A Survey on Link Based Algorithms for Web Spam Detection

Abstract

Web spamming techniques aim to achieve undeserved rankings in search results. Existing spam pages cause distrust to search engine results. This is not only wastes the time of visitors, but also wastes lots of search engine resources. Research has been widely conducted on identifying such spam and neutralizing its influence. Spammers use three kinds of spamming techniques to get higher score in ranking. These techniques are Link based techniques, hiding techniques and Content-based techniques. In turn, we perform a sub categorization of link-based category into five groups. These are labels propagation, link pruning, reweighting, labels refinement and graph regularization, and feature based. Experimental results show that some of these techniques are working well and can find spam pages more accurate than the others. This paper performs a survey on Link based algorithms for web spam detection.