An Intelligent Search Engine Using Improved Web Hyperlink Mining Algorithm

This paper presents a new search engine-Luka. Luka uses the cluster crawlers to enhance the search performance. We put forward a new web page ranking algorithm to improve the relevance of the results of search engines in Luka. This algorithm combines a link analysis algorithm with content correlation techniques to dynamically rank the pages, which are more relevant to the query. We use the anchor text of a Web page to compute its similarity to the query and use the similarity to readjust the rank of the page pre-calculated with the PageRank algorithm. Our experiments have shown a significant improvement to the current link analysis algorithm.