Research And Implementation Of Authoritative Web Information In Search Optimization

Search engine is the system that it can automatically collect information fromInternet, process them and get users to inquiry. It is convenient for us. But with thefurther development of Internet, the traditional Search engine exposes its deficiency:the users search too many pages that are irrelevant with their concern, and it takes lotsof time to find their concern. On that condition, our research is focused on searchengine and correlation technology.In this paper, we propose the Meta search system; it uses the technology of webtext-mining, Meta search and so on. There are two function modules in the system:Meta search module and text clustering module. In the system when the users use thekey words to search, Meta search module will do some things: transfer other searchengines to carry on the search, Removes these irrelevant search results and giveaccording to some evaluation rules, give the processing result to another module. Intext-clustering mod(?)le, firstly preprocess the search pages, and then (?)ses clusteringalgorithm to generate different clusters to help user search.The core of this system is text-clustering algorithm. We find K-means is easyand fast, and suit for large data. But we also find, there are different clustering results,if we use different initial points in k-means. So we choose authority web pages asinitial points to cluster, and it effects well.##原图像不清晰...