Research map / Label-efficient and robust learning / Recognition and long-tail
Exploiting both Vertical and Horizontal Dimensions of Feature Hierarchy for Effective Recommendation
Zhu Sun, Jie Yang, Jie Zhang, Alessandro Bozzon
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 horizontal complementary relationships to vertical-only model (HieV to HieVC) (ablation) | HieV · Clothing All Users | AUC: 0.7160 -> 0.7291 | Improved |
| add alternative relationships on top of complementary (HieVC to HieVH) (ablation) | HieVC · Clothing All Users | AUC: 0.7291 -> 0.7375 | Improved |
| add alternative relationships on top of complementary (HieVC to HieVH) (ablation) | HieVC · Electronics All Users | AUC: 0.7672 -> 0.7748 | Improved |
| use direct adaptation of item factors by feature factors (HieV) instead of ReMF regularization | ReMF · Clothing All Users | AUC: 0.7015 -> 0.7160 | Improved |
| add full HieVH over best baseline ReMF | ReMF · Home & Kitchen All Users | AUC: 0.7279 -> 0.7667 | Improved |
| add full HieVH over ReMF in cold start | ReMF · Clothing Cold Start | AUC: 0.7032 -> 0.7352 | Improved |
| increase horizontal regularization weight alpha (ablation) | HieVH · four Amazon datasets | AUC | Mixed |
| add horizontal relationships (ReMF to HieVH) for interpretability (ablation) | ReMF · Clothing Layer 1 | complementarity percentage Cp: 87.62% -> 95.65% | Improved |
| use flat feature models (CMF, FM) instead of hierarchy-aware models | FH-based methods · four Amazon datasets | AUC | Hurt |
Closest papers by meaning
- Tree of Preferences for Diversified Recommendation (NeurIPS 2025 · similarity 0.85)
- Language Representations Can be What Recommenders Need: Findings and Potentials (ICLR 2025 · similarity 0.85)
- Bridging the User-side Knowledge Gap in Knowledge-aware Recommendations with Large Language Models (AAAI 2025 · similarity 0.84)
- Temporally and Distributionally Robust Optimization for Cold-Start Recommendation (AAAI 2024 · similarity 0.84)
- Improving LLMs for Recommendation with Out-Of-Vocabulary Tokens (ICML 2025 · similarity 0.84)
- LIME: Link-based User-item Interaction Modeling with Decoupled XOR Attention for Efficient Test Time Scaling (NeurIPS 2026 · 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.
