Collaborative Filtering(CF)with Implicit Feedbacks,is one of the most popular method in recommender systems.CF with Implicit Feedbacks can be broadly classified into two main categories: point-wise regression based and pairwise ranking based.Implicit feedback is often very sparse,causing CF based methods to degrade significantly in recommendation performance.In this case,side information,abstracted via Stacked Denoising Autoencoder(SDA),is usually introduced and utilized to address the data sparsity problem.Nevertheless,the latent feature representation learned from side information by topic model may not be very effective when the data is too sparse.For that reason,Matrix Factorization(MF)is jointly used with SDA to address this problem under one model called SDAM.SDAM is a tightly coupled,and point-wise approach,which leverages deep feature representation of item content into a Bayesian framework of point-wise ranking model.The experimental analysis on realworld dataset show that the proposed method outperforms several state-of-the-art collaborative filtering methods. |