A Novel Evidence-Based Bayesian Similarity Measure for Recommender Systems

User-based collaborative filtering, a widely used nearest neighbour-based recommendation technique, predicts an item’s rating by aggregating its ratings from similar users. User similarity is traditionally calculated by cosine similarity or the Pearson correlation coefficient. However, both of these measures consider only the direction of rating vectors, and suffer from a range of drawbacks. To overcome these issues, we propose a novel Bayesian similarity measure based on the Dirichlet distribution, taking into consideration both the direction and length of rating vectors. We posit that not all the rating pairs should be equally counted in order to accurately model user correlation. Three different evidence factors are designed to compute the weights of rating pairs. Further, our principled method reduces correlation due to chance and potential system bias. Experimental results on six real-world datasets show that our method achieves superior accuracy in comparison with counterparts.

[1]  ZhangJingjing,et al.  Impact of data characteristics on recommender systems performance , 2012 .

[2]  Jure Leskovec,et al.  Effects of user similarity in social media , 2012, WSDM '12.

[3]  Gediminas Adomavicius,et al.  Impact of data characteristics on recommender systems performance , 2012, TMIS.

[4]  Jun Zhang,et al.  Single-trial ERPs denoising via collaborative filtering on ERPs images , 2015, Neurocomputing.

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

[6]  Pattie Maes,et al.  Social information filtering: algorithms for automating “word of mouth” , 1995, CHI '95.

[7]  David Heckerman,et al.  Empirical Analysis of Predictive Algorithms for Collaborative Filtering , 1998, UAI.

[8]  Peter Norvig,et al.  Artificial Intelligence: A Modern Approach , 1995 .

[9]  Xin Wang,et al.  Compressed knowledge transfer via factorization machine for heterogeneous collaborative recommendation , 2015, Knowl. Based Syst..

[10]  Joshua Zhexue Huang,et al.  Recommending high-utility search engine queries via a query-recommending model , 2015, Neurocomputing.

[11]  Anthony O'Hagan,et al.  Bayesian Statistics: Principles and Benefits , 2022 .

[12]  Neil Yorke-Smith,et al.  A Novel Bayesian Similarity Measure for Recommender Systems , 2013, IJCAI.

[13]  Hareton K. N. Leung,et al.  Improving network topology-based protein interactome mapping via collaborative filtering , 2015, Knowl. Based Syst..

[14]  George Karypis,et al.  A Comprehensive Survey of Neighborhood-based Recommendation Methods , 2011, Recommender Systems Handbook.

[15]  Dekang Lin,et al.  An Information-Theoretic Definition of Similarity , 1998, ICML.

[16]  Jun Zhang,et al.  The efficient imputation method for neighborhood-based collaborative filtering , 2012, CIKM.

[17]  Diego Fernández,et al.  Comparison of collaborative filtering algorithms , 2011, ACM Trans. Web.

[18]  Licia Capra,et al.  Private distributed collaborative filtering using estimated concordance measures , 2007, RecSys '07.

[19]  Martha Larson,et al.  Exploiting user similarity based on rated-item pools for improved user-based collaborative filtering , 2009, RecSys '09.

[20]  Michael R. Lyu,et al.  Effective missing data prediction for collaborative filtering , 2007, SIGIR.

[21]  Sahin Albayrak,et al.  Analyzing weighting schemes in collaborative filtering: cold start, post cold start and power users , 2012, SAC '12.

[22]  Naren Ramakrishnan,et al.  Scouts, promoters, and connectors: The roles of ratings in nearest-neighbor collaborative filtering , 2007, TWEB.

[23]  Yehuda Koren,et al.  Factor in the neighbors: Scalable and accurate collaborative filtering , 2010, TKDD.

[24]  Fernando Ortega,et al.  A collaborative filtering similarity measure based on singularities , 2012, Inf. Process. Manag..

[25]  Fernando Ortega,et al.  Improving collaborative filtering-based recommender systems results using Pareto dominance , 2013, Inf. Sci..

[26]  Congfu Xu,et al.  Adaptive Bayesian personalized ranking for heterogeneous implicit feedbacks , 2015, Knowl. Based Syst..

[27]  Hyung Jun Ahn,et al.  A new similarity measure for collaborative filtering to alleviate the new user cold-starting problem , 2008, Inf. Sci..

[28]  Françoise Fessant,et al.  Designing Specific Weighted Similarity Measures to Improve Collaborative Filtering Systems , 2008, ICDM.

[29]  Licia Capra,et al.  The effect of correlation coefficients on communities of recommenders , 2008, SAC '08.

[30]  Martha Larson,et al.  Mining contextual movie similarity with matrix factorization for context-aware recommendation , 2013, TIST.

[31]  Daniel Thalmann,et al.  A simple but effective method to incorporate trusted neighbors in recommender systems , 2012, UMAP.

[32]  Chao Liu,et al.  Recommender systems with social regularization , 2011, WSDM '11.