Next Item Recommendation with Self-Attention
arXiv:1808.06414
Abstract
In this paper, we propose a novel sequence-aware recommendation model. Our model utilizes self-attention mechanism to infer the item-item relationship from user's historical interactions. With self-attention, it is able to estimate the relative weights of each item in user interaction trajectories to learn better representations for user's transient interests. The model is finally trained in a metric learning framework, taking both short-term and long-term intentions into consideration. Experiments on a wide range of datasets on different domains demonstrate that our approach outperforms the state-of-the-art by a wide margin.
arXiv comment10 pages
References in corpus (11)
- BPR: Bayesian Personalized Ranking from Implicit Feedback
- Convolutional Sequence to Sequence Learning
- NAIS: Neural Attentive Item Similarity Model for Recommendation
- Session-based Recommendations with Recurrent Neural Networks
- Translation-based Recommendation
- Fusing Similarity Models with Markov Chains for Sparse Sequential Recommendation
- Outer Product-based Neural Collaborative Filtering
- Metric Factorization: Recommendation beyond Matrix Factorization
- NeuRec: On Nonlinear Transformation for Personalized Ranking
- CoupleNet: Paying Attention to Couples with Coupled Attention for Relationship Recommendation
- Sequence-Aware Recommender Systems
Cited by in corpus (14)
- Lifelong Sequential Modeling with Personalized Memorization for User Response Prediction
- SDM: Sequential Deep Matching Model for Online Large-scale Recommender System
- SANST: A Self-Attentive Network for Next Point-of-Interest Recommendation
- Dynamic Memory based Attention Network for Sequential Recommendation
- DeepRec: An Open-source Toolkit for Deep Learning based Recommendation
- Future Data Helps Training: Modeling Future Contexts for Session-based Recommendation
- Rethinking Lifelong Sequential Recommendation with Incremental Multi-Interest Attention
- Superbloom: Bloom filter meets Transformer
- Multi-Interest-Aware User Modeling for Large-Scale Sequential Recommendations
- Quaternion Collaborative Filtering for Recommendation
- Path-Based Reasoning over Heterogeneous Networks for Recommendation via Bidirectional Modeling
- Rethinking Item Importance in Session-based Recommendation
- Sequential Modelling with Applications to Music Recommendation, Fact-Checking, and Speed Reading
- TRec: Sequential Recommender Based On Latent Item Trend Information