Ontology supported inference system for Hindi and Marathi

In multilingual environment, there are multiple languages exist in a single database system. In order to retrieve correct and exact information from the system, we need an efficient semantic matching approach among the languages. These approaches should lead to compute inferences, which may be effectively utilized in decision making system. There are many approaches proposed for semantic matching, but either they are complex or will not give correct result for Indian languages like Hindi or Marathi. In this paper, we proposed semantic matching approach using ontology for Hindi and Marathi languages for inference. Our approach is supported by our integrated ontology building approach. In order to carry out inference, semantic matching is carried out by using bi-lingual dictionary, synset, ontology for an entered string. We achieved exact, accurate results in terms of precision and recall values for both ontology building and computing inference.