Sites Inria

English version

Equipe de recherche EDELWEISS

Rapports d'activité


We would like to pay a tribute to our dear friend and colleague Rose Dieng-Kuntz who was the founder and leader of the Acacia and Edelweiss teams.

Context and Objectives

Actors and interaction devices are becoming more and more mobile while knowledge sources, services and their networks are becoming ubiquitous. In this context we witness the emergence of communities of interest and/or practice, very light and agile structures that can be ephemeral and virtual. To assist the knowledge life-cycle of such communities we are interested in providing tools and methodologies supporting the interactions and the memories of these focused groups. Throughout its life time, a community uses, produces, exchanges, and shares resources materializing knowledge through various types of documents (that may be structured or not, textual, multimedia, etc.). A community may also rely on some services or programs available inside the community or outside. To ensure mutual understanding between community members, the exchanges inside a community rely on a common terminology and common concepts that may evolve throughout the life of the community. These exchanges can also use various media.

The context of the emergence of such virtual communities (inside organizations, across organizations or independently of any organization) is the use of the Web not only for information sharing but also for support to cooperation, the use of new interaction channels, the evolution of Web technologies (Semantic Web, social Web, Web services, mobile Web, ubiquitous Web).

Edelweiss aims at offering models, methods and techniques for supporting ergonomic, web-based knowledge management and collaboration in virtual communities interacting with information resources through the Web. We perform research on graph-based, ontology-based, web-based knowledge representation and inferences for interacting with or through information resources.

Research Topics

The support to virtual communities can be studied according to several viewpoints:

The activities of the community consist of structuring, searching, retrieving, reusing, and composing the community internal or external resources / services. A support to these activities can be offered through a Semantic Web based approach, by processing annotations of such resources / services.

Conceptual modeling of the interactions and collaboration among community members mediated by tools could enable us to propose ergonomic tools adapted to support such collaboration.

To achieve the development of such supporting tools and methodologies, basic blocks are needed to represent knowledge and to reason and perform inferences on this representation: we choose to rely on a graph-based representation.

Therefore, we study thoroughly two complementary research directions, corresponding to these viewpoints:

Interaction Design of Semantic Systems: Supporting human interoperability in semantic activities through articulating functionalities and in scenario management activities, Experimental evaluation of inferences for information retrieval and other tasks, Ontology-based intelligent interfaces.

Graph-based Knowledge Representation of the Semantic Web Knowledge: Scaling graph representations and operations, Ontology-based model driven engineering, Inferences characteristic to graphs and distributed Web sources, semantic annotation of information resources.

International and industrial relations

We collaborate or collaborated with industry in the following fields: telecommunications (Orange Labs, Telecom Valley), earth sciences and environment (Ademe, BRGM, IFP), biology (IPMC, Immunosearch), semi-conductors (Philips Semi-Conductors-NXP), manufacturing (Estanda) and car industry (ItalDesign, Renault). In the past, we took part in the Integrated Project Palette, in the STREPS projects SeaLife and SevenPro and in the Knowledge Web Network of Excellence.

Suivez Inria