AI Paper Atlas

Research map / Graph learning / Graph self-supervised learning

Uncovering Semantic Hierarchies in Text-Attributed Graphs via Variational EM-based LLM–GNN Synergy

Yunhui Liu, Xudong Jin, Qizhuo Xie, Chunhui Zhao, Xiao Luo, Tao Zheng, Bin Chong, Tieke He

NeurIPS 2026 · Graph learning › Graph self-supervised learning · approach: LLMs and graph foundation models · 0 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted summary. Written by AI from the paper's full text. Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Proposes SHiFT, a variational EM framework that couples LLM reasoning with GNN structure encoding to uncover coarse-to-fine semantic hierarchies in text-attributed graphs.

Problem

Representation learning on text-attributed graphs typically uses flat semantic spaces and ignores the coarse-to-fine hierarchical granularity present in real-world data.

Approach

Synergizes LLM reasoning with GNN structural representations inside an Expectation-Maximization (variational EM) loop to infer hierarchical semantic structure over text-attributed graph nodes, rather than a flat semantic space.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
add iterative EM refinement of the taxonomy (vs one-pass initial taxonomy) (ablation)SHiFT · Cora, Citeseer, WikiCSaccuracy: avg drop 1.27 without EM; Cora 84.83 -> 83.27Improved
add taxonomy partition alignment (TPA) (ablation)SHiFT · Cora, Citeseer, WikiCSaccuracy: avg drop 1.24 without TPAImproved
add taxonomy skeleton alignment (TSA) (ablation)SHiFT · Cora, Citeseer, WikiCSaccuracy: avg drop 1.39 without TSA; Citeseer 73.96 -> 72.31Improved
add taxonomy concept alignment (TCA) (ablation)SHiFT · Cora, Citeseer, WikiCSaccuracy: avg drop 0.48 without TCA; Cora 84.83 -> 84.14Improved
induce taxonomy with graph structure instead of PLM embeddings only (ablation)SHiFT taxonomy · Cora, Citeseer, WikiCSNPMI: Cora 0.1640 -> 0.1958Improved
EM refinement of taxonomy (T_init to T_final) (ablation)SHiFT taxonomy · Cora human evaluationHierarchy Validity: 3.65 -> 4.10Improved
append SHiFT-induced taxonomy path to raw text features (ablation)MLP classifier · Coraaccuracy: 76.05 -> 81.91Improved
append SHiFT-induced taxonomy path to raw text features (ablation)GCN classifier · Coraaccuracy: 81.70 -> 83.41Improved
append path from graph-agnostic taxonomy to raw text features (ablation)MLP classifier · Coraaccuracy: 76.05 -> 72.68Hurt
explicit semantic taxonomy instead of hyperbolic implicit hierarchySHiFT vs HGCN/HGCL · node classification avgaccuracy: 77.26 -> 81.40Improved
Mapping & Diagnosis incremental updates instead of full regenerationSHiFT E-step · later epochsLLM invocations: reduces by ~80%Improved

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