AI Paper Atlas

Research map / Graph learning / Graph self-supervised learning

Fair Attribute Completion on Graph with Missing Attributes

Dongliang Guo, Zhixuan Chu, Sheng Li

ICLR 2023 · Graph learning › Graph self-supervised learning · approach: Heterophily and imbalance on graphs · 2 citations (OpenAlex)

Read the paper

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 FairAC attention-based attribute completion with adversarial fairness losses (vs. BaseAC without them) (ablation)GCN on Pokec-z · Pokec-z, alpha=0.3ΔSP+ΔEO: 2.35 -> 0.31Improved
Add adversarial fairness losses to attribute completion (BaseAC to FairAC) (ablation)GCN on Pokec-z · Pokec-z, alpha=0.8ΔSP+ΔEO: 2.28 -> 0.70Improved
Use FairAC completion with fairness losses instead of average-neighbor completionGCN · NBAΔSP+ΔEO: 6.16 -> 0.91Improved
Use FairAC instead of FairGNN on incomplete attributesGCN · Pokec-z, Pokec-n, NBAΔSP+ΔEO: Pokec-z 2.58 -> 0.31; Pokec-n 1.24 -> 0.41; NBA 2.58 -> 0.91Improved
Increase beta (fairness trade-off weight) from 0 to 0.8 (ablation)FairAC · Pokec-z, alpha=0.3accuracy: slight declineHurt
Increase beta from 0 to 0.8 (ablation)FairAC · Pokec-z, alpha=0.3ΔSP+ΔEOImproved
Integrate FairAC with GAT on Pokec-nGAT · Pokec-n, alpha=0.7ΔSP+ΔEO: 10.84 -> 4.37Improved

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