Artificial Neural Networks For Content-based Web Spam Detection

Abstract

Web spam has become a big problem in the lives of Internet users, causing personal injury and economic losses. Although some approaches have been proposed to automatically detect and avoid this problem, the high speed the techniques employed by spammers are improved requires that the classifiers be more generic, efficient and highly adaptive. Despite of the fact that it is a common sense in the literature that neural based techniques have a high ability of generalization and adaptation, as far as we know there is no work that explore such method to avoid web spam. Given this scenario and to fill this important gap, this paper presents a performance evaluation of different models of artificial neural networks used to automatically classify and filter real samples of web spam based on their contents. The results indicate that some of evaluated approaches have a big potential since they are suitable to deal with the problem and clearly outperform the state-of-the-art techniques.