Research map / Label-efficient and robust learning / Recognition and long-tail
The Minority Matters: A Diversity-Promoting Collaborative Metric Learning Algorithm
Shilong Bao, Qianqian Xu, Zhiyong Yang, Yuan He, Xiaochun Cao, Qingming Huang
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (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 |
|---|---|---|---|
| add DCRS regularization (ablation) | DPCML2 · MovieLens-1m | MaxDiv@20: 292.089 -> 313.086 | Improved |
| add DCRS regularization (ablation) | DPCML1 · MovieLens-1m | MaxDiv@20: 140.057 -> 145.615 | Improved |
| add DCRS regularization (ablation) | DPCML1 · Steam-200k | MaxDiv@20: 155.193 -> 179.065 | Improved |
| add DCRS regularization (ablation) | DPCML2 · Steam-200k | MaxDiv@20: 365.458 -> 369.876 | Improved |
| replace single user vector with multiple user representations (DPCML2 vs HarS, same hard sampling) | CML · MovieLens-1m | NDCG@5: 25.77 -> 26.16 | Improved |
| replace single user vector with multiple user representations (DPCML2 vs HarS) | CML · Steam-200k | NDCG@5: 26.63 -> 30.56 | Improved |
| replace single user vector with multiple user representations (DPCML1 vs UniS) | CML · Steam-200k | MaxDiv@20: 117.927 -> 179.065 | Improved |
| replace single user vector with multiple user representations (DPCML2 vs HarS) | CML · MovieLens-1m | MaxDiv@20: 244.078 -> 313.086 | Improved |
| use too many user vectors C with hard negative sampling | DPCML2 · Steam-200k | accuracy metrics | Hurt |
| increase number of user vectors C (C>1 vs C=1) | DPCML · Steam-200k | training loss and test performance | Improved |
| tune diversity weight eta to about 10 (ablation) | DPCML · Steam-200k | recommendation metrics | Improved |
| use both DCRS terms instead of only delta1 or delta2 (ablation) | DPCML · Steam-200k | recommendation metrics | Improved |
Closest papers by meaning
- Improved Diversity-Promoting Collaborative Metric Learning for Recommendation (TPAMI 2024 · similarity 0.98)
- Rethinking Collaborative Metric Learning: Toward an Efficient Alternative Without Negative Sampling (TPAMI 2022 · similarity 0.89)
- ImplicitSLIM and How it Improves Embedding-based Collaborative Filtering (ICLR 2024 · similarity 0.85)
- Tree of Preferences for Diversified Recommendation (NeurIPS 2025 · similarity 0.85)
- On the Embedding Collapse when Scaling up Recommendation Models (ICML 2024 · similarity 0.85)
- DivGCL: A Graph Contrastive Learning Model for Diverse Recommendation (AAAI 2025 · similarity 0.84)
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.
