Automatic Annotation of Discourse and Semantic Relations supplemented by Terminology Extraction for Domain Ontology Building and Information Retrieval

摘要

In this article, we develop a framework for the building of domain ontologies and a semantic index based on two technologies: terminology extraction with LEXTER (EDF R&D) and discourse and semantic annotation with EXCOM. We have selected two specific points of view for this study: causality and part-whole notions. In the first part of this paper, we explain the contributions of a terminology and the discursive and semantic relations for domain ontology building. In the second part we propose a semantic based index for information retrieval.