SEMANTIC-BASED COMPOSITION OF MODULAR ONTOLOGIES APPLIED TO WEB QUERY REFORMULATION

When a Web query is submitted and relevant documents are not found, users are faced with the difficult task of reformulating the query. It may be argued that ontologies can be useful to automate the query reformulation process, taking advantage of the domain knowledge. This paper proposes a novel way of building a modular ontology that may be suitable for this task, composed of interrelated modules focused on specific topics. We propose to use well-known Web-based semantic relatedness measures to improve the content and structure of the modular ontology. Some experiments on query reformulation based on the obtained ontology modules show satisfactory results.