MSBPR: A multi-pairwise preference and similarity based Bayesian personalized ranking method for recommendation.
Liang ZengJiewen GuanBilian ChenPublished in: Knowl. Based Syst. (2023)
Keyphrases
- pairwise
- user preferences
- personal preferences
- personalized recommendation
- recommender systems
- user modeling
- user profiles
- personalized information
- pairwise comparison
- e learning
- product recommendation
- information filtering
- web personalization
- pairwise comparisons
- similarity measure
- recommendation systems
- spectral clustering
- collaborative filtering
- belief propagation
- news recommendation
- higher order
- preference relations
- point sets
- loss function
- high order
- posterior distribution
- bayesian inference
- decision theory
- context aware
- data driven
- markov random field
- multi attribute
- semi supervised
- bayesian networks
- similarity function