Research map / Learning theory and algorithms / Prediction and decision losses
Joint Optimization of Tree-based Index and Deep Model for Recommender Systems
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 |
|---|---|---|---|
| Jointly optimize tree index and preference model with hierarchical user representation instead of using the standalone DNN | DNN recommender · Amazon Books | Recall@200: 8.57% -> 12.45% | Helped |
| Jointly optimize tree index and preference model with hierarchical user representation instead of using the standalone DNN | DNN recommender · UserBehavior | Recall@200: 13.45% -> 14.71% | Helped |
| Replace inner-product preference scoring with a fully connected network | YouTube product-DNN · Amazon Books | Recall@200: 8.26% -> 8.57% | Helped |
| Replace inner-product preference scoring with a fully connected network | YouTube product-DNN · UserBehavior | Recall@200: 10.15% -> 13.45% | Helped |
| Add hierarchical user representation (ablation) | JTM-J · Amazon Books | Recall@200: 7.60% -> 12.45% | Helped |
| Add hierarchical user representation (ablation) | JTM-J · UserBehavior | Recall@200: 11.72% -> 14.71% | Helped |
| Learn the tree instead of keeping the fixed initial tree (ablation) | JTM-H · Amazon Books | Recall@200: 10.45% -> 12.45% | Helped |
| Learn the tree instead of keeping the fixed initial tree (ablation) | JTM-H · UserBehavior | Recall@200: 12.93% -> 14.71% | Helped |
| Replace the category-initialized tree with joint tree learning (ablation) | TDM · UserBehavior | Recall@200: +0.88% | Helped |
| Use hierarchical user representation (ablation) | TDM · UserBehavior | Recall@200: +2.09% | Helped |
| Replace Item-CF with JTM for candidate generation | Taobao App Homepage display advertising · Online traffic | CTR: +11.3% | Helped |
| Replace Item-CF with JTM for candidate generation | Taobao App Homepage display advertising · Online traffic | RPM: +12.9% | Helped |
Closest papers by meaning
- Actions Speak Louder than Words: Trillion-Parameter Sequential Transducers for Generative Recommendations (ICML 2024 · similarity 0.84)
- Can LLMs Outshine Conventional Recommenders? A Comparative Evaluation (NeurIPS 2025 · similarity 0.84)
- On Softmax Direct Preference Optimization for Recommendation (NeurIPS 2024 · similarity 0.84)
- Generalization Error Bounds for Two-stage Recommender Systems with Tree Structure (NeurIPS 2024 · similarity 0.83)
- SHARP-Distill: A 68× Faster Recommender System with Hypergraph Neural Networks and Language Models (ICML 2025 · similarity 0.83)
- Preference Diffusion for Recommendation (ICLR 2025 · similarity 0.83)
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.