Research map / Graph learning / Knowledge and temporal graphs
Bringing Light Into the Dark: A Large-Scale Evaluation of Knowledge Graph Embedding Models Under a Unified Framework
Mehdi Ali, Max Berrendorf, Charles Tapley Hoyt, Laurent Vermue, Mikhail Galkin, Sahand Sharifzadeh, Asja Fischer, Volker Tresp C, Jens Lehmann
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 |
|---|---|---|---|
| tune configuration (loss, training approach, inverse relations) with large HPO | RotatE · WN18RR | hits@10: 56.61% -> 60.09% | Improved |
| tune configuration with large HPO | MuRE · WN18RR | hits@10: 55.50% -> 57.90% | Improved |
| tune configuration with large HPO | ConvE · WN18RR | hits@10: 52.00% -> 56.33% | Improved |
| tune configuration with large HPO | TuckER · WN18RR | hits@10: 52.6% -> 56.09% | Improved |
| train with Adam instead of Adadelta | KGEMs · Kinships | hits@10 | Improved |
| train TransE for 4000 epochs instead of 1000 | TransE · WN18 | reproduction: 1000 -> 4000 epochs | Improved |
| use LCWA instead of sLCWA with BCEL/SPL (ablation) | KGEMs · WN18RR | hits@10 | Mixed |
| use LCWA instead of sLCWA (ablation) | TransH · WN18RR | hits@10 | Hurt |
| use LCWA instead of sLCWA (ablation) | RotatE · WN18RR | hits@10 | Improved |
| explicitly model inverse relations (ablation) | KGEMs · Kinships | hits@10 variance | Improved |
| use SPL instead of BCEL (ablation) | KGEMs · WN18RR | hits@10 variance | Improved |
| increase model size | KGEMs · Kinships | hits@10 | No clear effect |
Closest papers by meaning
- PyKEEN 1.0: A Python Library for Training and Evaluating Knowledge Graph Embeddings (JMLR 2021 · similarity 0.93)
- You CAN Teach an Old Dog New Tricks! On Training Knowledge Graph Embeddings (ICLR 2020 · similarity 0.92)
- Interpreting Knowledge Graph Relation Representation from Word Embeddings (ICLR 2021 · similarity 0.91)
- Toward a Unified Statistical Theory of Unsupervised Pretraining and Supervised Neural Knowledge Graph Learning (NeurIPS 2026 · similarity 0.89)
- Editing Language Model-Based Knowledge Graph Embeddings (AAAI 2024 · similarity 0.88)
- Generalizing Knowledge Graph Embedding with Universal Orthogonal Parameterization (ICML 2024 · 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.
