A collaborative recommendation framework for ontology evaluation and reuse

This is an electronic version of the paper presented at the International Workshop on Recommender Systems, held in Riva del Garda on 2006Ontology evaluation can be defined as assessing the quality and the adequacy of an ontology for being used in a spe-cific context, for a specific goal. Although ontology reuse is being extensively addressed by the Semantic Web community, the lack of appropriate support tools and automatic techniques for the evaluation of certain ontology features are often a barrier for the implementation of successful ontology reuse methods. In this work, we describe the recommender module of CORE [5], a sys-tem for Collaborative Ontology Reuse and Evaluation. This mod-ule has been designed to confront the challenge of evaluating those ontology features that depend on human judgements and are by their nature, more difficult for machines to address. Taking advan-tage of collaborative filtering techniques, the system exploits the ontology ratings and evaluations provided by users to recommend the most suitable ontologies for a given domain. Thus, we claim two main contributions: the introduction of collaborative filtering notion like a new methodology for ontology evaluation and reuse, and a novel recommendation algorithm, which considers specific user requirements and restrictions instead of general user profiles or item-based similarity measures