This paper introduces a new memoory based approach to ratings based Collaborative Filtering. Unlike existing memory based Collaborative Filtering approaches, this approach exploits the predictable portions of even some complex relationships between users while selecting the mentors for an active user through the use o f the novel notion of Selective Predictability, which can be measured using the Entropy measure.

The proposed approach has been tested using the MovieLens tlataset, and it is expected that this approach shouM work eqtially well for any given dataset This jlexibili~l would nzake it possible to make use o f this approach in a wide variev of applicatio~l slomains including e-commerce where recommendations neeti to be provided to users based on the ratings provided implicitly or explicitly by different users to different items in the past. However the items should represent a relatively homogeneous group like movies, music albums, compact disks, books. sojtware, research articles etc.