Analysis Of Search Engine

With the expensive development of the Internet, the contents on the Web are increasing exponentially. Ordinary Internet users who want to find the necessary information are very difficult to find them. Therefore, there is an urgent need for a superior search service, which is to facilitate the complicated contents access to information. Search engine technology as the solution to this problem has made outstanding contributions, but it is significant to the users of search engine that the search engine results on the quality of the performance and better quality of the pages can result in a better position, which is also the key indicator to measure the search engine technology. So on page assessment of the importance and the importance of search engines are the sort of technology to solve.First, this paper gives a brief on the composition of search engines, principles, processes, the development of the status and the existence of the advantages and disadvantages; Web mining includes three areas: content mining, structure mining, using mining.Secondly, through the analysis and research of PageRank, HITS, the two algorithms are compared on the link and the number of the links and their own web link structure model. This paper mainly researches the current mainstream of the PageRank algorithm, focusing on the formation of ideas, calculation methods.Finally, through analyzing the features of the PageRank to introduce the improved algorithm SP-PageRank, and I deeply study the principle on the exchange of internal memory and external memory. At last, I achieve a platform of PageRank and SP-PageRank which are based on data prefetching with Java language. The experiments are carried on 3 link datasets from the SOGOU Lab, and the experimental results show that PageRank and SP-PageRank algorithm based on data prefetching are more efficient than not using data prefetching.