Lighter and Better: Low-Rank Decomposed Self-Attention Networks for Next-Item Recommendation.
Xinyan FanZheng LiuJianxun LianWayne Xin ZhaoXing XieJi-Rong WenPublished in: SIGIR (2021)
Keyphrases
- low rank
- matrix factorization
- item recommendation
- collaborative filtering
- matrix completion
- low rank matrix
- missing data
- linear combination
- rank minimization
- convex optimization
- recommender systems
- negative matrix factorization
- nonnegative matrix factorization
- tensor factorization
- high dimensional data
- high order
- social networks
- complex networks
- singular value decomposition
- machine learning
- implicit feedback
- factorization method
- nearest neighbor
- pattern recognition