Research map / Graph learning / Knowledge and temporal graphs
INDIGO: GNN-Based Inductive Knowledge Graph Completion Using Pair-Wise Encoding
Shuwen Liu, Bernardo Cuenca Grau, Ian Horrocks, Egor V. Kostylev
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 changed | Where | Measured | Effect |
|---|---|---|---|
| Pair-wise GNN encoding with direct read-out instead of GraIL subgraph scoring | inductive KG completion · GraIL-BM FB15K-237 v1 | AUC: 78.6 -> 93.4 | Improved |
| Pair-wise encoding instead of GraIL | inductive KG completion · GraIL-BM NELL-995 v4 | accuracy: 49.7 -> 85.2 | Improved |
| Pair-wise encoding instead of GraIL | inductive KG completion · GraIL-BM NELL-995 v1 | accuracy: 97.3 -> 85.6 | Hurt |
| Pair-wise encoding instead of Hamaguchi et al. | inductive KG completion · Hamaguchi-BM h-1K | accuracy: 83.6 -> 75.3 | Hurt |
| Pair-wise encoding instead of Hamaguchi et al. | inductive KG completion · Hamaguchi-BM b-5K | AUC: 68.4 -> 96.8 | Improved |
| Pair-wise encoding instead of GraIL | inductive KG completion · GraIL-BM NELL-995 v1 | e-Hits@3: 51.0 -> 39.5 | Hurt |
| Pair-wise encoding instead of R-GCN/GraIL | inductive KG completion · INDIGO-BM | AUC: 94.2 -> 99.0 | Improved |
| INDIGO instead of GraIL for capturing type-implication rules | rule capture · INDIGO-BM | captured rules: 0 -> 207 | Improved |
| INDIGO instead of GraIL for chain rules | rule capture · INDIGO-BM | captured rules: 130 (19%) -> 150 (22%) | Improved |
Later papers that compared with this method
1 later accepted paper reports INDIGO as a baseline in their results tables; 1 of them report a better number on most of the results they share with it.
- Rule-Guided Graph Neural Networks for Explainable Knowledge Graph Reasoning (AAAI 2025 · better on 8 of 8 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
- Replacing Paths with Connection-Biased Attention for Knowledge Graph Completion (AAAI 2025 · similarity 0.89)
- Towards Global-Topology Relation Graph for Inductive Knowledge Graph Completion (AAAI 2025 · similarity 0.89)
- Context-aware Inductive Knowledge Graph Completion with Latent Type Constraints and Subgraph Reasoning (AAAI 2025 · similarity 0.88)
- KGCRR: An Effective Metric-Driven Knowledge Graph Completion Framework by Designing a Novel Upper Bound Function with Adaptive Approximation to Reciprocal Rank (AAAI 2025 · similarity 0.88)
- Knowledge Graphs Can be Learned with Just Intersection Features (ICML 2024 · similarity 0.88)
- Mixed Geometry Message and Trainable Convolutional Attention Network for Knowledge Graph Completion (AAAI 2024 · similarity 0.87)
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.
