A Study On The Method Of Ontology-Based Information Retrieval

Along with the advent of the information age, the Internet has become a big vessel, and the amount of information is growing in an exponential manner. Such a large amount of information has provided a good information source to Internet user. However, how to find the information we needed fast and accurately from a huge database is a thorny issue. The Information retrieval technology, especially the creation of Web information search engine enables users to seek useful information from the vast data easily.However, the current mainstream of information retrieval technology mainly bases on keywords and emphasizes the research of search algorithm, which relatively pays less attention to semantic retrieval. There are a lot of heterogeneity of information format,semantic multiplicity and non-uniform of information relations on the Internet, which brings a great inconvenience for users in information management and acquisition. Although the current search engine technology has improved a lot, enhanced efficiency by applying new technology, such as natural language processing and data mining, the recall rate is still not high on the whole.Semantic web model makes a new direction of development for information retrieval. In order to improve recall rate, this paper presents an ontology-based information retrieval framework, which is on the basis of some ontology-based information retrieval methods. This retrieval framework uses ontology-based local corpus analytical query extension technology, which bases on the combination of local-corpus-analysis-based query extension methods and ontology technology. Ontology's knowledge expression of domain information makes this new method achieves the support of semantic, and optimizes process of document analysis, as well as improve the analysis efficiency. Under the guarantee of the ontology and local corpus analysis, the accuracy and relevance of the extension result has improved a lot.In order to better improve the efficiency of ontology-based local corpus analytical query extension technology, this paper do some optimization to the ontology model, and deal with some horizontal linkages of concepts and entities which eliminate the network-like structure and maintains the tree hierarchy structure in the ontology model. This reduces the complexity of ontology model, as well as makes the ontology structure and relationship of concepts more clear,and similarity calculation more simpler.This paper applies hierarchical vector space model in the process of document. After analyzing the hierarchical vector space model, this paper points out the errors in the formula, and proposes the revised formula. This paper designs an experiment to verify the effectiveness of the retrieval framework and related methods. This paper builds an experimental ontology of athletics project, and selects 60 pieces of related news as the experimental document. After processing the document, this paper use ontology-based local corpus analytical query extension method to process the user retrieval requests, and search in the framework which is presented in this paper. The result shows that ontology-based local corpus analytical query extension method has an expansive coverage of users'retrieval keywords and high relevance, which is also high in the final retrieval document. The experimental result achieves the expected effect, and improves the retrieval recall rate with the guarantee of recall precision.