A Generalized Probabilistic Framework and its Variants for Training Top-k Recommender Systems

Accounting for missing ratings in available training data was recently shown [3, 17] to lead to large improvements in the top-k hit rate of recommender systems, compared to state-of-the-art approaches optimizing the popular root-mean-square-error (RMSE) on the observed ratings. In this paper, we take a Bayesian approach, which lends itself natu-rally to incorporating background knowledge concerning the missing-data mechanism. The resulting log posterior distri-bution is very similar to the objective function in [17]. We conduct elaborate experiments with real-world data, testing several variants of our approach under different hypotheti-cal scenarios concerning the missing ratings. In the second part of this paper, we provide a generalized probabilistic framework for dealing with possibly multiple observed rat-ing values for a user-item pair. Several practical applica-tions are subsumed by this generalization, including aggre-gate recommendations (e.g., recommending artists based on ratings concerning their songs) as well as collaborative filter-ing of sequential data (e.g., recommendations based on TV consumption over time). We present promising preliminary experimental results on IP-TV data