Deviation-Based Contextual SLIM Recommenders

摘要: Context-aware recommender systems (CARS) help improve the effectiveness of recommendations by adapting to users' preferences in different contextual situations. One approach CARS that has been shown be particularly effective is Context-Aware Matrix Factorization (CAMF). CAMF incorporates dependencies into standard matrix factorization (MF) process, where users and items are represented as collections weights over various latent factors. In this paper, we introduce another based on an extension factorization, namely, Sparse Linear Method (SLIM). We develop a family deviation-based SLIM (CSLIM) recommendation algorithms learning rating deviations conditions. Our CSLIM better at explaining underlying reasons behind recommendations, our experimental evaluations five context-aware data sets demonstrate these outperform state-of-the-art top-N task. also discuss criteria for selecting appropriate algorithm advance characteristics data.