SemOIR: An ontology-based semantic information retrieval system

Abstract

This paper describes a semantic vector space model (SeVSM) and an information retrieval system based on the model. The SeVsmaims to improve information retrieval performance for domain-specific systems. In this model, we use an ontology to build the relations between any two keywords to solve the performance deficiency caused by the basic hypothesis of a vector space model (VSM) where keywords are mutually independent. Then we designed and developed a semantic ontology-based information retrieval (SemOIR) system based on the SeVSM model. An experimental study using 15 queries from different domains confirms the effectiveness of the SeVSM and the usability of the SemOIR system. The proposed model and the system contribute significantly to the application of semantic retrieval for digital libraries and e-commerce systems.