Application And Research Of Document Cluster In Web Results Of Search Engine

With the rapid development of computer network, Internet is used to publish and share information, which brings forward more information and difficulty for common users to find information. Search engines are developed to solve this problem. At present, search engines have two distinct flaws, one is the number of results returned by search engine are numerous, the other is that the results are displayed linearly. Based on previous research results and technologies of search engines, we further the research on the document clustering and classifying web search results automatically.Firstly, not only the basic concept and principle of search engines, but also principles of document clustering are introduced, which are useful to form the solution based on Google Web API. The solution clusters the web results returned by search engines and displayed to users structured.The main achievements of the thesis are following:(1) We integrate document clustering with PAT-tree data structure, which improves the performance of document clustering.(2) We integrate document clustering with modern technologies of search engines and bring forward a new architecture of search engine, which solves some problems of modern search engines.(3) Based on previous research on search engine, we develop a demo system, which proves the feasibility of the new architecture of search engine.Search Engine is still a new conception, and many technologies involved are not yet mature and being developed. The future research problems are also discussed in the end of the paper.