Terminological and ontological resource and semantic annotations for knowledge management

Abstract

In dynamic contexts, ontologies and their lexical component terminological and ontological resource or TOR have to frequently adapt to domain evolutions, new uses and new user needs. Among all depending data, ontology-based semantic annotations also are regularly updated to tag new documents or to reflect new points of view. Within the TextViz ontology-based annotation framework, we propose EvOnto: a tool and a method that supports a coherent joint change management of termino-ontologies and semantic annotations. A special attention is paid to terms and concept labels as well as quality criteria to evaluate automatic text annotations and to detect lacks in the ontology. The method defines a cyclic process: after annotation evaluation, EvOnto proposes various change operators for the TOR and, for each of them; it allows tuning the consequences of this change within the TOR and back on the annotations. In this paper, the method will be motivated and detailed. The help provided by the support tool will be illustrated in a case-study where a collection of software maintenance files are semantically indexed by a domain TOR.