Conception d'un outil d'aide \`a l'indexation de ressources p\'edagogiques - Extraction automatique des th\'ematiques et des mots-clefs de documents UNIT

Abstract

Indexing learning documents using the Learning Object Metadata (LOM) is often carried out manually by archivists. Filling out the LOM fields is a long and difficult task, requiring a complete reading and a full knowledge on the topic dealt within the document. In this paper, we present an innovative model and method to assist the archivists in finding the important concepts and keywords of a learning document. The application is performed using wikipedia's category links.