Abstract: Improving retrieval systems performance involves discovering more relevant answers and at the same time reducing the number of non-relevant ones that will be retrieved. Providing advanced query functionalities can serve as support to the improvement. In the legal domain, known by its complexity, advanced querying requires on beyond a rich annotation of documents that takes into account their semantic content as well as their structure. The proposed approach demonstrates that the use of NLP techniques and tools in conjunction with a domain ontology that automate the process of information extraction from legal texts through a meaningful annotation is of paramount importance to provide a better access to Arabic legal information.

Keywords: Legal documents annotation, Document’s structure, Document’s semantic content, domain ontology.