Research map / Graph learning / Knowledge and temporal graphs
NodePiece: Compositional and Parameter-Efficient Representations of Large Knowledge Graphs
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 |
|---|---|---|---|
| Represent entities compositionally with NodePiece anchors and relations instead of separate entity embeddings | RotatE on YAGO 3-10 · YAGO 3-10 | Hits@10: 0.262 -> 0.488 | Helped |
| Use NodePiece representations with AutoSF rather than shallow AutoSF embeddings | AutoSF · OGB WikiKG 2 test | MRR: 0.546 $\scriptstyle{\pm 0.005}$ -> 0.570 $\scriptstyle{\pm 0.003}$ | Helped |
| Replace learned entity embeddings with NodePiece features | CompGCN · WD50K (5% labeled) | ROC-AUC: 0.836 -> 0.981 | Helped |
| Use NodePiece instead of oDistMult-ERAvg | out-of-sample link prediction · oFB15k-237 | MRR: 0.256$^\dagger$ -> 0.206 | Hurt |
| Remove relational context (ablation) | NodePiece + RotatE · FB15k-237 | Hits@10: 0.420 -> 0.425 | No clear effect |
| Remove relational context (ablation) | NodePiece + RotatE · WN18RR | Hits@10: 0.515 -> 0.465 | Hurt |
| Remove anchor distances (ablation) | NodePiece + RotatE · WN18RR | Hits@10: 0.515 -> 0.510 | No clear effect |
| Remove relational context (ablation) | NodePiece + AutoSF · OGB WikiKG 2 test | MRR: 0.570 $\scriptstyle{\pm 0.003}$ -> 0.592 $\scriptstyle{\pm 0.003}$ | Helped |
| Remove anchors and use relations only (ablation) | NodePiece + AutoSF · OGB WikiKG 2 test | MRR: 0.570 $\scriptstyle{\pm 0.003}$ -> 0.476 $\scriptstyle{\pm 0.001}$ | Hurt |
| Remove relational context (ablation) | NodePiece + GNN · WD50K (5% labeled) | Hard Accuracy: 0.513 -> 0.534 | No clear effect |
| Remove anchors and use relations only (ablation) | NodePiece + GNN · WD50K (5% labeled) | Hard Accuracy: 0.513 -> 0.532 | No clear effect |
| Remove relational context (ablation) | NodePiece + DistMult · oFB15k-237 | MRR: 0.206 -> 0.173 | Hurt |
Closest papers by meaning
- Knowledge Graph Completion with Relation-Aware Anchor Enhancement (AAAI 2025 · similarity 0.89)
- Select and Augment: Enhanced Dense Retrieval Knowledge Graph Augmentation (Abstract Reprint) (AAAI 2024 · similarity 0.88)
- Clustering then Propagation: Select Better Anchors for Knowledge Graph Embedding (NeurIPS 2024 · similarity 0.88)
- Toward a Unified Statistical Theory of Unsupervised Pretraining and Supervised Neural Knowledge Graph Learning (NeurIPS 2026 · similarity 0.88)
- KGDM: A Diffusion Model to Capture Multiple Relation Semantics for Knowledge Graph Embedding (AAAI 2024 · similarity 0.87)
- K-ON: Stacking Knowledge on the Head Layer of Large Language Model (AAAI 2025 · 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.