An Improved Weighted HITS Algorithm Based on Similarity andPopularity

The HITS algorithm is a very popular and effective algorithm to rank documents based on the link information among a set of documents. However, it assigns every link with the same weight which results in topic drift. In this paper, we generalize the similarity of web pages and propose a query-induced similarity describing how a webpage is similar to another on a query topic. Then, we provide a new improved weighted hits-based (I-HITS) algorithm by assigning appropriate weights to links with the similarity and popularity of web pages. Experiment results indicate that the improved HITS algorithm can find more relevant pages than HITS, ARC, SALSA and improve the relevance by 30%-50%. Furthermore, it can avoid the problem of topic drift and enhance the quality of web search effectively.