A Differentiable Ranking Metric Using Relaxed Sorting Opeartion for Top-K Recommender Systems.
Hyunsung LeeYeongjae JangJaekwang KimHonguk WooPublished in: CoRR (2020)
Keyphrases
- recommender systems
- ranking list
- implicit feedback
- collaborative filtering
- active user
- preference learning
- user feedback
- keyword search in relational databases
- learning to rank
- scoring function
- evaluation metrics
- query processing
- result set
- scoring functions
- matrix factorization
- ranking algorithm
- web search
- objective function
- loss function
- user preferences
- trust propagation
- cold start problem
- ranked list
- metric space
- ranking functions
- optimal solution
- link analysis
- user defined
- personalized search
- user profiles
- distance measure
- rating prediction
- database systems
- trust aware
- rank correlation
- recommendation quality
- user behavior
- information overload
- evaluation measures