AI Paper Atlas

Research map / Graph learning / Graph self-supervised learning

GraphCleaner: Detecting Mislabelled Samples in Popular Graph Learning Benchmarks

Yuwen Li, Miao Xiong, Bryan Hooi

ICML 2023 · Graph learning › Graph self-supervised learning · approach: Graph anomaly detection · 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 neighborhood-aware detection with synthetic mislabels generated from a learned class-transition matrixGraphCleaner · Cora, 10% symmetric noiseF1: 0.560 -> 0.790Improved
Use neighborhood-aware detection with synthetic mislabels generated from a learned class-transition matrixGraphCleaner · CiteSeer, 10% symmetric noiseF1: 0.432 -> 0.560Improved
Use neighborhood-aware detection with synthetic mislabels generated from a learned class-transition matrixGraphCleaner · PubMed, 10% symmetric noiseF1: 0.447 -> 0.616Improved
Use neighborhood-aware detection with synthetic mislabels generated from a learned class-transition matrixGraphCleaner · Computers, 10% symmetric noiseF1: 0.755 -> 0.844Improved
Use neighborhood-aware detection with synthetic mislabels generated from a learned class-transition matrixGraphCleaner · Photo, 10% symmetric noiseF1: 0.777 -> 0.902Improved
Use neighborhood-aware detection with synthetic mislabels generated from a learned class-transition matrixGraphCleaner · OGB-arxiv, 10% symmetric noiseF1: 0.500 -> 0.760Improved
Remove neighborhood-label agreement features (ablation)GraphCleaner · Cora, Computers, and OGB-arxiv, 10% symmetric noiseF1: Cora: 0.798 -> 0.790; Computers: 0.840 -> 0.844; OGB-arxiv: 0.750 -> 0.760Mixed
Remove neighborhood-prediction agreement features (ablation)GraphCleaner · Cora, Computers, and OGB-arxiv, 10% symmetric noiseF1: Cora: 0.676 -> 0.790; Computers: 0.844 -> 0.844; OGB-arxiv: 0.736 -> 0.760Mixed
Replace transition-matrix-guided synthetic noise with randomly selected incorrect labels (ablation)GraphCleaner · 10% symmetric noise, Cora, Computers, and OGB-arxivF1: Cora: 0.802 -> 0.790; Computers: 0.865 -> 0.844; OGB-arxiv: 0.819 -> 0.760Mixed
Replace transition-matrix-guided synthetic noise with randomly selected incorrect labels (ablation)GraphCleaner · 10% asymmetric noise, Cora, Computers, and OGB-arxivF1: Cora: 0.635 -> 0.669; Computers: 0.474 -> 0.778; OGB-arxiv: 0.295 -> 0.477Mixed
Remove noisy-labelled samples from the PubMed test setGCN+mixup · PubMed test setaccuracy: 86.71% -> 89.11%Improved
Use one-hop neighborhood information (ablation)GraphCleaner · OGB-arxiv, 10% noiseF1No 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