Gamma-Poisson Dynamic Matrix Factorization Embedded with Metadata Influence

A conjugate Gamma-Poisson model for Dynamic Matrix Factorization incorporated with metadata influence (mGDMF for short) is proposed to effectively and efficiently model massive, sparse and dynamic data in recommendations. Modeling recommendation problems with a massive number of ratings and very sparse or even no ratings on some users/items in a dynamic setting is very demanding and poses critical challenges to well-studied matrix factorization models due to the large-scale, sparse and dynamic nature of the data. Our proposed mGDMF tackles these challenges by introducing three strategies: (1) constructing a stable Gamma-Markov chain model that smoothly drifts over time by combining both static and dynamic latent features of data; (2) incorporating the user/item metadata into the model to tackle sparse ratings; and (3) undertaking stochastic variational inference to efficiently handle massive data. mGDMF is conjugate, dynamic and scalable. Experiments show that mGDMF significantly (both effectively and efficiently) outperforms the state-of-the-art static and dynamic models on large, sparse and dynamic data.

[1]  L. Eon Bottou.  Online Learning and Stochastic Approximations , 1998 .

[2]  Longbing Cao,et al.  Metadata-dependent Infinite Poisson Factorization for Efficiently Modelling Sparse and Large Matrices in Recommendation , 2018, IJCAI.

[3]  Ali Taylan Cemgil,et al.  Bayesian Inference for Nonnegative Matrix Factorisation Models , 2009, Comput. Intell. Neurosci..

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

[5]  Xindong Wu,et al.  Cross-Domain Collaborative Filtering over Time , 2011, IJCAI.

[6]  Wei Zhang,et al.  A Collective Bayesian Poisson Factorization Model for Cold-start Local Event Recommendation , 2015, KDD.

[7]  ChengYue Gong,et al.  Deep Dynamic Poisson Factorization Model , 2017, NIPS.

[8]  Lan Du,et al.  Leveraging Node Attributes for Incomplete Relational Data , 2017, ICML.

[9]  Yue Zhao,et al.  Dynamic Poisson Factor Analysis , 2016, 2016 IEEE 16th International Conference on Data Mining (ICDM).

[10]  David M. Blei,et al.  Bayesian Nonparametric Poisson Factorization for Recommendation Systems , 2014, AISTATS.

[11]  Longbing Cao,et al.  Modeling Asymmetry and Tail Dependence among Multiple Variables by Using Partial Regular Vine , 2014, SDM.

[12]  Michael I. Jordan,et al.  An Introduction to Variational Methods for Graphical Models , 1999, Machine Learning.

[13]  Longbing Cao,et al.  Non-IID Recommender Systems: A Review and Framework of Recommendation Paradigm Shifting , 2016, ArXiv.

[14]  Guillermo Sapiro,et al.  Online Learning for Matrix Factorization and Sparse Coding , 2009, J. Mach. Learn. Res..

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

[16]  David B. Dunson,et al.  Beta-Negative Binomial Process and Poisson Factor Analysis , 2011, AISTATS.

[17]  Mingyuan Zhou,et al.  Poisson-Gamma dynamical systems , 2016, NIPS.

[18]  Chong Wang,et al.  Stochastic variational inference , 2012, J. Mach. Learn. Res..

[19]  Longbing Cao,et al.  Coupled Poisson Factorization Integrated With User/Item Metadata for Modeling Popular and Sparse Ratings in Scalable Recommendation , 2018, AAAI.

[20]  Longbing Cao,et al.  Vine Copula-Based Asymmetry and Tail Dependence Modeling , 2018, PAKDD.

[21]  Yehuda Koren,et al.  Collaborative filtering with temporal dynamics , 2009, KDD.

[22]  Kush R. Varshney,et al.  Collaborative Kalman Filtering for Dynamic Matrix Factorization , 2014, IEEE Transactions on Signal Processing.

[23]  Joydeep Ghosh,et al.  Gamma Process Poisson Factorization for Joint Modeling of Network and Documents , 2015, ECML/PKDD.

[24]  Ali Taylan Cemgil,et al.  Conjugate Gamma Markov Random Fields for Modelling Nonstationary Sources , 2007, ICA.

[25]  Guandong Xu,et al.  Improving the Quality of Recommendations for Users and Items in the Tail of Distribution , 2017, ACM Trans. Inf. Syst..

[26]  Xi Chen,et al.  Temporal Collaborative Filtering with Bayesian Probabilistic Tensor Factorization , 2010, SDM.

[27]  Sotirios Chatzis.  Dynamic Bayesian Probabilistic Matrix Factorization , 2014, AAAI.

[28]  Longbing Cao,et al.  Coupling learning of complex interactions , 2015, Inf. Process. Manag..

[29]  David M. Blei,et al.  Scalable Recommendation with Hierarchical Poisson Factorization , 2015, UAI.

[30]  Joydeep Ghosh,et al.  Nonparametric Bayesian Factor Analysis for Dynamic Count Matrices , 2015, AISTATS.

[31]  J. Hanley,et al.  The meaning and use of the area under a receiver operating characteristic (ROC) curve. , 1982, Radiology.

[32]  Tina Eliassi-Rad,et al.  A Probabilistic Model for Using Social Networks in Personalized Item Recommendation , 2015, RecSys.

[33]  P. Cochat,et al.  Et al , 2008, Archives de pediatrie : organe officiel de la Societe francaise de pediatrie.

[34]  Michael W. Berry,et al.  Algorithms and applications for approximate nonnegative matrix factorization , 2007, Comput. Stat. Data Anal..

[35]  Peter Green,et al.  Markov chain Monte Carlo in Practice , 1996 .

[36]  Barbara E. Engelhardt,et al.  Hierarchical Compound Poisson Factorization , 2016, ICML.

[37]  Lawrence Carin,et al.  Topic-Based Embeddings for Learning from Large Knowledge Graphs , 2016, AISTATS.

[38]  Philip S. Yu,et al.  Coupled Behavior Analysis with Applications , 2012, IEEE Transactions on Knowledge and Data Engineering.

[39]  Longbing Cao,et al.  Learning Nonparametric Relational Models by Conjugately Incorporating Node Information in a Network , 2017, IEEE Transactions on Cybernetics.

[40]  D. Dunson,et al.  Bayesian latent variable models for mixed discrete outcomes. , 2005, Biostatistics.

[41]  Òscar Celma,et al.  Music recommendation and discovery in the long tail , 2008 .

[42]  Ghassen Jerfel,et al.  Dynamic Collaborative Filtering With Compound Poisson Factorization , 2016, AISTATS.

[43]  Jaana Kekäläinen,et al.  Cumulated gain-based evaluation of IR techniques , 2002, TOIS.

[44]  David M. Blei,et al.  Content-based recommendations with Poisson factorization , 2014, NIPS.

[45]  John F. Canny,et al.  GaP: a factor model for discrete data , 2004, SIGIR '04.

[46]  Ee-Peng Lim,et al.  Modeling Temporal Adoptions Using Dynamic Matrix Factorization , 2013, 2013 IEEE 13th International Conference on Data Mining.

[47]  David M. Blei,et al.  Dynamic Poisson Factorization , 2015, RecSys.

[48]  Hongyuan Zha,et al.  Recurrent Poisson Factorization for Temporal Recommendation , 2020, IEEE Transactions on Knowledge and Data Engineering.