Research On The Technology Of A Binary Search-Based Search Engine

Search engine has developed to today's fourth generation---topic-specific search engine. It is always developing toward humanization and personalization. Not only on the Technology of focused crawling but also on the promotion of Theme Service models, has Search engine gained good development. However, the current search engine can not satisfy the group, the members of which have common interests, on the demand of information.Considering the shortages of the current Search engine, this paper put the solution of the topic-specific search engine; realize the Search engine's thematic requirements. In addition, this paper, series of theory and technical problems involved in the solution are studied. The main frame of this paper is as follows:(1) The paper redefines the concept of topic-specific search engine; Improves the Model framework of topic-specific search engine; puts the working theory of Topic-oriented Meta search engine.(2) In this paper, the author introduces the ant colony algorithm into the technology of topic crawler which is one of the key technologies of topic-specific search engine. So, topic crawlers can traverse the web according to the history and the interests of the users, which can improve the efficiency of the topic crawlers apparently.(3) This paper puts forward Topic recommendation technology, in order to adapt more to the subject description of the group, the member of which have common interest. Through improving the usual method of subject description, this paper establishes subject tree based on the results of the crawling ant colony algorithm and gives the theme vectors.(4) This paper deepens the search concept of binary search and describes the operating principle of the General search engines and the search engines which use the binary search.And also, this paper introduces the Topic recommendation technology which is used in the binary search, giving the arithmetic of Topic recommendation technology.Finally, the evaluation indicator of topic-specific search engine is rebuilt. On base of those evaluation indicators, the experiment is conducted, in order to show the advantage of the topic-specific search engine which uses the binary search.