Research map / Graph learning / Knowledge and temporal graphs
Rethinking Knowledge Graph Evaluation Under the Open-World Assumption
Haotong Yang, Zhouchen Lin, Muhan Zhang
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 |
|---|---|---|---|
| Add missing true answers to the test answer set | BetaE · FB15k-237 Olympics query | MRR: 20% -> 82% | Improved |
| Add missing answers to test set | BetaE · FB15k-237 Olympics query | filtered ranking of swimming: 5 -> 1 | Improved |
| Increase required test queries when model strength gap shrinks from 0.05 to 0.01 | MRR evaluation · beta=0.35, l=0.7, N=43, p=5% | required N_q: 1140 -> 28500 | Hurt |
| Use less focus-on-top metrics (log-MRR, p-MRR) instead of MRR | KGC evaluation · Family tree KG | degradation and inconsistency | Improved |
| Make fact missingness correlated with a ComplEx model | KGC models on family tree KG · Family tree KG d=75% | sparse MRR gaps between models | Hurt |
| Lower KG density (sparser KG) | MRR · Family tree KG | metric degradation | Hurt |
Closest papers by meaning
- 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 0.89)
- Stability and Generalization Capability of Subgraph Reasoning Models for Inductive Knowledge Graph Completion (ICML 2025 · similarity 0.87)
- Diagnosing and Addressing Pitfalls in KG-RAG Datasets: Toward More Reliable Benchmarking (NeurIPS 2025 · similarity 0.86)
- Context-aware Inductive Knowledge Graph Completion with Latent Type Constraints and Subgraph Reasoning (AAAI 2025 · similarity 0.86)
- Knowledge Graphs Can be Learned with Just Intersection Features (ICML 2024 · similarity 0.85)
- Modeling Knowledge Graphs with Composite Reasoning (AAAI 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.
