Multi-level Link Structure Analysis Technqiue for Detecting Link Farm Spam Pages

Link based ranking algorithms have been adopted by most search engines in ranking the importance of Web pages. As a majority of Web accesses are through the use of search engines, Web spammers have taken advantage of the vulnerability of link based ranking algorithms by creating many artificial references or links in order to acquire higher-than-deserved ranking in search engines´ results which will generate higher traffic to their websites. Link farm spam refers to the technique used to create a network of heavy linked Web pages for this purpose. In this paper, an algorithm using multi-level link structure analysis technique is proposed for detecting link farm spam pages. This algorithm is an enhancement to seed and parental penalty algorithms proposed by Wu and Davison, 2005