Semantic Association Retrieval Based On Oil Domain Ontology

With the development of today's society,the amount of information is increasing rapidly.It is very important to find out the information needed by the users in the mass of data.Traditional search is based on query keywords.This retrieval model can only get query results only when query keywords are perfectly matched,which leads to missing a lot of information in the middle and not exactly meeting users' needs.The concept of ontology organized in semantic and knowledge level,with the concept of strong hierarchy and semantic relations system can provide logical reasoning support and extended query,semantic association retrieval ontology to enhance the system's recall rate and precision rate based on.Firstly,this paper makes an in-depth study about the definition of ontology,ontology description language,ontology and classification and the existing ontology construction method of ontology and semantic retrieval of relevant theoretical knowledge;secondly,analysis of the existing semantic similarity algorithm including semantic similarity algorithm based on path and based on the information,under the premise this paper uses the multi stage cascaded model of semantic similarity of concepts of phase calculation,and uses BP neural network to train results.The semantic similarity algorithm is very important in the application of ontology,it is an important part of semantic retrieval,semantic similarity algorithm accuracy directly affects the system search results.This paper researches on traditional query retrieval technology,analysis the advantages and disadvantages of different query expansion technology,comprehensive ontology and local analysis method extension technology advantage in the query,the complement each other,integrating the two into the query extension technology,so that the system can extend more query candidate words,so as to improve the accuracy of user query.Finally this paper builds the experimental environment,and construct the semantic association retrieval prototype system based on the oil domain ontology,the improved semantic similarity algorithm,the improved query extension technique and the prototype system were carried out respectively.The query results were analyzed by the parameters of precision,recall and F test.And checking the retrieval effect of the prototype system of semantic association retrieval.The experimental results show that the semantic association retrieval system based on oil domain ontology is higher than the traditional retrieval model in terms of recall and precision.