Research On Optimizing Strategies Of Search Engine Based On PageRank Algorithm

Along with rapid progress of computer network technology, WWW has become the main facility that people used to release, interchange and retrieve information. It is involved in many fields such as news, ad, consumption, finance, education and e-business. Four characteristics of Web are the following: Big scale, Dynamic, Isomerous and Half-structured data condition. Web abounds with hyperlink resources as well. Due to these characters, we can use Search Engine technology to get information and data from web. Three main contents of Web Mining are Content Mining, Structure Mining and Usage Mining. Web Structure Mining is to deduce some kind of knowledge from structures of WWW, Web documents and hyperlinks. As for Search Engine, We can establish a linking structure pattern by analyzing quantity and targets of a web page or website's in-links and out-links. By studying such algorithms based on hyperlink as PageRank, HITS and TSPR, we can guide our linking optimization and continuously improve the website's rank, avoid bad results of blindness. The article focuses on the PageRank, analyzes the idea and calculating method of the algorithm, establishes different models and advances related optimizing strategies. Furthermore, some program is made in Java to validate the result as well. Finally, the article concludes the disadvantages of the PageRank and introduces TSPR(Topic Sensitive PageRank), Hilltop. Web Mining is our request in gathering information from big scale knowledge. As for Search Engine, Web Mining technology plays an important role in development of the third generation of search engine, and meanwhile, it promotes the network information acquiring technology to a high precision and intelligent way.