Integrating Context Similarity with Sparse Linear Recommendation Model

摘要: Context-aware recommender systems extend traditional by adapting their output to users’ specific contextual situations. Most of the existing approaches context-aware recommendation involve directly incorporating context into standard algorithms (e.g., collaborative filtering, matrix factorization). In this paper, we highlight importance similarity and make attempt incorporate it recommender. The underlying assumption behind is that lists should be similar if situations are similar. We integrate with sparse linear model build a similarity-learning model. Our experimental evaluation demonstrates proposed able outperform several state-of-the-art for top-N task.