Research map / Graph learning / Graph self-supervised learning
Adversarial Directed Graph 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 |
|---|---|---|---|
| use the unified two-generator DGGAN instead of the strongest listed baseline | directed graph link prediction · Cora, reversed-edge test fractions 0%, 50%, and 100% | AUC (%): 85.3 -> 85.1; 76.4 -> 86.7; 74.3 -> 88.3 | Mixed |
| use the unified two-generator DGGAN instead of the strongest listed baseline | directed graph link prediction · Twitter, reversed-edge test fractions 0%, 50%, and 100% | AUC (%): 98.0 -> 99.7; 97.9 -> 99.7; 97.8 -> 99.7 | Helped |
| use the unified two-generator DGGAN instead of the strongest listed baseline | directed graph link prediction · Epinions, reversed-edge test fractions 0%, 50%, and 100% | AUC (%): 89.7 -> 96.1; 74.6 -> 86.4; 72.6 -> 85.1 | Helped |
| use the unified two-generator DGGAN instead of the strongest listed baseline | directed graph link prediction · Google, reversed-edge test fractions 0%, 50%, and 100% | AUC (%): 92.1 -> 92.3; 85.2 -> 93.4; 89.0 -> 94.4 | Helped |
| add the source-neighborhood generator to the target-only DGGAN* variant (ablation) | DGGAN · Cora, reversed-edge test fractions 0%, 50%, and 100% | AUC (%): 83.0 -> 85.1; 83.3 -> 86.7; 83.5 -> 88.3 | Helped |
| add the source-neighborhood generator to the target-only DGGAN* variant (ablation) | DGGAN · Twitter, reversed-edge test fractions 0%, 50%, and 100% | AUC (%): 99.4 -> 99.7; 99.3 -> 99.7; 99.2 -> 99.7 | Helped |
| add the source-neighborhood generator to the target-only DGGAN* variant (ablation) | DGGAN · Epinions, reversed-edge test fractions 0%, 50%, and 100% | AUC (%): 92.7 -> 96.1; 80.0 -> 86.4; 78.2 -> 85.1 | Helped |
| add the source-neighborhood generator to the target-only DGGAN* variant (ablation) | DGGAN · Google, reversed-edge test fractions 0%, 50%, and 100% | AUC (%): 91.6 -> 92.3; 89.2 -> 93.4; 87.7 -> 94.4 | Helped |
Closest papers by meaning
- GraphGAN: Graph Representation Learning With Generative Adversarial Nets (AAAI 2018 · similarity 0.89)
- Distribution-Induced Bidirectional Generative Adversarial Network for Graph Representation Learning (CVPR 2020 · similarity 0.88)
- Learning Representations for Hierarchies with Minimal Support (NeurIPS 2024 · similarity 0.87)
- TRACI: A Data-centric Approach for Multi-Domain Generalization on Graphs (AAAI 2025 · similarity 0.86)
- Enhancing Size Generalization in Graph Neural Networks through Disentangled Representation Learning (ICML 2024 · similarity 0.86)
- DUPLEX: Dual GAT for Complex Embedding of Directed Graphs (ICML 2024 · similarity 0.86)
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.