AI Paper Atlas

Research map / Graph learning / Graph self-supervised learning

Mitigating Noise and Imbalance in Social Governance Graphs for Multi-Type Risk Assessment

Di Jin, Haotian Zhao, Xiaobao Wang, Fengyu Yan, Dongxiao He

AAAI 2026 · Graph learning › Graph self-supervised learning · approach: Heterophily and imbalance on graphs · 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
Use HeCoGNN instead of RGCNHeCoGNN · SG-I Person nodesMacro-F1: 0.6874 -> 0.7471Improved
Use HeCoGNN instead of RGCNHeCoGNN · SG-I Location nodesMacro-F1: 0.6898 -> 0.8058Improved
Use HeCoGNN instead of RGCNHeCoGNN · SG-I Person nodesAUC: 0.8600 -> 0.8452Hurt
Use HeCoGNN instead of MLPHeCoGNN · SG-II Event nodesMacro-F1: 0.6791 -> 0.6774No clear effect
Use HeCoGNN instead of RGCNHeCoGNN · SG-II Person nodesMacro-F1: 0.4962 -> 0.6310Improved
Use HeCoGNN instead of RGCNHeCoGNN · ACMMacro-F1: 0.8325 -> 0.8324No clear effect
Use HeCoGNN instead of RGCNHeCoGNN · ACMMicro-F1: 0.8305 -> 0.8416Improved
Add GateNet edge gating to ContextNet-only model (ablation)HeCoGNN · SG-IAccuracyImproved
Add ContextNet attention to GateNet-only model (ablation)HeCoGNN · SG-IAccuracyImproved
Use MLP gating plus dynamic attention instead of simple or fixed-weight variants (ablation)HeCoGNN · Social governance graphsMacro-F1/AUCImproved
Add graph structure via HGT/RGCN over MLP for event nodesMLP · Social risk case graph (event nodes)AccuracyNo clear effect

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