Improving the knowledge discovery process using ontologies

Laurent Brisson (1) , Martine Collard (2) , Nicolas Pasquier (2)

1 Laboratoire d'Informatique, Signaux, et Systèmes de Sophia-Antipolis (I3S) / Equipe KEIA
2 I3S - Laboratoire d'Informatique, Signaux, et Systèmes de Sophia Antipolis

In this paper, we present the new ontology-based methodology ExCIS (Extraction using a Conceptual Information System) for integrating expert prior knowledge in a data mining process. This methodology describes guidelines for a data mining process like CRISP-DM. Its originality is to build a specific conceptual information system related to the application domain in order to improve datasets preparation and results interpretation. In this paper we specially present the CIS construction which consists of creating an ontology by information extraction from an initial raw database and building data to be mined.

Keywords: Data Mining

Domains: Information Retrieval [cs.IR]; Databases [cs.DB]

Laurent Brisson, Martine Collard, Nicolas Pasquier. Improving the knowledge discovery process using ontologies. IEEE MCD'2005 international workshop on Mining Complex Data, Nov 2005, Houston, United States. pp.25-32. ⟨hal-00363017⟩