SpotRank: a robust voting system for social news websites

In a social news website people share content they found on the web, called news, then vote for those they like the most. Voting for a news is then considered as a recommendation, and news with a sufficient number of recommendations are displayed on a front page. Malicious users of such websites boost their own content by manipulating the votes. We present SpotRank, an algorithm that can demote the effect of manipulations, thus leading to a better quality of service. We also present a website that implement this algorithm and show evidence of the efficiency of the approach, both from a statistical and human point of view. Categories and Subject Descriptors H.3.3 [Information Search and Retrieval]: Spam demotion General Terms Algorithms, Design, Experimentation, Human Factors

• Type: Conference
• Year: 2010
• Where: WWW
• Authors: Thomas Largillier, Guillaume Peyronnet, Sylvain Peyronnet
• Keywords: General Terms Algorithms, Internet Technology, Malicious Users, People Share Content, WWW 2010