Ontology-based information retrieval and extraction

Using domain knowledge and semantics to conduct effective information retrieval (IR) is one of the major challenges in IR. In this paper we introduce a framework that can facilitate information retrieval based on a sharable domain ontology. Specifically we conduct semantic annotation on the descriptive metadata of images or articles and convert metadata automatically into machine readable format in terms of semantic web representation. The retrieval of an image instance or information extraction of an historic question can then be conducted by matching the semantic and structural descriptions of the user query based on those of the annotated descriptive metadata. We show that with aid of domain ontology, an IR system can achieve a better performance than the keyword-based information retrieval systems.