AI Paper Atlas

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

MACRec: A Multi-View Subspace Alignment Framework for Contrastive Sampling Calibration in Recommendation

Junping Liu, Mingchao Yu, Xinrong Hu, Rui Yan, Wanqing Li, Jie Yang, Yi Guo

AAAI 2026 · Label-efficient and robust learning › Recognition and long-tail · approach: Uncertainty-aware multi-view learning · 0 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 MACRec sampling calibrationSGL · Amazon-BookRecall@10: 9.94 -> 10.62Improved
add MACRec sampling calibrationSGL · Amazon-MovieRecall@10: 12.39 -> 13.70Improved
add MACRec sampling calibrationDCCF · Amazon-BookRecall@10: 10.19 -> 10.51Improved
add MACRec sampling calibrationIRLLRec · Amazon-MovieRecall@10: 13.41 -> 13.92Improved
add MACRec sampling calibrationSimGCL · YelpRecall@10: 7.72 -> 7.95Improved
add calibration-based contrastive reweighting (CCR) (ablation)MACRec · three datasets, SGL/DCCF/IRLLRecRecall@10, NDCG@10Improved
add multi-view affinity (MVA) or cross-subspace alignment (CSA) (ablation)MACRec · three datasetsRecall@10, NDCG@10Improved
use subspace affinity instead of cosine similarity (ablation)MACRec · three datasetsRecall@10Improved
use calibrated reweighting instead of hard removal or relabeling as positives (ablation)MACRec · three datasetsRecall@10Improved
increase number of augmented views up to five (ablation)SGL+ · three datasetsRecall@10Mixed
low-rank instead of dense coefficient matrix (ablation)DCCF+ · Amazon-BookRecall@10: 10.67 -> 10.51 (time 302 -> 160)No clear effect
low-rank with sparsity regularization instead of no sparsity (ablation)DCCF+ · Amazon-BookRecall@10: 10.03 -> 10.51Improved

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