Research map / Image generation / Fast diffusion sampling
Modeling Item-Level Dynamic Variability with Residual Diffusion for Bundle Recommendation
Dong Zhang, Lin Li, Ming Li, Amran Bhuiyan, Meng Sun, Xiaohui Tao, Jimmy 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 residual diffusion (RDiffBR) to six BR backbones under item removal | BR backbones · Youshu rho=-4 | Recall@20: BGCN .1897 -> .2178 | Improved |
| add RDiffBR when bundle-item affiliations are added (rho>0) | BR backbones · iFashion/others | Recall/NDCG@20: up to 24.83% | Improved |
| add RDiffBR when affiliations unchanged (rho=0) | BR backbones · Youshu | Recall@20: CrossCBR .2776 -> .2827 | Improved |
| add RDiffBR on already robust backbone | MultiCBR · Youshu rho=0 | Recall@20: MultiCBR .2791 -> .2824 | Improved |
| remove residual connection from approximator (ablation) | RDiffBR on BundleGT · Youshu rho=3 | Recall@20: BundleGT+RDiffBR .2925 -> w/o R .2519 | Hurt |
| remove residual connection from approximator (ablation) | RDiffBR on CrossCBR · MealRec+ rho=3 | Recall@20: CrossCBR+RDiffBR .2474 -> w/o R .1843 | Hurt |
| remove residual connection from approximator (ablation) | RDiffBR on BGCN · Youshu rho=3 | Recall@20: BGCN+RDiffBR .2394 -> w/o R .2337 | Hurt |
| use too large loss balance factor lambda (ablation) | RDiffBR · bundle datasets | Recall/NDCG | Hurt |
| use too high or too low residual weight delta (ablation) | RDiffBR · bundle datasets | Recall/NDCG | Hurt |
| remove items from bundles at test time (rho<0) | HyperMBR (no RDiffBR) · Youshu | Recall@20: rho=0 .2649 -> rho=-4 .1464 | Hurt |
| add RDiffBR to BGCN training | BGCN · Youshu training time | one-epoch time overhead: extra cost 3.42% | No clear effect |
Closest papers by meaning
- Disentangled Contrastive Bundle Recommendation with Conditional Diffusion (AAAI 2025 · similarity 0.87)
- Learning Multiple User Distributions for Recommendation via Guided Conditional Diffusion (AAAI 2025 · similarity 0.84)
- Enhancing Diffusion Model with Auxiliary Information Mining-Exploration and Efficient Sampling Mechanism for Sequential Recommendation (AAAI 2025 · similarity 0.83)
- Enhancing Sequential Recommendation with Global Diffusion (AAAI 2025 · similarity 0.83)
- Addressing Cold-Start Problem in Click-Through Rate Prediction via Supervised Diffusion Modeling (AAAI 2025 · similarity 0.82)
- On Efficiency-Effectiveness Trade-off of Diffusion-based Recommenders (NeurIPS 2025 · similarity 0.82)
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.
