Research map / Graph learning / Graph self-supervised learning
Diffusion Maps for Textual Network Embedding
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (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 changed | Where | Measured | Effect |
|---|---|---|---|
| Add diffusion convolution to text inputs (ablation) | DMTE · Cora link prediction, 95% edges | AUC: 96.7 -> 98.8 | Helped |
| Add diffusion convolution to text inputs (ablation) | DMTE · Zhihu link prediction, 95% edges | AUC: 75.1 -> 81.5 | Helped |
| Add structure embeddings alongside text embeddings (ablation) | DMTE · Cora link prediction, 95% edges | AUC: 94.2 -> 98.8 | Helped |
| Add structure embeddings alongside text embeddings (ablation) | DMTE · Zhihu link prediction, 95% edges | AUC: 74.1 -> 81.5 | Helped |
| Use Bi-LSTM rather than word-average text inputs (ablation) | DMTE · Cora link prediction, 95% edges | AUC: 98.8 -> 98.1 | Hurt |
| Use Bi-LSTM rather than word-average text inputs (ablation) | DMTE · Zhihu link prediction, 95% edges | AUC: 81.5 -> 82.2 | Helped |
| Increase diffusion hops from H=1; performance initially rises, then levels off for sufficiently large H (ablation) | DMTE (WAvg) · Link prediction, Cora and Zhihu | AUC | Mixed |
| Use global graph structure to inform text representations and the training objective | DMTE · DBLP multi-label classification | Macro-F1 | Helped |
| Use DMTE rather than CANE for link prediction | DMTE · Cora and Zhihu link prediction | AUC: +4.5 and 6.8 | Helped |
Closest papers by meaning
- Relation Rectification in Diffusion Model (CVPR 2024 · similarity 0.87)
- Cross-Domain Graph Data Scaling: A Showcase with Diffusion Models (NeurIPS 2025 · similarity 0.87)
- Unifying Text Semantics and Graph Structures for Temporal Text-attributed Graphs with Large Language Models (NeurIPS 2025 · similarity 0.86)
- Diffusion Adaptive Text Embedding for Text-to-Image Diffusion Models (NeurIPS 2025 · similarity 0.86)
- TGLsta: Low-resource Textual Graph Learning with Semantic and Topological Awareness via LLMs (AAAI 2025 · similarity 0.85)
- TEG-DB: A Comprehensive Dataset and Benchmark of Textual-Edge Graphs (NeurIPS 2024 · similarity 0.85)
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.