Deep Learning For Recommender Systems

Please enable javascript in your browser.

Abbasi, Omid | 2017

917 Viewed

Type of Document: M.Sc. Thesis
Language: Farsi
Document No: 49663 (19)
University: Sharif University of Technology
Department: Computer Engineering
Advisor(s): Soleimani, Mahdieh
Abstract:
Collaborative fltering (CF) is one of the best and widely employed approaches in Recommender systems (RS). This approach tries to fnd some latent features for users and items so it would predict user rates with these features. Early CF methods used matrix factorization to learn users and items latent features. But these methods face cold start as well as sparsity problem. Recent years methods employ side information along with rating matrix to learn users and items latent features. On the other hand, deep learning models show great potential for learning effective representations especially when auxiliary information is sparse. Due to this feature of deep learning, we use deep learning to learn good representation for items.In particular, we propose a hybrid method with deep learning alongside matrix factorization which creates a two-way interaction between latent features learnt from matrix factorization and features extracted from content of the items with deep leaning methods and simultaneously optimize the parameters of these models. Experiments on real-world datasets show that the proposed method outperform state-of-the-art RS methods
Keywords:
Recommender System ; Deep Learning ; Collaborative Filtering