have not been able to test on very large databaseto determine the length of time and other factors. We are however convinced that, even if the search timeincreases because of the complexities, the accuracy of match would not be compromised. We have testedthe system on systems with 2.0 GHz, Intel Pentium IV minimum system requirement. The implementationtools are JAVATM, MATLAB version 7.3.0. and WAMP5. JAVATM language is of choice because of itsportability and robust inbuilt 270 functions for string manipulation. MATLAB provides excellent environmentfor implementing fuzzy logic and its inference system. The JMATLINK provide the link to integrate thefunctionality of fuzzy system with Java. The WAMP5 houses the database employed for the search tests.

6. New Results

6.1. Collaborative information retrievalParticipants: Victor Odumuyiwa, Amos David.

Observations of users information behaviour show that users collaborate during information seeking andretrieval processes [26]. Collaborative Information Retrieval (CIR) could be seen to be at the intersectionof information search, communication, knowledge management and social networks [15]. CIR consists ofmethods and systems for managing collective activities of users in information retrieval process in order tofacilitate direct collaboration among the users thereby enabling knowledge sharing among them. The emphasishere is not only on the collective activities but also on the direct collaboration among the users which leadsto sharing of tacit knowledge. We proposed two models and developed a collaborative information retrievalsystem (CIRS) to manage collective activities and to facilitate synchronous and explicit collaboration amongusers during information retrieval (IR) process [15]. The first model is a collaboration pyramid made upof six important phases necessary for the success of CIR. These phases include: (1) initial trust phase, (2)

Team SITE 11


consensus building phase, (3) communication phase, (4) knowledge sharing phase, (5) group awareness phaseand (6) division of labour phase. The second model is a communication model for collaborative informationretrieval (COCIR). This model is a representation of the collaborative context for knowledge sharing in IR. Theattributes of this model facilitate the contextualisation of interactions during collaboration in order to manageexpressed knowledge and consequently allow knowledge sharing among collaborators. The two models wereimplemented in a CIRS called MECOCIR which we developed to validate our propositions.

6.2. Document annotationParticipants: Olusoji Okunoye, Amos David.

The study focuses on design and development of model for capturing and exploiting user interpretationsof document through annotation. Annotation as knowledge is the expressed knowledge of annotator onthe contextualized document. Actors knowledge is elicited by capturing his interpretation on the documentof interest. Such annotation may likely base on his intellectual capacity, experience, cultural backgroundand environment. Storing information on who annotates what and how personal characteristics influenceannotations made could serve as form of knowledge on the annotator. Annotation patterns of individual actorovertime could reveal interesting patterns that could assist in discovery new knowledge about him as wellas in decision problem resolution [17]. We propose the use of annotation representation as Attribute-Value-Pair (AVP) [28] as a medium of capturing EI actors interpretation to document of interest. AVP is a datarepresentation methodology which allows data element to be represented with a set of attributes and values.Ability of an actor to express his observation and/or contextualize document object as AVP annotation couldimprove significantly the effectiveness of such value-added information. While most of existing annotationmodels allows annotators to add annotation to document of interest as atomic object, it is believed thatannotation represented as an attribute-value pair will offer enhanced interpretation functionalities. Therefore,the use of AVP will provide a good basis for data restructuring, data mining, robust exploitation, knowledgeelicitation, among others. We also proposed AVP search algorithm retrieving required information fromInformation System based on the semantic context of user query. Based on the above, we have come up withannotation model called Annotation Model for Economic Actors (AMTEA) [16]. We have used the modelto develop an annotation system. In the system, user can browse, annotate, query and search using our novelAVP search algorithm. The system exploitation unit is based on EQuA2te architecture.

6.3. Multimedia document modelingParticipants: Hanene Maghrebi, Amos David.

User studies focusing upon multimedia information needs, its uses and representation are still very limitedin the information retrieval world. Our goal is to provide an empirical basis to acquire information that canhelp user in his retrieval, to better satisfy his information needs, and therefore focus on the problem of howto reduce the distance between multimedia information representation and the user’s needs. The complexity isthat multimedia information is compound information composed of a variety of information elements (sound,text and image), which are difficult to represent, index and access. In fact, in multimedia information, the textor the sound elements accompanying the image guides the user in his interpretation of the image and bringout a meaning in which the user never thought of. The semantic concepts depicted in, or otherwise emergingfrom, an image is specific to a user. We can summarize our approach, on which we base the development ofour system, on three phases. Firstly, we believe that the starting point of our research is a group of problemsthat can be summarized as the distance between user’s information needs and the multimedia information thathe needs. Secondly, in order to reduce the gaps, we propose to model user at the same level as multimediadocument for better satisfaction of information needs in terms of time and response relevance [25]. Themodeling that we propose is done by contextualizing the use of the retrieved information. Thus the user’sinformation need is translated into information use context. We noticed that it is easier for a user to expressinformation use context than to specify his information need. We think that there is a high dependency betweenuser needs and information use context. In our research, the context of use is integrated into informationrepresentation that can help refine user’s information needs. Thirdly, we propose an approach that allows a



user to carry out information retrieval and at the same time participate in indexing multimedia documents inthe database through annotation process since the semantic aspect of multimedia information varies from oneuser to another. The resulting thesis4 was defending in february 2010.

6.4. Knowledge capitalizationParticipants: Fausat Oladejo, Amos David.

The goal of this study is to enable the preservation of knowledge in the various phases of the EI process byacquiring, organizing and integrating such knowledge in order to exploit it in terms of re-use and distribution ofknowledge. Our approach is to harmonize EI models (knowledge resources) for capitalization and optimizationof decision-making process. The result of this study will contribute to the development of the knowledge baseas presented in figure 1. Finally, we have developed an application package for EI Knowledge Capitalization(EIKC) System [18]. This prototype system allows actors to log in based on their roles. In the course of solvinga decision problem, the system tracks and stores the activities of each EI actor as well as the interaction amongthem including the initiation of a decision problem (DP), concessions among actors, definition of the decisionalproblem, information research specification, its reformulation, etc. Other activities for which knowledge isgenerated include the resolution of information search problem such as identification of relevant informationsources, collection of relevant information from databases and open sources, analysis of collected informationfor producing indicators for the resolution of the DP. EIKC can produce the following types of knowledgeamong others:

• Details of each past DP : the initiator, who participated in its specification, the contributions of eachparticipant, temporal evolution of the specification of a DP.

• Yearly distribution of DP• The actors involved in past DP : cross analysis of DP and actors, cross analysis of actors and roles.• Information (solution) retained for the past DP : details of retained solution with relevance justifica-

tion, cross analysis of solutions and the DP.• Sources used for resolving past DP : why the choice ?, who identified the source ?, cross analysis of

sources and DP

About application areas, the system can be used in any socio-economic organization such as educational,economic, politics, banking, production industries, etc.

6.5. Knowledge and strategy elicitationParticipant: Stéphane Goria.

To resolve the information retrieval and innovation problems [12], we created a elicitation knowledgemethodology with a tool. The second part of this work is in representing the activities of the person chargedwith collecting the information to use in resolving a decision-problem. A complementary approach is basedon social and technical analysis of user needs and context in an organisation. The main purpose is torepresent the different actors based on the artefacts in the project. The main issue here is on a particularinformational problem : organisation strategy identification and explanation applied to territorial intelligenceconcept, creativity methods (particulary TRIZ and Synectics), war-rooms and wargame practices, informationvisualization and the strategy board game metaphor [14]. Its principal objective is to develop a visualizationtool to map a situation or a strategy. At this moment the Materalia lorrain competitiveness cluster is interestedin the development of this project.

6.6. Risk management in Economic Intelligence6.6.1. Information risk management

Participants: Olufade Onifade, Gérald Duffing, Odile Thiéry.

4H. Maghrebi, "La représentation des informations multimedias à partir des besoins informationnels des utilisateurs : approchesd’intelligence économique", Thèse université Nancy 2, 2010.

Team SITE 13


To Information Risk Management in EI context, we are presently hybridizing principle of Ontology and Fuzzyin what we call FuzzOntolgy [13]. The rationale is from the Ontological framework to facilitate appropriateuse of linguistic variables in declaring the membership functions. The development is in progress and isoriented with a Cognitive-based Risk Factor Model [27] for identification of possible occurrence of risksto monitoring. We focus on the risk in decision process when the data in data warehouse are not consistant.Then, Knowledge is seen as legitimate and meaningful resources that strengthens the overall managementperformance. As a result, knowledge management is viewed as a sine qua non towards creation, storage,sharing, and reusing of the organization’s knowledge, employing advances in today’s technology. EconomicIntelligence (EI) is saddled with usage of timely availability of information towards strategic decision making.However, while decision making is based on available information, it has been observed with concern thatreconciling the "need for decision" and subsequent "search for relevant information" poses a serious threat tothe overall decision because of some intangible factor that are difficult to be expressed culminating into non-quality of retrieved data, and sometime time taken to adequately mapped the decision maker’s "mind-set" intoan appropriate object for information retrieval. Ontology potentially enable automated knowledge sharing andreuse among both human and computer agents; this is facilitated based on their ability to interweave humanand machine understanding through formal and real-world semantics. Our framework attempts to captureoperational complexities and human issues into a singular unit to present the global view of inherent risk inknowledge reconciliation that could result into non-quality of data/information and consequently affect thedecision maker adversely in the discharge of his duties. The Models developed for the research include:

• Knowledge Reconciliation and Ontological Framework tagged (KNOREM). This model attemptsto determine the level of understanding that results from the interaction between some set of actors(decision maker and watcher) towards the need to deliver strategic decision.

• Decisionability succinctly describes the problems associated with increase volume of informationon the ability of a decision maker to adequately perform his roles otherwise referred to as cognitiveoverload.

• Fuzzontology is a compound word derived from Fuzzy Inference System (FIS) and Ontology. Thisfollowed from the result of KNOREM taking into cognizance the opinion of Langefors, (1966)dealing with "interpretation". The model thus interprets what ontology captures with fuzzy inferencesystem.

• FUZZYWATCH is the last of the proposition, it combines the tolerance in FIS to address the issueof missing data that could result from missing data (not wrong, but ambiguous), more importantmissed/spelling errors.

6.6.2. Risk among decision actorsParticipants: Olufade Onifade, Odile Thiéry.

Subjective estimation and perception, complexity of the environment under study, interaction amongst sub-systems,lack of precise data, missing data, limited information processing capacity and ambiguity in naturallanguages are major forms of uncertainty facing Decision Makers in the process of delivering strategic deci-sions in economic intelligent systems. This study employs soft computing paradigm to capture and analyzeuncertainty based on information risk factors via our proposed knowledge reconciliation model based on on-tology and the FuzzyMatch model. We modeled the process of decision making from the point of problemdefinition to decision delivery (translation credibility) and include intangible factors with the fish-bone archi-tecture. Ontological framework for Knowledge Reconciliation was developed to facilitate knowledge sharingand reuse among both human and computer agents while intangible factors, emotions and ambiguities in nat-ural languages were captured with fuzzy membership function. We extended this operation with fuzzy thatis what ontology captures is interpreted by fuzzy techniques (FuzzOntology) [13]. The fuzzy match relationfor information retrieval tagged FuzzyMatch improves the information search result thus reducing the risk ofmissing data which is of grave consequence in Economic Intelligence process [28]. FuzzOntological modelfacilitates a flexible means of capturing intangible and uncertain factors as a means of resolving the ambiguityin natural languages. FuzzyMatch assists in reducing missing data problems. Future decisional process willcontend with lesser information retrieval risks in Economic Intelligence process using this model.



6.7. Economic Intelligence organizations and systems6.7.1. The coordinator: the manager of EI process

Participant: Audrey Knauf.

The Economic Intelligence Devices (EID) recently introduced in France, is concerned with seven identifiedactions of EI (watch, KM, animation, training, councelling, security, influence). We took part in the estab-lishment of one of the devices, in which infomediaries are responsible for promoting and coordinating theactions of EI to the profit of decision-makers. Within the framework of her PhD studies sought to identifycompetencies necessary for a better piloting of the EID. The study led to the proposal of a guide (CEID:Specifications of competences of the Coordinator in a EID) allowing to take into account the necessary rolesand competencies of the coordinator according to the actions of EI implemented in the EID [24]. This toolmakes it possible to generate a form of questionnaire which can be used to prepare role specification in orderto recruit a coordinator. CEID allows the definition of a new profession and thus its integration as elements of atraining curriculum, such as that proposed by Alain Juillet, Senior Director in charge of Economic Intelligencein France. Earlier results from this work has been tested at the regional level such as in the project DECILOR.

6.7.2. Territorial intelligence: practices and conceptionsParticipants: Audrey Knauf, Stéphane Goria.

Territorial intelligence is a new concept focusing on the concept of "territory". Generally territorial intelligenceis understood with a geographic referent. With this interpretation, territorial intelligence can be interpreted asan economic intelligence practice. We study particulary development of company clusters and competitivenessclusters. We focus on their needs, governance, practices and tools of these clusters to help their members [23].Another interpretation of territorial intelligence is to consider it as an intellectual construction which has asense only in relation with others territories. This year, we focused on territory as a Market [11].

7. Contracts and Grants with Industry

7.1. Institutional partnershipsThe research team is in partnership with the following structures:

• GDR-IE: a research group on Economic Intelligence created by French National Center for ScientificResearch (CNRS).

• The French senior director in charge of Economic Intelligence: Some of the proposals of the researchteam are exploited in some work groups working with the senior director. The master studies inEconomic Intelligence, Nancy 2, has also received a formal support from the director.

• The INIST-CNRS: We are working on a research agreement between the INIST-CNRS and Univer-sité Nancy 2 in order to officialize a collaboration between the research teams at INIST-CNRS andour team. Two research subjects at masters level were jointly defined by our research team and twounits at INIST-CNRS in 2005 as well as in 2006.

• The CRCIL5: We are collaborating with the CRCIL for technology transfer by publication of theresults of research in the field of EI, by training in form of central internship offer managementthereby putting our students in contact with the companies. We are also studying other means ofcollaboration. For example The CRCIL has made a formal request to use two models defined by ourresearch team within companies6.

5CRCIL: Chambre Régionale de Commerce et d’Industrie de Lorraine6DMP (Decision-Maker Problem) model and WISP (Watcher’s Information Search Problem)

Team SITE 15


• The Ministry for Foreign Affairs: We develop our international relations with the French Ministryfor Foreign Affairs. Indeed, by collaboration with the ministry, we were able to organize twoconferences in Nigeria and obtained 6 PhD study grants since 2003. This last collaboration withthe French Ministry for Foreign Affairs is evolving towards the proposal of a joint masters degree inEconomic Intelligence and in Computer Science that will involve Université Nancy 2 on EconomicIntelligence and 4 Nigerian Universities (University of Ibadan, Obafemi Awolowo University, LagosState University and University of Lagos) on Information Systems.

7.2. Contracts, research agreements and industrial actionsWe took part in three European projects with industrial partners in the field of research. Similarly, withinthe framework of Master ISTIE (a professional course on Scientific and Technical Information - EconomicIntelligence), technology transfer projects are carried out.

We have had four industrial contracts with companies relating to the applications of our proposals for thedevelopment of strategic information systems used in the context of EI:

• CIFRE contract with COM-MEDIC, 2006-2009.

• CIFRE contract with NANCIE: 2003-2006.

• CIFRE contract with CEIS for the lorrain project DECILOR: 2002-2005.

• Research agreement with the Materalia competitiveness cluster, France, signed in 2009.

• Research agreement with the Company HAYET, Tunisia, 2002-2006.

• Research agreements with CVCE (Centre Virtuel de la Connaissance sur l’Europe), Luxembourg,signed in 2006.

7.3. Visit, InvitationThe research team received two visiting lectures in 2010:

• Sawyerr Babatunde, Computer Science department, University of Lagos Nigeria.

• Adeyemo Omowunmi, Computer Science department, University of Ibadan, Nigeria.

8. Dissemination

8.1. Animation of scientific communityMembers of the team are members of several scientific organizations. They take part or are responsible for theorganization of several scientific conferences.

• Amos David was nominated as member of the National University Commission (Conseil Nationaldes Universités - CNU) section 71 of Information and Communication Sciences.

• The research team organized ISKO 2005 conference on "Knowledge Organisation in Utilisationoriented Information Systems: Watch and Economic Intelligence Context", April 2005, Nancy.

• We take charge of the teaching and evaluation of a master course at the University of Ibadan,Nigeria, since 2004 on "Advanced Data modeling and development". The program involves threeother universities/institutions in Nigeria.

• Amos David is a member of the scientific committee of ISKO (International Society for KnowledgeOrganization), french chapter.

• Odile Thiéry is a member of AIM (Association of Information and Management) and ADELI(Association for information system mastery).



• She was some years back a regular member of the program committee of the congress INFORSIDwhich comprises a session on SIS.

• The research group GDR-IE to which our research team belong, has been involved in the productionof a book and two special journals on economic intelligence.

• We have collaboration with the University of Malaga, Spain, on user modeling in the context ofinformation retrieval.

• Audrey Knauf and Stéphane Goria are a members of SFSIC (French Society of Commucation andInformation Sciences).

• Audrey Knauf and Stéphane Goria are a members of RRI (French Research Network on Innovation).

8.2. TeachingWe participate in teaching at second and third cycle in the university. The modules we teach relate primarily tothe fields developed in our team: economic intelligence process, data modeling and design of traditional andstrategic information systems. We intervene in the following degree programmes:

• Master’s Degree– Domain: Information and Communication sciences, Economic Intelligence– Domain: MIAGE (Computer Science methods applied to Management)– DRT SIO

• Bachelor’s Degree– ICN (School of Management) Organizational Information System option– Elective courses in Documentation

8.3. Administrative loads• Amos DAVID is the coordinator of an agreement between University Nancy 2 and University of

Ibadan Nigeria. He is director of studies for Master’s in STI-EI. Within the framework of thisagreement, a session of summer school is organized each year. And for this specific action of summerschool, we drew up a tripartite agreement (Nancy 2, University of Ibadan, and the French Ministryfor Foreign Affairs), over 3 years (2004-2011). This programme is under evolution to arrive at a jointMaster’s degree in STI-EI and subsequently in Computer Science.

• Odile THIERY is responsible for DRT SIO. She has been reelected as member of Nancy 2University’s administrative council. She is the director of the Campus Manufacture s General Servicesince 2005. Since 4 years ago she has been in charge of "patrimoine immobilier" under the Presidentof Université Nancy 2.

• Gérald DUFFING is the head of department ’Management Information Systems’ at ICN School ofManagement. He is also in charge of the ICN Master’s Degree.

• Audrey KNAUF is responsable for M1 of Master’s in STI-EI (University Nancy 2).• Stéphane GORIA is responsible of the Preparation to Professional Project for the computer sciences

students at Nancy Charlemagne Technology University Institute.

9. BibliographyMajor publications by the team in recent years

[1] B. AFOLABI, O. THIÉRY. Système d’intelligence économique et paramètres sur l’utilisateur : application à unentrepôt de publications, TIC et Territoires, in "Informations, Savoirs, Décisions, Médiations", 2005, no 22,p. 1-15, http://hal.inria.fr/inria-00001072/en/.

Team SITE 17


[2] N. BOUAKA, A. DAVID. Modèle pour l’explicitation d’un problème décisionnel: un outil d’aide à la décisiondans un contexte d’intelligence économique, in "International Journal of Infomation and CommunicationSciences for Decision Making", 2003, no 11, p. 12-18, http://hal.inria.fr/inria-00099692/en/.

[3] A. DAVID. , P. U. DE NANCY (editor)Organisation des connaissances dans les systèmes d’informationsorientés utilisation : contexte de veille et d’intelligence économique, Presses Universitaires de Nancy, 2005.

[4] A. DAVID. , L. C. DU NUMÉRIQUE (editor)Intelligence économique, Information, Evaluation, Stratégies,Hermes - Lavoisier, 2010, vol. 5, http://hal.inria.fr/inria-00465681/en.

[5] A. DAVID, O. THIÉRY. L’architecture EQuA2te et son application à l’Intelligence Economique, in "ISDM :Informations, Savoirs, Décisions et Médiations", 2003, no 11, p. 1-8, http://hal.inria.fr/inria-00099690/en/.

[6] P. KISLIN, A. DAVID. De la caractérisation de l’espace-problème décisionnel à l’élaboration des éléments desolution en recherche d’information dans un contexte d’Intelligence Economique : le modèle WISP, in "ISDM(Informations, Savoirs, Décisions et Médiations)", 2003, no 11, p. 1-10, http://hal.inria.fr/inria-00099701/en/.

[7] A. KNAUF. Les dispositifs d’intelligence économique : compétences et fonctions utiles à leur pilotage,Intelligence économique, L’Harmattan, May 2010, http://hal.inria.fr/inria-00545013/en.

[8] A. KNAUF, S. GORIA. Spécification des métiers et compétences impliqués dans le dispositif régionald’intelligence économique, in "Intelligence Territoriale : L’intelligence économique appliquée au territoire",L. FRANÇOIS (editor), TEC & DOC, Lavoisier, 2008, p. 71-86, http://halshs.archives-ouvertes.fr/halshs-00359173/en/.

[9] H. MAGHREBI, A. DAVID. Exploiting Context awareness and annotation process to support multimedia in-formation retrieval, in "Content Architecture: Exploiting and Managing Diverse Resources - The first bien-nial Conference of the British Chapter of the International Society for Knowledge Organization - ISKO UK",Royaume-Uni London, June 2009, 9 pages, http://www.iskouk.org/conf2009/papers/maghrebi_ISKOUK2009.pdf, http://hal.inria.fr/inria-00433830/en.

[10] O. THIÉRY, A. DAVID. Modélisation de l’utilisateur, Systèmes d’Informations Stratégiques et IntelligenceEconomique, in "Revue Association pour le Développement du Logiciel (ADELI)", 2002, no 47, http://hal.inria.fr/inria-00100742/en/.

Year PublicationsArticles in International Peer-Reviewed Journal

[11] S. GORIA. Intelligence économique, Intelligence Territoriale et cabinets de conseil, in "Revue Internationaled’Intelligence Economique", October 2010, vol. 2, no 1, p. 99-116, http://hal.inria.fr/halshs-00526781/en.

[12] S. GORIA. Proposition d’une méthode d’expression d’idées et de problèmes d’innovation, in "ESSACHESS(Journal of Social and Cultural Studies)", December 2010, no 5, p. 5-20, http://hal.inria.fr/halshs-00550000/en.

[13] O. THIÉRY, G. DUFFING, A. OSOFISAN, W. ONIFADE. Fuzzontology: Resolving Information Mining,Ambiguity in Economic Intelligent Process, in "Lecture notes in computer science", 2010, p. pp. 232–243,http://hal.inria.fr/inria-00545656/en.



International Peer-Reviewed Conference/Proceedings

[14] S. GORIA. Mise en évidence d’informations stratégiques à partir de l’analogie du jeu de plateau, in "Colloqueinternational Veille Stratégique Scientifique & Technologique - VSST 2010", France Toulouse, B. DOUSSET(editor), October 2010, vol. 1, p. 1-16, http://hal.inria.fr/hal-00532584/en.

[15] V. ODUMUYIWA, A. DAVID. Collaborative knowledge creation and management in information retrieval,in "Fifth International Conference on Knowledge Management in Organizations - KMO 2010", Hongrieveszprém, May 2010, http://hal.inria.fr/inria-00546802/en.

[16] O. OKUNOYE, A. DAVID, C. UWADIA. AMTEA: Tool for Creating and Exploiting Annotations in the Contextof Economic Intelligence (Competitive Intelligence), in "11th IEEE International Conference on InformationReuse and Integration - IRI 2010", United States Las Vegas, IEEE, August 2010, p. 249–252, ISBN : 978-1-4244-8097-5 [DOI : 10.1109/IRI.2010.5558933], http://hal.inria.fr/inria-00550115/en.

[17] O. OKUNOYE, B. OLADEJO, V. ODUMUYIWA. Dynamic Knowledge Capitalization through Annotationamong Economic Intelligence Actors in a Collaborative Environment, in "Colloque International VeilleStratégique Scientifique et Technologique - VSST 2010", France Toulouse, October 2010, p. 1-17, http://hal.inria.fr/inria-00546809/en.

[18] B. OLADEJO, V. ODUMUYIWA, A. DAVID. Dynamic Capitalization and Visualization Strategy in Collabora-tive Knowledge Management System for EI Process, in "International Conference in Knowledge Managementand Knowledge Economy - ICKMKE 2010", France paris, June 2010, p. 1-9, http://hal.inria.fr/inria-00546808/en.

[19] W. ONIFADE, O. THIÉRY, A. OSOFISAN, G. DUFFING. A Fuzzy Model for Improving Relevance Ranking inInformation Retrieval Process, in "International Conference on Artificial Intelligence and Pattern Recognition- AIPR-10", états-Unis Orlando, Z. MAJKIC, D. TAMIR, G. WANG (editors), ISRST, July 2010, p. 189-195,ISBN : 978-1-60651-015-5, http://hal.inria.fr/inria-00545654/en.

[20] W. ONIFADE, O. THIÉRY, A. OSOFISAN, G. DUFFING. Dynamic Fuzzy String-Matching Model for Informa-tion Retrieval Based on Incongruous User Queries, in "Proceedings of the World Congress on Engineering -WCE 2010", Royaume-Uni Londres, July 2010, ISBN : 978-988-17012-9-9, http://hal.inria.fr/inria-00545650/en.

Scientific Books (or Scientific Book chapters)

[21] A. DAVID. Intelligence économique, Information, Evaluation, Stratégies, Hermes - Lavoisier, 2010, vol. 5,http://hal.inria.fr/inria-00465681/en.

[22] A. KNAUF. Les dispositifs d’intelligence économique : compétences et fonctions utiles à leur pilotage,Intelligence économique, L’Harmattan, May 2010, http://hal.inria.fr/inria-00545013/en.

[23] A. KNAUF, S. GORIA. L’intelligence économique et la gestion des connaissances au sein des organisa-tions, in "Communication des organisations : recherches récentes", C. L. ET BERTRAND PARENT (editor),L’Harmattan, October 2010, p. 174-188, http://hal.inria.fr/halshs-00526790/en.

Team SITE 19


[24] A. KNAUF. Vers l’émergence de nouvelles compétences pour le pilotage de l’intelligence économique enrégion, in "Intelligence économique et problèmes décisionnels", A. DAVID (editor), Hermès - Lavoisier, June2010, http://hal.inria.fr/inria-00545016/en.

[25] H. MAGHREBI. La recherche et l’accès à l’information multimédia par et à travers un processus d’IntelligenceEconomique, in "Intelligence économique et problèmes décisionnels", A. DAVID (editor), Traité STI - sérieEnvironnements et services numériques d’information, Hermes Science, 2010, p. 135-157, http://hal.inria.fr/inria-00547939/en.

[26] V. ODUMUYIWA. La recherche collaborative d’informations dans le processus d’intelligence économique, in"Intelligence économique et problèmes décisionnels", A. DAVID (editor), Traité STI - série Environnementset services numériques d’information, Hermes, 2010, ISBN : 978-2-7462-2503-9, http://hal.inria.fr/inria-00546807/en.

[27] O. THIÉRY, W. ONIFADE, A. OSOFISAN, G. DUFFING. Etude des facteurs de risque pour la prise de décisionen intelligence économique : une approche cognitive, in "Intelligence économique, Information, Evaluation,Stratégies", A. DAVID (editor), Hermes-Lavoisier, June 2010, http://hal.inria.fr/inria-00548169/en.

Other Publications

[28] W. ONIFADE, O. OKUNOYE, A. DAVID. Embedded fuzzontological model for document interpretation inattribute-value-pair annotation in economic intelligent systems, May 2010, http://hal.inria.fr/hal-00545204/en.


site modeling and development of economic intelligence systems

Documents