AI Paper Atlas

Research map / Graph learning / GNN architectures

Evaluating Graph Neural Networks for Link Prediction: Current Pitfalls and New Benchmarking

Juanhui Li, Harry Shomer, Haitao Mao, Shenglai Zeng, Yao Ma, Neil Shah, Jiliang Tang, Dawei Yin

NeurIPS 2023 · Graph learning › GNN architectures · approach: Graph structure learning · 25 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
Replace global random evaluation negatives with per-positive corruptions selected using RA, PPR, and feature similarityLink prediction evaluation · Cora, Citeseer, Pubmed, ogbl-collab, ogbl-ddi, ogbl-ppa, ogbl-citation2mean model standard deviation: 5.19 -> 0.79; 5.94 -> 0.88; 4.14 -> 0.35; 1.49 -> 0.96; 2.13 -> 0.36; 6.77 -> 3.49; 1.39 -> 0.59Improved
Use heuristic-selected, personalized hard negatives instead of existing randomly sampled negativesLink prediction evaluation · Most datasets; especially Cora, Citeseer, Pubmed, and ogbl-citation2MRRHurt
Use HeaRT negatives rather than existing negativesLink prediction evaluation · ogbl-ppaMRRMixed
Convert ogbl-citation2 to an undirected graph when computing heuristic scoresHeuristic methods · ogbl-citation2MRR: around 50% -> around 75%Improved
Include validation edges in the graph during testingNeo-GNN · ogbl-collabHits@50: 57.52 -> 66.13Improved
Train BUDDY with one negative per positive instead of sixBUDDY · ogbl-ddiHits@20: 78.51 -> 29.60Hurt
Remove pretraining of the GNN used to initialize Neo-GNNNeo-GNN · ogbl-ddiHits@20: 63.57 -> 20.95Hurt
Use ranking-based metrics rather than relying on AUC aloneLink prediction evaluation · ogbl-ppaHits@K: 0; >99%Mixed

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