Section: Other Grants and Activities
National Initiatives
AxIS is involved in several national working groups.
CNRS Action Concertée Incitative: ``Histoire des savoirs''
This initiative (ACI RNR TTT Grammaire et mathématique dans le monde indien 17/01/03 - 17/01/06) associates several French research teams from various research fields, such as computer science, data analysis, and Sanskrit literature. The main goal of this action is to provide help for the construction of critical edition of Indian manuscripts in Sanskrit, and to provide pertinent information about the manuscripts classification (construction of cladistic trees). The expected tools will not be restricted to Sanskrit language in every aspects. This action is complemented by the European AAT project support which allows us to collect more Sankrit manuscripts and to care about some interactive aspect that we where not able to take into account with the ACI dotation.
The end of action has been delayed until december 2007.
EGC Association: National Group on Mining Complex Data
AxIS members participated actively this year to the Working Group ``Fouille de données complexes'' created by D.A Zighed in June 2003 in the context of the EGC association:
-
B. Trousse with O. Boussaïd (ERIC, Lyon) co-organised and co-chaired the third workshop ``Fouille de données complexes dans un processus d'extraction de connaissances'' (January 17, 2006) [16] . M-A. Aufaure, F. Masseglia and Y. Lechevallier were members of the program committee.
-
F. Masséglia with O. Boussaïd co-animate one of the three topics: ``Organisation and Structuration of Complex Data''.
-
B. Trousse with S. Després (CRIP5 - Université de Paris V) co-animate the topic `` Knowledge in Complex Data Mining''.
SFDS association: InfoStat Group
SFDS is the French Society of Statistics : URL: http://www.sfds.asso.fr/ .
AxIS members participated actively this year to the workshops "Les après-midis d'InfoStat" of the InfoStat Group which is leaded by Y. Lechevallier (president):
-
January 12, Rennes: ``Visualisation statistique: du PDA à l'écran géant''.
-
March 23, Paris: "Les prévisions ont-elles un avenir ?".
-
October 24, Paris: "Classification avec des modèles de mélange: la nouvelle version 2.0 du logiciel libre MIXMOD".²
GDR-I3
AxIS participated to three working groups of the GDR-PRC~I3 National Research Group ``Information - Interaction - Intelligence'' of CNRS:
-
Working Group 3.4 (GT) on Data Mining animated by P. Poncelet and J.M. Petit: MA. Aufaure, F. Masseglia, B. Trousse
-
Gracq (Groupe de Recherche en Acquisition des Connaissances ) (GRACQ ): B. Trousse.
-
Working Group 3.7 ``Sécurité des Systèmes d'Information'' animated by D. Boulanger and A. Gabillon: F. Masseglia and B. Trousse.
Other Collaborations
-
LePont Laboratory of the South Toulon University (E. Boutin) and LIRMM 5(M. Nanard). Our work was on the pre-processing and clustering of a test collection of Web pages returned by different search engines in response to a specific set of queries (cf. section 6.5.6 .
-
LIP6: We work with Bénédicte Le Grand and Michel Soto in the Eiffel RNTL project on visualisation and navigation for enhancing semantic web retrieval in the tourism domain. Marie-Aude Aufaure works also with them on semantic and conceptual context-aware information retrieval [48] ..
-
ENST Paris: Y. Lechevallier collaborated with Georges Hébrail (ENST) [38] .
-
ENST Bretagne: In the framework of the ACI "Histoire des savoirs" we have a regular collaboration with some ENST B researchers, namely P. Bertrand, M Le Pouliquen, J-P Barthélémy on classification and comparison of the Sanskrit Manuscripts. With P. Bertrand (ENST B) and Annie Morin (TEXMEX, IRISA Rennes) we had a meeting on the possible use of the N-GRAMM for the sanskrit texts comparisons.
-
CNAM and Loria (Cortex Team): contacts have been established with research teams in human and social sciences.
-
University of Bordeaux 1 and 2 (MAP laboratory): Y. Lechevallier collaborated with M. Chavent [79] ,[32] .
-
Two ARC proposals were submitted: 1) ARC SéSur: ``Sécurité et Surveillance dans les data streams'' (resp: F. Masseglia) with M.O Cordier (DREAM, IRISA) P. Poncelet (LGI2P, Alès) and M. Teisseire (LIRMM, Montpellier); 2) ARC Valex: `` Vérification et exploitation de collections de documents semi-structurés'' (resp: A.-M. Vercoustre) with Annie Morin (TEXMEX, IRISA Rennes), A. Napoli (Orpailleur, INRIA, Nancy), E. de la Clergerie (ATOLL) and B. Sagot (Signes), INRIA Rocq.
-
ANR: We have proposed a project in response to the ANR call for proposals on ``Masses de Données / Connaissances Ambiantes''. The ``MIDAS'' (MIning DAta Streams ) project gathered together 6 academic partners: INRIA's team-projects AxIS and DREAM, ENST (Paris), LIRMM (Montpellier), LGI2P (Nimes), GRIMAAG (Martinique); and two industrial partners: EDF R&D and France Télécom R&D. Despite this project has not been accepted, we plan to work on further proposal on this subject.
-
GRIMM-SMASH team (Université Toulouse Le Mirail): F. Rossi works with N. Villa on Support Vector Machines and functional data (cf section 6.3.3 and [62] , [27] ).
-
LITA EA3097 (Université de Metz): F. Rossi works with Brieuc Conan-Guez on the Self Organizing Map for dissimilarity matrices (see section 6.3.2 ) [20] , [22] , [34] and others topics [24] .
-
A. Michard (Inria Rocquencourt): we hold different meetings with A. Michard to discuss a service based on modeling the dependency of Information and Organization in order to improve the quality of Entreprise Information Systems. We also discussed the possible application of data mining in anticipating possible breakdown of the system. Such a model was submitted to Bouygues Telecom and France Telecom in the prospect of a future R&D project.