点击次数：

论文类型： 会议论文

发表时间： 2007-11-01

收录刊物： EI、CPCI-S、CPCI-SSH、Scopus

页面范围： 149-151

摘要： Link Analysis plays an important role in Web Information Retrieval. Most of the existing link analysis algorithms use link as the main factor to determine the importance of a web page, which can not reflect complicated relationships among web pages. In this paper, we propose G-HITS (Gravitation-Based HITS), which condenses useful information associate to a web as its "mass", and reflects relationship between two web pages as "attractive force" inspired by Newton's theory of gravitation, thus giving a physical interpretation for link analysis. Experimental results show that G-HITS is more reasonable to interpret link analysis problems and it can identify higher quality authorities on given topics.