Project : axis
Section: Other Grants and Activities
IST European Network : Ontoweb
AxIS participated in the European network Ontoweb (Ontology-based Information Exchange for Multilingual Electronic Commerce and Information Integration) proposed in 2000 by Dieter Fensel (Division of Mathematics & Computer Science, Vrije Universiteit Amsterdam).
The project ended in May 2004 with an excellent evaluation.
IST European Project: ASSO
ASSO was a project of the European Union Fifth framework Research and Development program in the Information Society Technology strand, number IST-2000-25161. ASSO offered methods, methodology and software tools for the analysis of multidimensional complex data (numerical or non numerical) coming from databases in statistical offices and administrations using Symbolic Data Analysis. ASSO ended this year.
ASSO was based around nine working groups ("Workpackages") with the participation of fifteen partners ("Consortium").
AxIS contributions to program
Benchmark and prototype evaluation The first part in this task consists in the benchmark definition: choice of data, variables, population, encoding, weights, rules and taxonomies with a special interest to data on unemployment, social insertion as well as business registers.
Data Format SOM library Every algorithm within the project uses the SOM library, which interfaces the different algorithms with the data. They will take into account the metadata, introduce a new representation in order to provide to the different algorithms a way to use the knowledge provide by the rules.
Metadata Model Design included in SOM The mission of this task is to design a semantically rich metadata model that will hold additional information for the assertions that construct the SO. Such a model should be carefully designed in order to ensure that the data consumers needs are not underestimated (missing information) or that data providers are not forced to allocate resources in capturing metainformation of little value (overestimation).
DB2SO, which creates SO from a database, has a first version in the previous project but many improvements have to be done. In the new version every groups of individuals is not generalized by one Symbolic Description but by many disjunctive Symbolic Descriptions.
DIV, SCLUST Partitional clustering In this task we propose clustering algorithms in order to partition a set of SO into a predefined number of classes. The classes are interpreted and represented by suitably generalized class prototypes (again in the form of symbolic objects).
Interpretation of partition, robustness of methods and cluster validation included in SCLUSTER We select a cluster and interpret it for a selected set of variables. We will measure the robustness of a cluster and a symbolic object related to extent, intent dissimilarity and algorithm. We will measure the degree of isolation and compactness of each cluster and its symbolic intent.
COST Action 282
In 2004, we participed actively in the COST Action 282 (2001-2005): ``Knowledge Exploration in Science and Technology''.
Organisation of the KELSI 2004 workshop in Nice in January 2004 : S. Honnorat and B. Trousse
Short Term Cost missions to the Bioinformatics Research Group at the University of Ulster (Nth. Ireland) July 2004: D. Tanasa and B. Trousse
Writing a common paper with Jesus Lopez accepted to the KELSI04 Symposium (Milano) .
Participation in two Management Committees (January in Nice, November in Milan): B. Trousse is one of the two French representatives in the Management Committee.
EuropAid project: Sanskrit
An Asia-Information Technology and Communications project (EuropeAid) has been submitted and accepted by the EEC. This project aims to design a most advanced IT tool for an "archaeology of ancient Asian texts" (mainly written in Sankrit). It will start in 2005 for one year.
Consortium members of this projet are: 1) Inria, AxIS team 2) Bhandrakar Institute, India, 3) University Sapienza di Roma, Italy and 4) Mahendra Sanskrit University, Nepal.
Italy, University of Napoli II (Profs C. Lauro and R. Verde)  : Y. Lechevallier, A. El Golli, M.Csernel
Belgium, Facultés Universitaires Notre-Dame de la Paix à Namur (Profs A. Hardy, M. Noirhomme and J.-P. Rasson) : Y. Lechevallier.
Belgium, Université Catholique de Louvain-la-Neuve, DICE Laboratory ( N Delannay, M Verleysen)  : B. Conan-Guez, F. Rossi