In this paper, we define social parcimonious language models that emphasize the most important terms in documents, and lower less important terms. The detection of important terms relies on the document itsef and on the tags that were employed by users to describe the document. Conversely to classical personalization approaches that focus first on user’s profiles or on the matching function, ou...

In this paper we introduce a machine learning approach for automatic text segmentation. Our text segmenter clusters text-segments containing similar concepts. It first discovers the different concepts present in a text, each concept being defined as a set of representative terms. After that the text is partitioned into coherent paragraphs using a hard clustering technique based on the Classific...

We introduce a personal assistant based on a distributed architecture. A portal connects user to applications. Applications are installed by a user who compose his own assistant according to his needs. Mots-clés : Système de dialogue, applications du traitement automatique du langage naturel, assistant personnel.

This note characterizes the optimal base for commodity taxation in the presence of administrative fixed costs varying across goods. For low tax rates, the optimal base comprises all commodities whose discouragement index is greater than the ratio of their administrative costs to the tax they yield. JEL classification numbers: H21. Mots-clés : Taxation indirecte, TVA, assiette fiscale, coûts de ...

We construct a KK-theory for Banach algebras, equivariant with respect to the action of a groupoid. We prove the Baum–Connes conjecture with commutative coefficients for hyperbolic groups and for the Poincaré groupoids of foliations with a compact base and a longitudinal Riemannian metrics with negative sectional curvature. Mots clés : théorie de Kasparov ; conjecture de Baum–Connes ; algèbres ...

DAnIEL, parsimonious yet high-coverage multilingual epidemic surveillance DAnIEL is a multilingual epidemic surveillance system. DAnIEL relies on a parsimonious scheme making it possible to process new languages at small cost. MOTS-CLÉS : extraction d’information, recherche d’information, veille, multilinguisme, genre journalistique, grain caractère.

Résumé : Cet article présente les résultats d’une évaluation sur l’intégration des données issues d’un lexique syntaxique, le Lexique-Grammaire, dans un analyseur syntaxique. Nous montrons qu’en modifiant le jeu d’étiquettes des verbes et des noms prédicatifs, un analyseur syntaxique probabiliste non lexicalisé obtient des performances accrues sur le français. Mots clés : Analyse syntaxique pro...

Knowledge management systems rely upon ontologies in order to describe a domain's knowledge. Concepts evolve in a dynamic multi actor context. Therefore, it is necessary to study the ontology evolution. In this paper, we present an approach for a dynamic evolution of ontologies in semantic web using the agent paradigm. MOTS-CLÉS : web sémantique, évolution d’ontologie, système multi-agents, ges...

We describe an approach focussing on hierarchical composition of components during both design and development levels. We present an overview of our UML composition model which is based on well-defined composition properties. Thus we present how we have implemented a control mechanism of theses properties into Julia, an implementation of the Fractal component model. MOTS-CLÉS : UML, composition...

This paper presents the URPAH team’s participation in DEFT 2012.Our approach uses noun phrases in the automatic identification of keywords indexing the content of scientific papers published in a review of Human and Social Sciences, with assistance from the terminology of keywords (piste1) and without terminology (piste2 ) MOTS-CLÉS : syntagmes nominaux, patrons syntaxiques, recherche d’informa...