Review Paper: New Advances in Recommender Systems

Abstract: Recommender Systems have been known for the past forty plus years but with the explosion of ecommerce sites and social media. In the new era it has become a major need to personalize the recommender system for every user, collaborative filtering does this job well but only but as users and products scale up we need newer techniques. Privacy is being demanded by internet users so how can one build recommender systems with privacy embedded in it is now one of the issue discussed in this paper. A recommender system also faces attacks on it like fake users, fake ratings the methods to deal with such attacks are listed. Lastly deep learning has shown several advances in the field of computer vision how to use the power of deep learning and Convolution and Recurrent Neural Networks in Recommender systems this paper explicitly covers that.

Keywords: Recommender Systems, NLP,RBM,CNN, RNN