Research Of Meta-based Web Military Intelligence Search Technology

Along with the development of the Internet, the net has already become the main mains to gain information. However, it is extremely difficult to find the needed information in a great deal of resource on the Internet. The coming of the search engine solved the information search issue; The meta search engine, on the basis of profession search engine, make full uses performances and features of profession search engine. It can provide entirely and accurate researched results for user by call a lot of profession search engine. The related technology for web net military intelligence information based on Meta Search Engine is studied. The main works in this paper were as follows:The paper introduced the related theories of the search engine and meta search engine, including the emergence, development and the research difficulties. The structure, classification, characteristics and evaluation indicators of the Meta Search Engine are focused on. The key technologies such as personal interest models, members of the search engine selection and merging the results were introduced. The meta-search engine personalization model is proposed. The personal interest model is built based on detailed analysis the of meta-search engine personalized service. Its structure, classification, organization and the vector structure has been focused on and the demand for personalized search conversion algorithm is gived out.The members search engine selection strategy are proposed. The evaluation and parameterized method of the members engine are given in contrast to the existing the members search engine selection method. The members'engine selection algorithm is focused based on personal interest model. According to personal interests model from meta search engine model, the parameterized member engine are related calculated, and then the reasonable member engine are selected. The returning results synthetic methods of members'engine are proposed. The results page content extraction is discussed. It is raised that Keywords correlation matrix and the results are expressed as a function TFIDF feature vectors. an improved k-means clustering algorithm is focused on and tested, analysised.