Ontology-based semantic retrieval for mechanical design knowledge

Abstract

The semantic retrieval of mechanical domain knowledge is critical in product design process. To address the problems with existing keyword-based and semantic-enabled methods, an ontology-based search system (OBSS) for knowledge search and retrieval from knowledge base is proposed. In our scheme, the mechanical design knowledge ontology is first constructed, and query semantic extension and retrieval are then adopted for knowledge retrieval. For the query semantic extension, the semantic similarity analysis is adopted to discover the semantic distance between the semantic key in query and its extended semantic keys. Then, the query boosts of extended semantic keys are modified by the semantic similarity. In the precision and recall experiments, the results demonstrate that our method outperforms the keyword-based search technique. Further, two query cases ensure that the semantic retrieval algorithm based on domain ontology could understand the latent search intents. This research contributes to the query intent capture and the application of engineering ontology for design knowledge retrieval.