Research Of Search Engine

The network search engine is very important for searching information from the network resources. Using search engine ,user can gain the network information that he need from a great deal of network information. if people input the keyword of query, the search engine will find relational information for them..But today the technique of those search engines which is not perfect, exist some problem that need to be resolved, such as the comparative low recall rate and precision rate of the search' results, the existing of many duplicated Web pages that not only waste storage resource but also bring many burdens to user in retrieval and reading, etc.When we need some information, generally we shall use a certain search engine, then input keyword, then the search engine would return for us a flood of search results. Usually we can discover many irrelative item and duplicated item from all the search results.The research of the results that returned by search engine take up with clustering analysis or removing the duplicated web page that based on the results of search engine, in order to reduce inconvenience towards user because of unperfect technique of the search engine.Researching the results of search engine, we don't need to extensive collect network information resources data using automatic index program Robot(or spider). The research system only combine the results that other search engine returns, such as URLS, caption, abbreviated describe for web page, etc. The Research System also can combine the results that is from several different search engines. Then it will process those results again.In this paper, the design and simple implement of a research system for the results of search engine will be described. The purpose of this paper is to design and implement a Intelligence Agent System, that automatically compares the results of search engine, and removes those duplicated web pages, then return the left results to user. So the system can aim user to save time and energy.In this paper, we use the search results that provided by existing search engine ( here baidu will be used). The abbreviated describe for web page of these results will be analysed in order to judge whether current web page is duplicated. If current web page is duplicated, it will be removed from the result set.