Modeling Trust for Recommender Systems using Similarity metrics Georgios Pitsilis and Lindsay F. marshall Abstract. In this paper we present novel techniques for modeling trust relationships that can be used in recommender systems. Such environments exist with the volun- tary collaboration of the community members who have as a common purpose the provision of accurate recommendations to each other. The performance of such sys- tems can be enhanced if the potential trust between the members is properly ex- ploited. This requires that trust relationships are appropriately established betweer them. Our model provides a link between the existing knowledge, expressed in simi larity metrics, and beliefs which are required for establishing a trust community. Al- though we explore this challenge using an empirical approach, we attempt a com parison between the alternative candidate formulas with the aim of finding the optimal one. A statistical analysis of the evaluation results shows which one is the best. We also compare our new model with existing techniques that can be used for 1 Introduction Recommender systems have become popular nowadays as they are widely used in e- commerce. Examples of services which use recommender systems for helping user to choose products they might like are epinions [1], eBay [2] and Amazon [3]. The contribution of recommender systems comes in two forms, either as predicted rat of services that a user wants to know about, lists of services that users might find of interest. The effectiveness of a Recommender system can be measure by the accuracy of the predictions that it makes. Collaborative filtering(CF)[4] is os Pitsilis and Lindsay F. M: of Computing Science, Univ Newcastle Upon-Tyne, Newcastle Upon Tyne, NEI K, e-mail: Georgios. Pitsil Please arse the following farma when citing this chapterModeling Trust for Recommender Systems using Similarity Metrics Georgios Pitsilis and Lindsay F. Marshall1 Abstract. In this paper we present novel techniques for modeling trust relationships that can be used in recommender systems. Such environments exist with the volun￾tary collaboration of the community members who have as a common purpose the provision of accurate recommendations to each other. The performance of such sys￾tems can be enhanced if the potential trust between the members is properly ex￾ploited. This requires that trust relationships are appropriately established between them. Our model provides a link between the existing knowledge, expressed in simi￾larity metrics, and beliefs which are required for establishing a trust community. Al￾though we explore this challenge using an empirical approach, we attempt a com￾parison between the alternative candidate formulas with the aim of finding the optimal one. A statistical analysis of the evaluation results shows which one is the best. We also compare our new model with existing techniques that can be used for the same purpose. 1 Introduction Recommender systems have become popular nowadays as they are widely used in e￾commerce. Examples of services which use recommender systems for helping users to choose products they might like are epinions [1], eBay [2] and Amazon [3]. The contribution of recommender systems comes in two forms, either as predicted rat￾ings of services that a user wants to know about, or as lists of services that users might find of interest. The effectiveness of a Recommender system can be measured by the accuracy of the predictions that it makes. Collaborative filtering (CF) [4] is Georgios Pitsilis and Lindsay F. Marshall School of Computing Science, University of Newcastle Upon-Tyne, Newcastle Upon Tyne, NE1 7RU, U.K., e-mail: Georgios.Pitsilis@ncl.ac.uk Please use the following format when citing this chapter: Pitsilis, G. and Marshall, L. F., 2008, in IFIP International Federation for Information Processing, Volume 263; Trust Management II; Yücel Karabulut, John Mitchell, Peter Herrmann, Christian Damsgaard Jensen; (Boston: Springer), pp. 103–118