Context based information retrieval based on ontological concepts

Ontology has rich internal structure. There are relations and constraints between the concepts that is why Ontology has a richer internal structure. Ontology can be used for information retrieval. Ontology is a halfway determination of a conceptual vocabulary to be utilized for formulating knowledge-level hypothesis around a domain of discourse. The key part of ontology is to help knowledge sharing and reuse. The process of allotting descriptions to documents in an IRS is called indexing. In this paper a technique is proposed which improves results. In this technique web pages are stored in xml database. WordNet is used as dictionary for finding synonyms of user´s query. The technique is based on context of word. Using context of words helps improvise the search results. The technique is called ontological indexing. This technique is compared with text based search. The words on the web pages are mapped to concepts in ontology. Mapping score is generated for each word. Results of search depend on value of mapping score. Moreover the issue of word sense disambiguation is solved up to some extent using parts of speech tagger.