Reducing Web Spam using Content and Link Analysis

Abstract

Techniques of search engine manipulation are increasing rapidly making the importance of anti-web spam filters evident. In this paper, we fuse content analysis metrics and link analysis algorithms to retrieve relevant documents while blocking spam pages. We compare the efficiency of the algorithm with well-known link algorithms. This implementation aims to maintain a high recall/precision ratio while using two levels of filtering. The hybrid implementation outperforms the popular HITS.