We present a novel approach to combating web spam. In the spirit of Luis von Ahn's games with a purpose, we propose using a two player game to identify spam pages within search results. Our game asks users to classify a page as either highly relevant to a query or not relevant to a query, with the option of passing. We use data from the game as the input to a simple voting algorithm which determines whether a page is spam. We show that the best strategy for users playing the game for fun is to answer truthfully, and that spammers have difficulty obstructing the game.

22 pages

Return to: SCS Technical Report Collection
School of Computer Science

This page maintained by reports@cs.cmu.edu