| With the rapid development of internet, resources on the web are increasingly rich. In order to help people find information in the vast internet, search engines came into being and have been developed rapidly. However, people's information demands become more and more specialized and refinement, the traditional search engines have not been nice to meet people's demands, but the vertical search engine oriented to specific domain is more and more favourite for people. Vertical search is a kind of special information integration on page, directed points to extract field data processing, then returned to the user in some form, which is subdivision and extension of search engines.In information retrieval, based on the mechanical characters of matching keywords may cause some relevant documents queried by users, which is a key to retrieve the performance problem. Query expansion in a certain extent can solve the phenomenon which words do not match, then the paper on the basis of the study and based on the analysis of existed query expansion methods, has a query expansion strategy based on ontology and its research. This paper main contents are as follows:(1) the research based on the analysis of the traditional links (PageRank algorithm, HITS algorithm and the HillTop algorithm), and combining the characteristics of perpendicular search for vertical search algorithm for the v_PageRank query results returned to sort and experimental results verify the new algorithm is better than the original algorithm is improved.(2) based on the analysis of the limitations of keywords and query expansion method, the query expansion strategy is proposed based on ontology, and construct a vertical search engine prototype system in a phone domain ontology.(3) with phone domain ontology, a prototype system of vertical search engine based on query expansion is designed, the experiment result proves that it has better recall and precision than traditional search engine. |
PDF Full Text Request