A Context-Aware Method for Top-k Recommendation in Smart TV

We discuss the video recommendation for smart TV, an increasingly popular media service that provides online videos by TV sets. We propose an effective video recommendation model for smart TV service (RSTV) based on the developed Latent Dirichlet allocation(LDA) to make personalized top-k video recommendation. In addition, we present proper solutions for some critical problems of the smart TV recommender system, such as sparsity problem and contextual computing. Our analysis is conducted using a real world dataset gathered from Hisense smart TV platform, JuHaoKan Video-on-Demand dataset(JHKVoD), which is an implicit watch-log dataset collecting sets of videos watched by each user with their corresponding timestamps. We fully portray our dataset in many respects, and provide details on the experimentation and evaluation framework. Result shows that RSTV performs better comparing to many other baselines. We analyse the influence of some of the parameters as well as the contextual granularity.

[1]  Bernd Ludwig,et al.  Matrix factorization techniques for context aware recommendation , 2011, RecSys '11.

[2]  Bamshad Mobasher,et al.  Splitting approaches for context-aware recommendation: an empirical study , 2014, SAC.

[3]  Michael I. Jordan,et al.  Latent Dirichlet Allocation , 2001, J. Mach. Learn. Res..

[4]  Bart Goethals,et al.  Top-N Recommendation for Shared Accounts , 2015, RecSys.

[5]  Bamshad Mobasher,et al.  CSLIM: contextual SLIM recommendation algorithms , 2014, RecSys '14.

[6]  Yifan Hu,et al.  Collaborative Filtering for Implicit Feedback Datasets , 2008, 2008 Eighth IEEE International Conference on Data Mining.

[7]  Jennifer Golbeck,et al.  Generating Predictive Movie Recommendations from Trust in Social Networks , 2006, iTrust.

[8]  Nuria Oliver,et al.  Multiverse recommendation: n-dimensional tensor factorization for context-aware collaborative filtering , 2010, RecSys '10.

[9]  Alexander Tuzhilin,et al.  Experimental comparison of pre- vs. post-filtering approaches in context-aware recommender systems , 2009, RecSys '09.

[10]  Ruslan Salakhutdinov,et al.  Bayesian probabilistic matrix factorization using Markov chain Monte Carlo , 2008, ICML '08.

[11]  Ruslan Salakhutdinov,et al.  Probabilistic Matrix Factorization , 2007, NIPS.

[12]  Yehuda Koren,et al.  Factorization meets the neighborhood: a multifaceted collaborative filtering model , 2008, KDD.

[13]  Bamshad Mobasher,et al.  CARSKit: A Java-Based Context-Aware Recommendation Engine , 2015, 2015 IEEE International Conference on Data Mining Workshop (ICDMW).

[14]  Xiaohui Yan,et al.  A biterm topic model for short texts , 2013, WWW.