Research map / Video / Video-language understanding
ProRec-Video: Guiding Hierarchical Interest Transitions for Proactive Short Video Recommendation with Dynamic Feedback Adaptation
Weizhi Chen, Baoyun Peng, Bo Liu, Xingkong Ma, Houjie Qiu
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 both positive and negative feedback reflexion (ablation) | ProRec-Video · KuaiRec-Small (interest-level) | Success Rate: 0.6634 -> 0.6860 | Improved |
| add both positive and negative feedback reflexion (ablation) | ProRec-Video · MicroLens-50K (interest-level) | Success Rate: 0.6939 -> 0.8500 | Improved |
| add only positive feedback (ablation) | ProRec-Video · KuaiRec-Small (interest-level) | Success Rate: 0.6634 -> 0.6000 | Hurt |
| add only negative feedback (ablation) | ProRec-Video · KuaiRec-Small (interest-level) | Success Rate: 0.6634 -> 0.6207 | Hurt |
| add only one feedback type (ablation) | ProRec-Video · MicroLens-50K (interest-level) | Success Rate: 0.6939 -> 0.7500 (positive), 0.7832 (negative) | Improved |
| split planning into two-step LLM pipeline (path then video) instead of single step (ablation) | ProRec-Video · KuaiRec-Small and MicroLens-50K | all five metrics | Improved |
| use larger / stronger LLM as planner (GPT-4 vs 7B-34B open models) (ablation) | ProRec-Video · KuaiRec-Small and MicroLens-50K | all five metrics | Improved |
| use ProRec-Video instead of Pf2Inf-Dijkstra at item level | item-level proactive recommendation · KuaiRec-Small | Success Rate: 0.5609 -> 0.4561 | Hurt |
| use interest-level planning with receptivity profiling vs best baselines | proactive recommendation · KuaiRec-Small and MicroLens-50K (interest-level) | Success Rate: 0.5780 -> 0.6860; 0.6988 -> 0.8500 | Improved |
| use ProRec-Video for resistant users | user receptivity groups · KuaiRec-Small resistant group | Success Rate: 0.2979 -> 0.3333 | Improved |
Closest papers by meaning
- Exploiting Fine-Grained Skip Behaviors for Micro-Video Recommendation (AAAI 2025 · similarity 0.86)
- MTRec: Learning to Align with User Preferences via Mental Reward Models (NeurIPS 2025 · similarity 0.86)
- UxSID: Semantic-Aware User Interests Modeling for Ultra-Long Sequence (NeurIPS 2026 · similarity 0.85)
- Plug-In Diffusion Model for Sequential Recommendation (AAAI 2024 · similarity 0.84)
- Ready for You When You Are Back: Content-Driven Session-Based Recommendation for Continuity of Experience (AAAI 2025 · similarity 0.84)
- Calibrating Video Watch-time Predictions with Credible Prototype Alignment (ICML 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.
