Philippe PICOUETMaître de conférences / Assistant Professor
|
En français !
In english |
Main activities :
Based on
previous experiences in the database area (PhD thesis in 1995) and cognitive
modelling of expertise (COMAPS European Project in 1998), my current interest fields
are clearly having roots in both interest domains:
Aided design of
bio-informatics scenario
In
cooperation with OUEST-genopole®, we are studying the possibilities to
facilitate the use of bio computing by biologists. The work is twofold: first,
a front end should allow biologists to define and execute scenarios of
functions independently from the location of databases and libraries; second, a
study of the expertise of biologists (in defining biologic functions
parameters) should help them to define some massive experimental execution
scenarios. This project, based on high level description of bioInformatics
resources, should lead to a easy to use tool in early 2007.
Cognitive consistency of
documental itineraries
In this
work, involving a PhD Student and supported by an IST European Project
(Mesmuses), we focus on the automatic generation of coherent itineraries
through previously indexed resources. The context of the study is Community
Web, where a community of users index themselves resources according to a
semantic but simple data model. We base our approach on some semiotics
considerations and cognitive constraints, thus compensating the weakness of
indexing. Alternative to more sophisticated hyper document generation
techniques, our approach can even be applied with a downgraded indexing: The
information space is modelled as an hyper graph of documents and document
descriptors. A spanning tree of descriptors is extracted to constitute the
guided tour structure. We use a matching between spanning tree edges and a
subset of transition documents to strengthen guided tour coherence and avoid
redundancy. The model relies on an analysis of cognitive issues of guided tour
consultation.
Expert based evaluation of
rule quality in KDD
Data mining algorithms, especially
those used for unsupervised learning, generate a large quantity of rules, from
which it is necessary to extract the most interesting ones, according to some
expert preferences that are intrinsically hard to express and formalize.
Several measures try to capture such interestingness of rules, but they all
have very different characteristics and drawbacks. We propose in this work a
double-step solution to the choice problem of an expert-adapted interestingness
measure: Firstly, a characterization of interestingness measures, based on
meaningful classical properties, is provided; Secondly, a multicriteria
decision assistance process is applied on this characterization and illustrates
the benefit that a user, who is not a data mining expert, can achieve with such
methods. This work is supported by the French CNRS institution through the STIC
activity Gafodonnées (subgroup Gafoqulité).
Selected publications
- B. Buffereau, P. Picouet: STIL: an
extended resource description framework and an advanced query language for
metadatabases, DASFAA 2006, Singapour, Avril 2006
- B. Buffereau, P. Picouet: A Thematic Guided Tour Model for Contextualized Concept Presentations. AH 2004: 384-388
- P. Lenca, P. Meyer, B. Vaillant, P. Picouet et S. Lallich , Evaluation et analyse multi-critères des mesures de qualité des règles d'association, Revue des Nouvelles Technologies de l'Information, Mesures de Qualité pour la Fouille de Données (rédacteurs invités: Briand H., Sebag M., Gras R., Guillet F.), RNTI-E-1, ISBN 2-85428-646-4, pp. 219-246, 2004.
- M. Hallard, P. Picouet, L.-F. Rodriguez, S. Bigaret: Bioside: faciliter l'accès des biologistes aux ressources bioinformatiques, JOBIM 2004: 64
- P. Picouet, P. Lenca,
"Extraction de connaissances à partir de données", chapitre 13,
dans 'Bases de données et Internet', Série Informatique et Systèmes
d'information, Ed. Hermès,
p 395-420, 2001. E. LE SAUX, P. LENCA, P. PICOUET, J.-P. BARTHELEMY,
"An anthropocentric tool for decision support", in Proceedings
of the sixteenth International Joint Conference on Artificial Intelligence
(IJCAI'99), Stockholm, August 1999.
- P. PICOUET, V. VIANU, "Semantics And
Expressiveness Issues in Active Databases", Journal of Computer and
System Sciences, 57(3), p 325-355 (1998).
|
Email: Philippe.Picouet@enst-bretagne.fr Téléphone / Phone : (+33 / 0) 2 29 00 13 95 Télécopie / Fax : (+33 / 0) 2 29 00 10 30 |
Postal address |