Seyedeh Homa Alizadeh - Computer EngineeringDepartment Imam Reza International University Mashhad, Iran
Majid Vafaei Jahan - Computer Engineering Department Islamic Azad University Mashhad, Iran
N. R. Arghami - R&D Department FAMA Technology Naperville, IL, USA

One of the most popular techniques used in recommender systems is collaborating filtering. In this technique it is usual that Pearson’s correlation is used to find thesimilarity between users. It is a known fact that Pearson’s correlation is not suitable for measuring the strength of nonlinearrelations. Since Spearman’s correlation is in fact Pearson’s correlation applied to ranks and does not work well in non-monotone relationships and since measures like Kendal’stau do not work well in small samples, we introduce a new measure of association to be used in collaborative systems whichwe shall call alpha. Our investigations show it leads to better MAE. We also propose a method by combining Alpha and trust propagation and add anew algorithm to semantic similarity for confronting the problems with cold startand it leads to better coverage.

recommender systems, alpha measure of association, semantic similarity, collaborative filtering, cold start