AI Paper Atlas

Research map / Graph learning / Graph self-supervised learning

Multi-graph Fusion Cross-model Contrastive Learning for Recommendation

Shengjun Ma, Yuhai Zhao, Fenglong Ma, Baoyin Liu, Zhengkui Wang, Wen Shan

AAAI 2026 · Graph learning › Graph self-supervised learning · approach: Contrastive learning with negatives and views · 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 cross-model contrastive learning (CMC) (ablation)MFCCL · Alibaba-iFashionrecall@20: 0.1094 -> 0.1288Improved
add cross-model contrastive learning (CMC) (ablation)MFCCL · Last-FMrecall@20: 0.0883 -> 0.1025Improved
use virtual relations instead of original KG relations (ablation)MFCCL · Alibaba-iFashionndcg@20: 7.21% relativeImproved
use virtual relations instead of original KG relations (ablation)MFCCL · Last-FMndcg@20: 0.0897 -> 0.0978Improved
add user-interest graph (UiG) (ablation)MFCCL · Alibaba-iFashionrecall@20: 0.1276 -> 0.1288Improved
add bipartite graph for user representation (ablation)MFCCL · Alibaba-iFashionrecall@20: 0.1264 -> 0.1288Improved
add item-association graph (IaG) (ablation)MFCCL · Last-FMrecall@20: 0.1012 -> 0.1025Improved
add KG for item representation (ablation)MFCCL · Last-FMrecall@20: 0.1012 -> 0.1025Improved
MFCCL over best baselineMFCCL · Alibaba-iFashionrecall@20: 0.1196 -> 0.1288Improved
MFCCL over best baselineMFCCL · Yelp2018ndcg@20: 0.0604 -> 0.0501Hurt
more virtual relations than moderate value (ablation)MFCCL · Alibaba-iFashion, Last-FMrecall@20, ndcg@20Hurt
set CL loss coefficient to 0.0001 (ablation)MFCCL · three datasetsrecall@20Improved

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