Research map / Label-efficient and robust learning / Recognition and long-tail
ATRank: An Attention-Based User Behavior Modeling Framework for Recommendation
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| Replace sequence encoders with multi-space self-attention behavior modeling | ATRank · Amazon Electro. | AUC: 0.8804 -> 0.8921 | Helped |
| Replace sequence encoders with multi-space self-attention behavior modeling | ATRank · Amazon Clothing. | AUC: 0.7869 -> 0.7905 | Helped |
| Add vanilla attention on top of Bi-LSTM | Bi-LSTM · Amazon Electro. | AUC: 0.8757 -> 0.8769 | Helped |
| Add vanilla attention on top of Bi-LSTM | Bi-LSTM · Amazon Clothing. | AUC: 0.7869 -> 0.7835 | Hurt |
| Add vanilla attention on top of Bi-LSTM | Bi-LSTM · Ali Multi Behavior, item prediction | AUC: 0.6779 -> 0.6754 | Hurt |
| Add vanilla attention on top of Bi-LSTM | Bi-LSTM · Ali Multi Behavior, query prediction | AUC: 0.6019 -> 0.5999 | Hurt |
| Add vanilla attention on top of Bi-LSTM | Bi-LSTM · Ali Multi Behavior, coupon prediction | AUC: 0.8500 -> 0.8413 | Hurt |
| Use all behavior types as history instead of only the target behavior type | ATRank · Ali Multi Behavior, item prediction | AUC: 0.6785 -> 0.6825 | Helped |
| Use all behavior types as history instead of only the target behavior type | ATRank · Ali Multi Behavior, query prediction | AUC: 0.6132 -> 0.6297 | Helped |
| Use all behavior types as history instead of only the target behavior type | ATRank · Ali Multi Behavior, coupon prediction | AUC: 0.8601 -> 0.8725 | Helped |
| Jointly train one model to predict all behavior types | ATRank · Ali Multi Behavior, item prediction | AUC: 0.6825 -> 0.6759 | Hurt |
| Jointly train one model to predict all behavior types | ATRank · Ali Multi Behavior, query prediction | AUC: 0.6297 -> 0.6199 | Hurt |
Closest papers by meaning
- UxSID: Semantic-Aware User Interests Modeling for Ultra-Long Sequence (NeurIPS 2026 · similarity 0.88)
- Long-Sequence Recommendation Models Need Decoupled Embeddings (ICLR 2025 · similarity 0.87)
- Would You Like Your Data to Be Trained? A User Controllable Recommendation Framework (AAAI 2024 · similarity 0.87)
- Customizing Language Models with Instance-wise LoRA for Sequential Recommendation (NeurIPS 2024 · similarity 0.86)
- LIME: Link-based User-item Interaction Modeling with Decoupled XOR Attention for Efficient Test Time Scaling (NeurIPS 2026 · similarity 0.86)
- BeFA: A General Behavior-driven Feature Adapter for Multimedia Recommendation (AAAI 2025 · similarity 0.86)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.