AI Paper Atlas

Research map / Label-efficient and robust learning / Recognition and long-tail

DRSoRec: Dual-Rectification of Social Networks for Recommendation

Liangxun Yang, Tianzi Zang, Jiayi Sun, Juan Li, Yicong Li

AAAI 2026 · Label-efficient and robust learning › Recognition and long-tail · approach: Recommendation representation learning · 4 citations (OpenAlex)

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 changedWhereMeasuredEffect
add invariant social rationale discovery module (ablation)DRSoRec · CiaoRecall@10: 0.4775 -> 0.5046Improved
add adaptive social connection refinement module (ablation)DRSoRec · CiaoRecall@10: 0.4796 -> 0.5046Improved
add social-knowledge contrastive loss (ablation)DRSoRec · CiaoRecall@10: 0.4963 -> 0.5046Improved
add adaptive multi-view fusion instead of average pooling (ablation)DRSoRec · YelpRecall@10: 0.7248 -> 0.7278Improved
add rationale module on Yelp (ablation)DRSoRec · YelpRecall@10: 0.7067 -> 0.7278Improved
add social network denoising/refinement over DiffNet-style indiscriminate social tiessocial recommendation · DeliciousRecall@10: DiffNet++ 0.3071 vs DRSoRec 0.3485Improved
add indiscriminate social ties (DiffNet++) vs no social info (LightGCN)LightGCN · DeliciousRecall@10: 0.3290 -> 0.3071Hurt
add reconstruction loss weight (lambda1 = 0.1 vs 0) (ablation)DRSoRec · YelpNDCG@10: +1.32%Improved
increase hyperparameters (lambda, tau, layers) beyond optimum (ablation)DRSoRec · YelpRecall@10/NDCG@10Hurt

Closest papers by meaning

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.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy