IUT de Villetaneuse

This work reports on the applicability of the Broadway recommendation computation ap-proach for implementing a query renement (QR) service in the context of a distributed infor-mation retrieval and gathering system. The Broadway recommendation approach is ...

This work copes with the problem of link prediction in large-scale two-mode social networks. Two variations of the link prediction tasks are studied: predicting links in a bipartite graph and predicting links in a unimodal graph obtained by the projection of a bipartite graph over one of its node sets. For both tasks, we show in an empirical way, that taking into account the bipartite nature of the graph can enhance substantially the performances of prediction models we learn. This is achieved by introducing new variations of topological atttributes to measure the likelihood of two nodes to be connected. Our approach, for both tasks, consists in expressing the link prediction problem as a two class discrimination problem. Classical supervised machine learning approaches can then be applied in order to learn prediction models. Experimental validation of the proposed approach is carried out on two real data sets: a co-authoring network extracted from the DBLP bibliographical database and bipartite graph 8-years history of transactions on an on-line music e-commerce site.

This paper describes the design choices and the prototype implementation of CoopScan, a generic framework for synchronous groupware development. We focus on architectural issues and on strategies for integration of existing single−user applications into a collaborative environment. In this work, we propose a generic approach to application re−use. This approach is validated through the development of a testbed synchronous collaborative editor.

This paper describes the design of a peer-to-peer system for collaborative management of distributed bibliographical databases. The goal of this system is twofold: firstly, it aims at providing help for users to manage their local bibliographical databases. Secondly, it offers the possibility to exchange bibliographical data among like-minded user groups in an implicit and intelligent manner. Each user is assisted by a personal agent that provides help such as: filling in bibliographical records, verifying the correctness of information entered and more importantly, recommendation of relevant bibliographical references. To do this, the personal agent needs to collaborate with its peers in order to get relevant recommendations. Each agent applies a case-based reasoning approach in order to provide peers with requested recommendations. The paper focuses mainly on describing the recommendation computation approach.

Replicated data consistency is a key issue in the design of distributed real time groupware applications. In this paper, we propose a new protocol to cope with this problem. The proposed algorithm guarantees an optimal response time while ensuring data consistency at system quiescence. The originality of our proposition relies on the fact that neither locks nor clocks nor global information are required to establish data consistency. Instead, direct dependency relations between generated operations as well as operation transformation mechanism are used. The coupling of the above two mentioned mechanisms is shown to realize a good trade-off between the different requirements of groupware applications. Advantages of our approach are illustrated by comparing the algorithm to two well known optimistic concurrency control protocols for groupware applications: dOPT and ORESTE.

In this paper we describe a multi-agent system, called COWING that aims at enabling an organised group of people to share their results of information searching on the World Wide Web. Users normally save relevant sites they found in their private collections of bookmarks. The goal of the proposed system is to enable users to share their bookmarks, in an implicit, secure and effective way. By implicit we mean that users are not required to spend extra effort in order to use the system. Secure sharing refers to the capacity of each user to control who knows what about her/his own bookmark collections. Finally effectiveness is ensured by recommending users with relevant bookmarks that are computed by applying a distributed collaborative filtering algorithm. Recommended bookmarks are automatically inserted in the most appropriate bookmark folder in the user's local collection. When accessing a folder the user can evaluate (accept or reject) provided recommendations. Our system is a multi-agent system where each user is associated to an assistant agent, called a WING agent. A WING agent observes the user's behaviour when managing her/his own bookmark collection. It learns how the user classifies her/his bookmarks. A hybrid neural/case-based reasoning supervised classifier is used for this purpose. WING agents exchange bookmarks according to a defined collaboration protocol that protects associated users privacy.

In this paper, we describe a cooperative P2P bibliographical data management and recommendation system (COBRAS). In CO-BRAS, each user is assisted by a personal software agent that helps her/him to manage bibliographical data and to recommend new bibliographical references that are known by peer agents. Key problems are:

In this work we propose a new efficient algorithm for communities construction based on the idea that a community is animated by a set of leaders that are followed by a set of nodes. A node can follow different leaders animating different communities. The algorithm is structured into two mains steps: identifying nodes in the network playing the role of leaders, then assigning other nodes to some leaders. We provide a general framework for implementing such an approach. First experimental results obtained by applying the algorithm on different real networks show the effectiveness of the proposed approach.

Bookmark systems are now recognised as a practical mean that allow users to land mark interesting sites in the huge information space that constitute the World Wide Web. In this paper we argue that, by building collaborative bookmark management tools where a group of people can ...

This paper presents a new approach in the management of mobile ad hoc networks. Our alternative, based on mobile agent technology, allows the design of mobile centralized server in ad hoc network, where it is not obvious to think about a centralized management, due to the absence of any administration or fixed infrastructure in these networks. The aim of this centralized approach is to provide permanent availability of services in ad hoc networks which are characterized by a distributed management. In order to evaluate the performance of the proposed approach, we apply it to solve the problem of mobile code localization in ad hoc networks. A comparative study, based upon a simulation, of centralized and distributed localization protocols in terms of messages number exchanged and response time shows that the centralized approach in a distributed form is more interesting than a totally centralized approach.

Après une introduction à la recherche d'informations sur le Web et une classification des assistants d'aide à la navigation sur le Web, cet article décrit une nouvelle approche pour l'aide à la navigation sur le Web centrée fouille de données nommé Broadway. L'approche Broadway repose sur l'utilisation des techniques de raisonnement à partir de cas (RàPC) afin de recommander à un utilisateur des pages qui ont satisfait des utilisateurs ayant navigué d'une manière similaire. Un exemple de l'implantation de l'approche Broadway est décrit. Enfin, avant de conclure, une comparaison avec d'autres approches est donnée.

Bookmark systems are now recognised as a practical mean that allow users to land mark interesting sites in the huge information space that constitute the World Wide Web. In this paper we argue that, by building collaborative bookmark management tools where a group of people can share their experiences and their results in feeding and organising personal bookmark sets, bookmarks

Groupware applications introduce new requirements in terms of access control that are not adequately handled by access control models devel- oped for non collaborative applications. The central presence of humans in the context of collaborative applications adds additional constraints on the access model and data protection mechanism to be used. In this paper, we propose a hierarchy-based general data access

In this paper we present an original approach for community detection in complex networks. The approach belongs to the family of seed-centric algorithms. However, instead of expanding communities around selected seeds as most of existing approaches do, we explore here applying an ensemble clustering approach to different network partitions derived from ego-centered communities computed for each selected seed. Ego-centered communities are themselves computed applying a recently proposed ensemble ranking based approach that allow to efficiently combine various local modularities used to guide a greedy optimisation process. Results of first experiments on real world networks for which a ground truth decomposition into communities are known, argue for the validity of our approach.

Nous nous intéressons dans ce travail à la problématique du développement des applications pour le travail coopératif, dites aussi collecticiels. Un collecticiel est à la fois une application multi-utilisateurs, répartie et interactive. La somme des trois propriétés précédentes rend le développement de ce type d'applications particulièrement difficile. Une approche souvent empruntée pour la construction des collecticiels consiste à développer une plate-forme qui fournit les services requis pour la coopération. Dans ce rapport nous identifions les principaux services demandés à une telle plate-forme et nous décrivons et nous comparons les différentes approches possibles pour réaliser ces services. Nous proposons ensuite un nouvel environnement de coopération appelé Colt (pour Collaboration Terrain). Une première qualité de Colt est l'intégration des deux modes de travail : individuel et en groupe. Les utilisateurs partagent un espace d'information où chacun a le droit de ...

We define an adaptive Web site as a site that can automatically change itself in order to optimise its effectiveness viewed from one or more site's actor point of view. The aim of this paper is two folds: firstly we propose a new classification scheme for Web site adaptation approaches. Secondly, we propose a new approach for web site adaptation, called the OBRA approach. This approach is based on reusing past experiences of navigation of a group of users in order to propose shortcuts in the same site. Case-based reasoning methodology is used to implement our approach.

This paper provides a state of art survey of works using social network analysis (SNA) techniques for improving performances of an on-line help desk. The processing of a client request Qc can be achieved following oe ofg the following scenarios: 1) sending Qc to the help desk. 2) sending QC to other clients that may have an appropriate answer., and 3) Navigating in FAQ pages in order to find an answer to Qc.

Pollutant transport by shallow water flows on nonflat topography is presented and numerically solved using a finite volume scheme. The method uses unstructured meshes, incorporates upwinded numerical fluxes and slope limiters to provide sharp resolution of steep bathymetric gradients that may form in the approximate solution. The scheme is non-oscillatory and possesses conservation property that conserves the pollutant mass during the transport process. Numerical results are presented for three test examples which demonstrate the accuracy and robustness of the scheme and its applicability in predicting pollutant transport by shallow water flows. In this paper, we also apply the developed scheme for a pollutant transport in the strait of Gibraltar. The scheme is efficient, robust and may be used for practical pollutant transport phenomena.

In this paper, hydrogeological and geophysical data are used to validate a numerical model developed to predict seawater intrusion into coastal aquifers. The cell-centered finite volume method is adopted here to solve the set of coupled partial differential equations describing the motion of saltwater and freshwater separated by a sharp interface. These equations are based on the Dupuit approximation and are obtained from integration of 3D flow equations for fresh and salt water zones over the vertical dimension. In order to have flexibility upon complex configurations domain, non structured grid meshing is utilized. To approximate the diffusion fluxes, Green-Gauss type reconstruction, based on diamond-cell and least squares interpolation, is performed. The model is first validated using academic test case studies with known closed form solutions. The mathematical model has been calibrated using hydrogeochemical and geophysical data. The geophysical method applied in this study has been a frequency domain electromagnetic method. In this method the apparent electrical conductivity is measured by induction using two separate hand-held transmitter and receiver coils. During the operation the transmitter coil is energized by a low frequency alternating current that radiates an electromagnetic field and the receiver coil detects the resulting field. Taking into account the relationship between the bulk conductivity of the subsoil and the conductivity of groundwater, EM soundings have been interpreted to provide complementary information to hydrogeochemical data to outline the fresh–saltwater interface. This methodology has been applied to the case of saltwater intrusion into the Llobregat delta aquifer, near Barcelona, Spain.