AI Paper Atlas

Research map / Graph learning / Graph self-supervised learning

Forest-Based Graph Learning for Semi-Supervised Node Classification

Jin Li, Shenghao Gao, Kaichen Zhang, Xinlong Chen, Ying Sun, Hui Xiong

ICLR 2026 · Graph learning › Graph self-supervised learning · approach: Semi-supervised node representation 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
add global tree-aggregation submodule (ablation)FGL · Coraaccuracy (%): 80.00 -> 85.46Improved
add global tree-aggregation submodule (ablation)FGL · Wisconsinaccuracy (%): 83.92 -> 86.27Improved
add local submodule (ablation)FGL · Texasaccuracy (%): 69.93 -> 91.89Improved
add local submodule (ablation)FGL · Flickraccuracy (%): 32.17 -> 47.22Improved
use homophily-guided tree sampling instead of uniform (ablation)FGL · Coraaccuracy (%): 83.63 -> 85.46Improved
sample a forest instead of a single homophily-guided tree (ablation)FGL · Pubmedaccuracy (%): 78.55 -> 81.00Improved
sample a forest instead of a single tree (ablation)FGL · OGBN-Arxivaccuracy (%): 55.17 -> 56.47Improved
use one homophily-guided tree instead of uniform random trees (ablation)FGL · Cornellaccuracy (%): 72.97 -> 76.35Improved
use two-stage homophily estimator instead of naive attention estimator (ablation)FGL · Pubmedaccuracy (%): 75.10 -> 81.00Improved
use homophily-guided sampling instead of uniform in estimator study (ablation)FGL · Coraaccuracy (%): 78.40 -> 85.46Improved
increase number of trees beyond 6-10 (ablation)FGL · Cora, Pubmed, Actoraccuracy (%)No clear effect
improve homophily estimator accuracy (ablation)FGL · Cora, Pubmed, Actoraccuracy (%)Improved

Earlier methods this paper compares with

Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.

Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.

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