Fast Greedy MAP Inference for Determinantal Point Process to Improve Recommendation Diversity

The determinantal point process (DPP) is an elegant probabilistic model of repulsion with applications in various machine learning tasks including summarization and search. However, the maximum a posteriori (MAP) inference for DPP which plays an important role in many applications is NP-hard, and even the popular greedy algorithm can still be too computationally expensive to be used in large-scale real-time scenarios. To overcome the computational challenge, in this paper, we propose a novel algorithm to greatly accelerate the greedy MAP inference for DPP. In addition, our algorithm also adapts to scenarios where the repulsion is only required among nearby few items in the result sequence. We apply the proposed algorithm to generate relevant and diverse recommendations. Experimental results show that our proposed algorithm is significantly faster than state-of-the-art competitors, and provides a better relevance-diversity trade-off on several public datasets, which is also confirmed in an online A/B test.

[1]  M. L. Mehta,et al.  ON THE DENSITY OF EIGENVALUES OF A RANDOM MATRIX , 1960 .

[2]  O. Macchi.  The coincidence approach to stochastic point processes , 1975, Advances in Applied Probability.

[3]  M. L. Fisher,et al.  An analysis of approximations for maximizing submodular set functions—I , 1978, Math. Program..

[4]  Michel Minoux,et al.  Accelerated greedy algorithms for maximizing submodular set functions , 1978 .

[5]  Maurice Queyranne,et al.  An Exact Algorithm for Maximum Entropy Sampling , 1995, Oper. Res..

[6]  G. W. Stewart,et al.  Matrix Algorithms: Volume 1, Basic Decompositions , 1998 .

[7]  Jade Goldstein-Stewart,et al.  The use of MMR, diversity-based reranking for reordering documents and producing summaries , 1998, SIGIR '98.

[8]  Ellen M. Voorhees,et al.  The TREC-8 Question Answering Track Report , 1999, TREC.

[9]  Jaana Kekäläinen,et al.  IR evaluation methods for retrieving highly relevant documents , 2000, SIGIR '00.

[10]  George Karypis,et al.  Evaluation of Item-Based Top-N Recommendation Algorithms , 2001, CIKM '01.

[11]  Barry Smyth,et al.  Improving Recommendation Diversity , 2001 .

[12]  Sean M. McNee,et al.  Improving recommendation lists through topic diversification , 2005, WWW '05.

[13]  Mi Zhang,et al.  Avoiding monotony: improving the diversity of recommendation lists , 2008, RecSys '08.

[14]  Sreenivas Gollapudi,et al.  Diversifying search results , 2009, WSDM '09.

[15]  Cong Yu,et al.  It takes variety to make a world: diversification in recommender systems , 2009, EDBT '09.

[16]  Yehuda Koren,et al.  Matrix Factorization Techniques for Recommender Systems , 2009, Computer.

[17]  Malik Magdon-Ismail,et al.  On selecting a maximum volume sub-matrix of a matrix and related problems , 2009, Theor. Comput. Sci..

[18]  Ben Taskar,et al.  Structured Determinantal Point Processes , 2010, NIPS.

[19]  Harald Steck,et al.  Item popularity and recommendation accuracy , 2011, RecSys '11.

[20]  Tova Milo,et al.  Diversification and refinement in collaborative filtering recommender , 2011, CIKM '11.

[21]  Thierry Bertin-Mahieux,et al.  The Million Song Dataset , 2011, ISMIR.

[22]  Gediminas Adomavicius,et al.  Maximizing Aggregate Recommendation Diversity: A Graph-Theoretic Approach , 2011, RecSys 2011.

[23]  Neil J. Hurley,et al.  Novelty and Diversity in Top-N Recommendation -- Analysis and Evaluation , 2011, TOIT.

[24]  Ben Taskar,et al.  Learning Determinantal Point Processes , 2011, UAI.

[25]  Ben Taskar,et al.  k-DPPs: Fixed-Size Determinantal Point Processes , 2011, ICML.

[26]  Alexander J. Smola,et al.  Fair and balanced: learning to present news stories , 2012, WSDM '12.

[27]  Ben Taskar,et al.  Near-Optimal MAP Inference for Determinantal Point Processes , 2012, NIPS.

[28]  Ben Taskar,et al.  Determinantal Point Processes for Machine Learning , 2012, Found. Trends Mach. Learn..

[29]  Yuli Ye,et al.  Max-Sum diversification, monotone submodular functions and dynamic updates , 2012, PODS '12.

[30]  Joseph Naor,et al.  A Tight Linear Time (1/2)-Approximation for Unconstrained Submodular Maximization , 2012, 2012 IEEE 53rd Annual Symposium on Foundations of Computer Science.

[31]  Jingrui He,et al.  GenDeR: A Generic Diversified Ranking Algorithm , 2012, NIPS.

[32]  Xiaoyan Zhu,et al.  Promoting Diversity in Recommendation by Entropy Regularizer , 2013, IJCAI.

[33]  Ben Carterette,et al.  Preference based evaluation measures for novelty and diversity , 2013, SIGIR.

[34]  Tevfik Aytekin,et al.  Clustering-based diversity improvement in top-N recommendation , 2013, Journal of Intelligent Information Systems.

[35]  Katja Niemann,et al.  A new collaborative filtering approach for increasing the aggregate diversity of recommender systems , 2013, KDD.

[36]  Yong Yu,et al.  Set-oriented personalized ranking for diversified top-n recommendation , 2013, RecSys.

[37]  Kristen Grauman,et al.  Diverse Sequential Subset Selection for Supervised Video Summarization , 2014, NIPS.

[38]  Ben Taskar,et al.  Expectation-Maximization for Learning Determinantal Point Processes , 2014, NIPS.

[39]  Saul Vargas,et al.  Coverage, redundancy and size-awareness in genre diversity for recommender systems , 2014, RecSys '14.

[40]  Jennifer Gillenwater.  Approximate inference for determinantal point processes , 2014 .

[41]  Zheng Wen,et al.  Optimal Greedy Diversity for Recommendation , 2015, IJCAI.

[42]  Suvrit Sra,et al.  Fixed-point algorithms for learning determinantal point processes , 2015, ICML.

[43]  Ulrich Paquet,et al.  Bayesian Low-Rank Determinantal Point Processes , 2016, RecSys.

[44]  Edward Y. Chang,et al.  Tweet Timeline Generation with Determinantal Point Processes , 2016, AAAI.

[45]  Choon Hui Teo,et al.  Adaptive, Personalized Diversity for Visual Discovery , 2016, RecSys.

[46]  Suvrit Sra,et al.  Gaussian quadrature for matrix inverse forms with applications , 2015, ICML.

[47]  Yves Grandvalet,et al.  A Coverage-Based Approach to Recommendation Diversity On Similarity Graph , 2016, RecSys.

[48]  Nicholas Jing Yuan,et al.  Relevance Meets Coverage , 2016, ACM Trans. Intell. Syst. Technol..

[49]  Sunju Park,et al.  A Single-Step Approach to Recommendation Diversification , 2017, WWW.

[50]  Jinwoo Shin,et al.  Faster Greedy MAP Inference for Determinantal Point Processes , 2017, ICML.

[51]  Ulrich Paquet,et al.  Low-Rank Factorization of Determinantal Point Processes , 2017, AAAI.

[52]  Wei Zeng,et al.  Adapting Markov Decision Process for Search Result Diversification , 2017, SIGIR.