Génération automatique d'une ontologie dans le domaine des ressources humaines

The current era is increasingly influenced by the prominence of smart data and the semantic web. This does not make recruitment processes easier, especially when seeking candidates and skills. Most systems matching job offers to profiles build on one or more language resources. Developing and maintaining specific resources in each field is difficult. From millions of user profiles gathered on multiple social networks and tens of thousands of job offers collected on the internet, we describe an approach to generating a semi-automatic ontology. A manual evaluation of the results has shown excellent results. MOTS-CLÉS : Ontologie, réseaux sociaux, ressources humaines, application industrielle.