Research On Ontology Retrieval System Based On User Implicit Knowledge Mining

With ontology widely used in the semantic web, more and more users access toknowledge from Ontology retrieval systems instead of traditional IR system. Thetraditional keyword-based retrieval system can’t meet user implicit needs, neither canuse ontology effectively.This paper presents an algorithm for ontology searching and ranking based onuser background information mining, and implemented an ontology retrieval systemfor searching ontology more accurately and efficiently. By using association rulemining on an improved data model to obtain user background information and findcommunity (theme) classification of them. We can obtain relevant descriptiondocuments of the community (theme) with association rules, which can extend theoriginal query by using relevance feedback. So the system can identify the user’simplicit demand with the combination of user background knowledge and query. Usepattern matching method to classify semantic relationship of the parsed RDF triples,that to extract the logical view of ontology concepts. A CommunityRank algorithm isproposed for get the semantic ranking with calculation of Cosine Similarity betweenexpanded query and the RDF logical view.In conclusion, the retrieval scheme can sort the query results using semanticranking algorithm based on the background information and implicit needs of the user,thus providing the ontology retrieval services more in line with user needs. This workhas a certain guiding significance for the research of ontology on semantic web.