A novel approach for spam detection using boosting pages

Link spam techniques are widely used in commercial web pages to achieve higher ranking in search results which may reduce the quality of the results. Most of techniques require some boosting pages to increase score of the target pages. In this paper, we aim to detect boosting pages and then spam pages to increase the quality of results. Our approach consists of two steps: (1) find boosting pages from spam seed set, and (2) detect web spam from the discovered boosting pages. An experimental result shows that our approach can detect 93.20 % of web spam pages with 79.96% precision.