Research map / Label-efficient and robust learning / Recognition and long-tail
Tenrec: A Large-scale Multipurpose Benchmark Dataset 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 |
|---|---|---|---|
| pre-train SBR model on QK-video before fine-tuning on QB-video (ablation) | SASRec · QB-video | HR@20: 0.11715 -> 0.12612 | Helped |
| pre-train SBR model on QK-video before fine-tuning on QB-video (ablation) | NextItNet · QB-video | HR@20: 0.11922 -> 0.12291 | Helped |
| add self-supervised pre-training on QK-video (ablation) | BERT4Rec · cold-start QK-article | HR@20: 0.03555 -> 0.04963 | Helped |
| add pre-training (ablation) | PeterRec · user profile age prediction | accuracy: 0.68671 -> 0.69712 | Helped |
| use multi-task MMOE instead of single objective optimization | click and like prediction · QK-video-1M | click-AUC: 0.7957 -> 0.7900 | No clear effect |
| use ESMM instead of MMOE | MTL · QK-video-1M | click-AUC: 0.7900 -> 0.7940 | Helped |
| use bidirectional BERT4Rec instead of unidirectional models | session-based rec · QK-video-1M | HR@20: 0.05112 -> 0.04902 | Hurt |
| compress with CpRec | NextItNet · QK-video-1M | HR@20: 0.05112 -> 0.05001 | Hurt |
| train with StackRec | NextItNet · QK-video-1M | training time: 66880s -> 24320s | Helped |
| use SkipRec adaptive layer skipping | SASRec · QB-video | inference time: 305s -> 206s; HR@20 0.11715 -> 0.11086 | Mixed |
| use shared historical item embeddings instead of separate embeddings (ablation) | CTR models · QK-video-1M | AUC: NFM 0.7957 -> 0.7924 | Hurt |
| add Conure lifelong pre-training from past tasks (ablation) | Conure-NextItNet · task 4 (QB-article) | NDCG@20: 0.0931 -> 0.1074 | Helped |
Closest papers by meaning
- RecFlow: An Industrial Full Flow Recommendation Dataset (ICLR 2025 · similarity 0.88)
- ORBIT - Open Recommendation Benchmark for Reproducible Research with Hidden Tests (NeurIPS 2025 · similarity 0.88)
- AgentRecBench: Benchmarking LLM Agent-based Personalized Recommender Systems (NeurIPS 2025 · similarity 0.87)
- Can LLMs Outshine Conventional Recommenders? A Comparative Evaluation (NeurIPS 2025 · similarity 0.87)
- Would You Like Your Data to Be Trained? A User Controllable Recommendation Framework (AAAI 2024 · similarity 0.87)
- MTRec: Learning to Align with User Preferences via Mental Reward Models (NeurIPS 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.