Is trust robust?

John O'Donovan, Barry Smyth
2006 Proceedings of the 11th international conference on Intelligent user interfaces - IUI '06  
Systems that adapt to input from users are susceptible to attacks from those users. Recommender systems are common targets for such attacks since there are financial, political and many other motivations for the false promotion or demotion of recommendable items [2] . Recent research has shown that incorporating trust and reputation models into the recommendation process can have a positive impact on the accuracy of recommendations. In this paper we examine the effect of using five different
more » ... st models in the recommendation process on the robustness of collaborative filtering in an attack situation. In our analysis we also consider factors of quality and accuracy of recommendations. Our results caution that including trust models in recommendation can either reduce or increase prediction shift for an attacked item depending on the model-building process used, while highlighting approaches that appear to be more robust.
doi:10.1145/1111449.1111476 dblp:conf/iui/ODonovanS06 fatcat:xe5y6bcg4bgspdjnvdt6gbh2oy