Research map / Graph learning / Knowledge and temporal graphs
Probability Calibration for Knowledge Graph Embedding Models
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 isotonic regression calibration using ground truth negatives | TransE · WN11 | Brier score: .443 -> .087 | Helped |
| Add isotonic regression calibration using weighted synthetic negatives | TransE · WN11 | Brier score: .443 -> .088 | Helped |
| Use synthetic rather than ground truth negatives for isotonic calibration | TransE · FB13 | Brier score: .124 -> .141 | Hurt |
| Use isotonic regression rather than Platt scaling | TransE · WN11 | Brier score: .089 -> .087 | Helped |
| Replace self-adversarial loss with pairwise loss (ablation) | TransE with isotonic calibration · WN11 | Brier score: .087 -> .198 | Hurt |
| Replace self-adversarial loss with NLL loss (ablation) | TransE with isotonic calibration · WN11 | Brier score: .087 -> .088 | No clear effect |
| Replace self-adversarial loss with Multiclass-NLL loss (ablation) | TransE with isotonic calibration · WN11 | Brier score: .087 -> .189 | Hurt |
| Use one calibrated-probability threshold instead of learned per-relation thresholds | TransE · WN11 triple classification | accuracy: 88.9 vs 88.9 | No clear effect |
Closest papers by meaning
- Toward a Unified Statistical Theory of Unsupervised Pretraining and Supervised Neural Knowledge Graph Learning (NeurIPS 2026 · similarity 0.88)
- You CAN Teach an Old Dog New Tricks! On Training Knowledge Graph Embeddings (ICLR 2020 · similarity 0.88)
- Interpreting Knowledge Graph Relation Representation from Word Embeddings (ICLR 2021 · similarity 0.88)
- Improved Knowledge Graph Embedding Using Background Taxonomic Information (AAAI 2019 · similarity 0.87)
- ProjE: Embedding Projection for Knowledge Graph Completion (AAAI 2017 · similarity 0.87)
- Uncertain Knowledge Graph Completion via Semi-Supervised Confidence Distribution Learning (NeurIPS 2025 · 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.