AI Paper Atlas

Research map / Graph learning / GNN architectures

Fair Graph Machine Learning under Adversarial Missingness Processes

Debolina Halder Lina, Arlei Silva

ICLR 2026 · Graph learning › GNN architectures · approach: Graph structure learning

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 BFtS worst-case 3-player imputation instead of RNF when no sensitive attributes are availablefair GNN node classification · BAIL, no sensitive infoΔDP (%): 11.65 -> 8.01Improved
BFtS vs RNF accuracy on NBA and Pokec-n without sensitive infoBFtS · NBA and Pokec-n, no sensitive infoF1 (%): NBA 70.2 -> 69.8; Pokec-n 72.2 -> 69.6Mixed
select low-degree nodes to have missing sensitive values (degree heuristic)independent GCN imputation · multiple graph datasetsbias underestimationHurt
targeted missingness heuristic vs degree heuristic (ablation)independent imputation · multiple graph datasetsbias discrepancyNo clear effect
BFtS worst-case imputation under adversarial missingnessBFtS · imputed vs true biasbias underestimationImproved
BFtS with higher assortativity vs low assortativity on synthetic graphbaselines FairGNN and Debias · Simulation datasetlabel learningHurt
BFtS vs Debias on running timeBFtS · six datasetsrunning time: German 21.74 (Debias) vs 28.91 (BFtS)Hurt
vary alpha vs beta hyperparameters (ablation)BFtS · sensitivity analysisfairness-accuracyMixed

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