Research map / Graph learning / Graph self-supervised learning
Deep Gaussian Embedding of Graphs: Unsupervised Inductive Learning via Ranking
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 |
|---|---|---|---|
| Embed nodes as Gaussians with attributes (G2G) instead of one-hot encoding (G2G oh) (ablation) | Graph2Gauss · Link prediction, Cora-ML | AUC: 98.01 vs 96.95 | Helped |
| Use node attributes (G2G) rather than one-hot encoding (G2G oh) (ablation) | Graph2Gauss · Citeseer link prediction | AUC: 96.09 vs 95.89 | Helped |
| Use Gaussian embedding with attributes vs GAE baseline | Graph2Gauss · Link prediction across five datasets (Table 1) | AUC: G2G exceeds GAE on all datasets, e.g. Citeseer 96.09 vs 92.31 | Helped |
| Inductive learning: unseen nodes embedded via attribute encoder only | Graph2Gauss · Inductive link prediction, Cora-ML, 10% hidden | AUC: 90.93 vs 75.95 for logistic regression | Helped |
| Increase fraction of unseen nodes from 10% to 25% in inductive setting | Graph2Gauss · Inductive link prediction, Cora-ML | AUC: 87.83 vs 90.93 | Hurt |
| Node-anchored sampling strategy instead of naive uniform triplet sampling (ablation) | Graph2Gauss training · Link prediction convergence, Cora-ML | validation AUC/AP vs number of pairs seen | Helped |
| Dropping highly uncertain latent dimensions (those with increasing variance) (ablation) | Graph2Gauss · Link prediction, Cora-ML | AUC/AP | No clear effect |
| Attributed G2G embedding vs logistic regression on attributes only | Graph2Gauss · Link prediction, Cora-ML | AUC: 98.01 vs 90.01 | Helped |
Closest papers by meaning
- Watch Your Step: Learning Node Embeddings via Graph Attention (NeurIPS 2018 · similarity 0.91)
- Fisher Information Embedding for Node and Graph Learning (ICML 2023 · similarity 0.90)
- iN2V: Bringing Transductive Node Embeddings to Inductive Graphs (ICML 2025 · similarity 0.89)
- How Low Can You Go? Searching for the Intrinsic Dimensionality of Complex Networks using Metric Node Embeddings (ICLR 2025 · similarity 0.88)
- Learning Representations for Hierarchies with Minimal Support (NeurIPS 2024 · similarity 0.88)
- Node Identifiers: Compact, Discrete Representations for Efficient Graph Learning (ICLR 2025 · similarity 0.88)
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.