Research map / Graph learning / Knowledge and temporal graphs
Knowledge graph embeddings
Embed entities and relations with geometric transformations for link prediction.
110 accepted papers at ICML, NeurIPS, ICLR, CVPR and AAAI (2016–2026) · an approach in Knowledge and temporal graphs, keyword cluster “knowledge graph · entities · kg”
State History Benchmarks Tried Baselines Checklist Preprints Who Export
Where it stands
Quiet None of the tracked benchmarks has a report from 2025–26.
20 papers in 2023–24 and 13 in 2025–26 (at ICML, NeurIPS, ICLR, CVPR and AAAI); its share of the topic shrank, from 7.8% to 4.6%.
Open questions the data points to
- Papers report RotatE on WN18RR (MRR) anywhere from 0.411 to 0.476: comparisons against it depend on whose number you take.
- Papers report RotatE on FB15k-237 (MRR) anywhere from 0.296 to 0.338: comparisons against it depend on whose number you take.
No longer reported by papers here: FB15k-237 (MRR, last in 2024); WN18RR (MRR, last in 2024); YAGO3-10 (MRR, last in 2024); FB15K (MRR, last in 2022); WN18 (MRR, last in 2022); FB15k-237 (Hits@10, last in 2022).
How this is decided
A benchmark is tracked when at least four papers of this approach report their own method on the same dataset and metric in at least two different years. Its record is the best number reported up to 2024 compared with the best reported in 2025–26: still improving means at least 1% better, flat means at most 0.2% better with at least two papers from 2025–26 trying. Numbers on a different scale from the rest (a fraction among percentages) and far outliers are left out. Results come from different papers with different setups, so they are signals, not head-to-head comparisons. The share of papers is counted at ICML, NeurIPS, ICLR, CVPR and AAAI.
How it got here
Most cited paper of each year
- 2016 · Holographic Embeddings of Knowledge Graphs (AAAI 2016 · 893 citations)
- 2017 · ConceptNet 5.5: An Open Multilingual Graph of General Knowledge (AAAI 2017 · 2,147 citations)
- 2018 · Convolutional 2D Knowledge Graph Embeddings (AAAI 2018 · 2,484 citations)
- 2019 · RotatE: Knowledge Graph Embedding by Relational Rotation in Complex Space (ICLR 2019 · 756 citations)
- 2020 · Learning Hierarchy-Aware Knowledge Graph Embeddings for Link Prediction (AAAI 2020 · 413 citations)
- 2021 · Dual Quaternion Knowledge Graph Embeddings (AAAI 2021 · 160 citations)
- 2022 · Geometry Interaction Knowledge Graph Embeddings (AAAI 2022 · 86 citations)
- 2023 · Analogical Inference Enhanced Knowledge Graph Embedding (AAAI 2023 · 19 citations)
- 2024 · KGTS: Contrastive Trajectory Similarity Learning over Prompt Knowledge Graph Embedding (AAAI 2024 · 26 citations)
- 2025 · HyperGraphRAG: Retrieval-Augmented Generation via Hypergraph-Structured Knowledge Representation (NeurIPS 2025 · 9 citations)
- 2026 · DANS-KGC: Diffusion Based Adaptive Negative Sampling for Knowledge Graph Completion (AAAI 2026 · 2 citations)
Earliest papers here
- Holographic Embeddings of Knowledge Graphs (AAAI 2016 · 893 citations)
- Representation Learning of Knowledge Graphs with Entity Descriptions (AAAI 2016 · 734 citations)
- Knowledge Graph Completion with Adaptive Sparse Transfer Matrix (AAAI 2016 · 440 citations)
Benchmark records
The best number each year that beat every earlier one, among papers of this approach reporting their own method. Numbers are extracted by AI from each paper's text; check the paper before citing.
FB15k-237 · MRR (higher is better)
24 papers, 2018–2024 · not reported since 2024
| Year | Best so far | Set by |
|---|---|---|
| 2018 | 0.325 | Convolutional 2D Knowledge Graph Embeddings (AAAI 2018 · 2,484 citations) |
| 2019 | 0.404 | TransGate: Knowledge Graph Embedding with Shared Gate Structure (AAAI 2019 · 28 citations) |
| 2021 | 0.415 | Neural Bellman-Ford Networks: A General Graph Neural Network Framework for Link Prediction (NeurIPS 2021 · 113 citations) |
| 2024 | 0.52 | KGDM: A Diffusion Model to Capture Multiple Relation Semantics for Knowledge Graph Embedding (AAAI 2024 · 11 citations) |
WN18RR · MRR (higher is better)
23 papers, 2018–2024 · not reported since 2024
| Year | Best so far | Set by |
|---|---|---|
| 2018 | 0.43 | Convolutional 2D Knowledge Graph Embeddings (AAAI 2018 · 2,484 citations) |
| 2019 | 0.488 | Quaternion Knowledge Graph Embeddings (NeurIPS 2019 · 281 citations) |
| 2020 | 0.497 | Learning Hierarchy-Aware Knowledge Graph Embeddings for Link Prediction (AAAI 2020 · 413 citations) |
| 2021 | 0.5 | 5* Knowledge Graph Embeddings with Projective Transformations (AAAI 2021 · 30 citations) |
| 2022 | 0.511 | HousE: Knowledge Graph Embedding with Householder Parameterization (ICML 2022 · 20 citations) |
| 2024 | 0.525 | Generalizing Knowledge Graph Embedding with Universal Orthogonal Parameterization (ICML 2024 · 0 citations) |
YAGO3-10 · MRR (higher is better)
14 papers, 2018–2024 · not reported since 2024
| Year | Best so far | Set by |
|---|---|---|
| 2018 | 0.44 | Convolutional 2D Knowledge Graph Embeddings (AAAI 2018 · 2,484 citations) |
| 2019 | 0.495 | RotatE: Knowledge Graph Embedding by Relational Rotation in Complex Space (ICLR 2019 · 756 citations) |
| 2020 | 0.545 | Learning Hierarchy-Aware Knowledge Graph Embeddings for Link Prediction (AAAI 2020 · 413 citations) |
| 2022 | 0.579 | Geometry Interaction Knowledge Graph Embeddings (AAAI 2022 · 86 citations) |
| 2024 | 0.588 | Generalizing Knowledge Graph Embedding with Universal Orthogonal Parameterization (ICML 2024 · 0 citations) |
FB15K · MRR (higher is better)
8 papers, 2018–2022 · not reported since 2022
| Year | Best so far | Set by |
|---|---|---|
| 2018 | 0.657 | Convolutional 2D Knowledge Graph Embeddings (AAAI 2018 · 2,484 citations) |
| 2019 | 0.833 | Quaternion Knowledge Graph Embeddings (NeurIPS 2019 · 281 citations) |
| 2021 | 0.84 | 5* Knowledge Graph Embeddings with Projective Transformations (AAAI 2021 · 30 citations) |
WN18 · MRR (higher is better)
7 papers, 2018–2022 · not reported since 2022
| Year | Best so far | Set by |
|---|---|---|
| 2018 | 0.943 | Convolutional 2D Knowledge Graph Embeddings (AAAI 2018 · 2,484 citations) |
| 2019 | 0.95 | Quaternion Knowledge Graph Embeddings (NeurIPS 2019 · 281 citations) |
| 2021 | 0.96 | 5* Knowledge Graph Embeddings with Projective Transformations (AAAI 2021 · 30 citations) |
FB15k-237 · Hits@10 (higher is better)
5 papers, 2018–2022 · not reported since 2022
| Year | Best so far | Set by |
|---|---|---|
| 2018 | 0.491 | Convolutional 2D Knowledge Graph Embeddings (AAAI 2018 · 2,484 citations) |
| 2019 | 0.54 | End-to-End Structure-Aware Convolutional Networks for Knowledge Base Completion (AAAI 2019 · 665 citations) |
| 2020 | 0.576 | ParamE: Regarding Neural Network Parameters as Relation Embeddings for Knowledge Graph Completion (AAAI 2020 · 53 citations) |
What has been tried
99 papers here report 968 measured changes (a component added, removed or swapped, compared with a baseline): helped 69%, no clear effect 9%, hurt 19%, mixed 4%. Papers mostly publish changes that helped, so "helped" is over-represented.
Ablations in recent papers
| What was changed | Measured on | Result | Effect | Paper |
|---|---|---|---|---|
| add GRot grid embedding | Porto · HR@1 | 0.3243 -> 0.5244 | Helped | KGTS: Contrastive Trajectory Similarity Learning over Prompt Knowledge Graph Embedding (AAAI 2024 · 26 citations) |
| Add context to R-GCN link prediction | DocRED · Hits@1 | 0.06 -> 0.11 | Helped | Revisiting Document-Level Relation Extraction with Context-Guided Link Prediction (AAAI 2024 · 15 citations) |
| add the Scoring Module | FB15k-237 · MRR | 0.449 -> 0.520 | Helped | KGDM: A Diffusion Model to Capture Multiple Relation Semantics for Knowledge Graph Embedding (AAAI 2024 · 11 citations) |
| add entity retrieval | Medicine · F1 | 35.4 -> 29.8 | Helped | HyperGraphRAG: Retrieval-Augmented Generation via Hypergraph-Structured Knowledge Representation (NeurIPS 2025 · 9 citations) |
| Use relational clustering-based anchor selection instead of random selection | FB15k-237 entity prediction · MRR | 0.249 -> 0.265 | Helped | Clustering then Propagation: Select Better Anchors for Knowledge Graph Embedding (NeurIPS 2024 · 5 citations) |
| remove retriever (random relevant descriptions) from DRKA | FB15K link prediction, Overall · Hits@10 | 54.8 -> 51.5 | Hurt | Select and Augment: Enhanced Dense Retrieval Knowledge Graph Augmentation (Abstract Reprint) (AAAI 2024 · 1 citation) |
| Remove the adaptive dangling indicator r_ei | DBP2.0 entity alignment · alignment performance | Hurt | Lambda: Learning Matchable Prior For Entity Alignment with Unlabeled Dangling Cases (NeurIPS 2024 · 1 citation) | |
| Use MAYPL's structure-driven initializer rather than learnable vectors for each entity and relation | WikiPeople´ · MRR | 0.388 -> 0.521 | Helped | Structure Is All You Need: Structural Representation Learning on Hyper-Relational Knowledge Graphs (ICML 2025 · 0 citations) |
| Replace Euclidean geometry with elliptic geometry | WN18RR · MRR | .496 -> .505 | Helped | Generalizing Knowledge Graph Embedding with Universal Orthogonal Parameterization (ICML 2024 · 0 citations) |
| use mean aggregator instead of sum aggregator in RAMP encoder | FB15K-237, CoDEx, UMLS · generalization error | Helped | PAC-Bayesian Generalization Bounds for Knowledge Graph Representation Learning (ICML 2024 · 0 citations) | |
| Replace late interaction with early interaction | AIDS and NCI · MAP | 0.557 → 0.726; 0.529 → 0.624 | Helped | Charting the Design Space of Neural Graph Representations for Subgraph Matching (ICLR 2025) |
Baseline numbers that disagree
The same baseline on the same benchmark and metric, reported more than 5% apart by different papers here. Training budgets, data splits and reimplementations differ, so state which number you compare with.
| Baseline | Benchmark | Metric | Reported | Papers |
|---|---|---|---|---|
| RotatE | WN18RR | MRR | 0.411 – 0.476 | 7 papers, e.g. AAAI 2023 and ICLR 2022 |
| RotatE | FB15k-237 | MRR | 0.296 – 0.338 | 7 papers, e.g. AAAI 2023 and ICLR 2022 |
| RotatE | YAGO3-10 | MRR | 0.121 – 0.495 | 3 papers, e.g. AAAI 2023 and AAAI 2020 |
| ComplEx | WN18RR | MRR | 0.44 – 0.49 | 3 papers, e.g. ICLR 2019 and AAAI 2021 |
| ConvE | NELL-995 | MRR | 49.1 – 75.42 | 2 papers, e.g. AAAI 2021 and AAAI 2020 |
| RESCAL | FB15K test | HITS@10 (%) | 44.1 – 59.6 | 2 papers, e.g. NeurIPS 2018 and AAAI 2018 |
What reviewers will expect
Datasets and baselines named in the 5 papers of this approach from 2025–2026 that have a full-text summary. Leaving out the common ones is a frequent review question.
Baselines
- DistMult · 2 of 5 papers
Recent preprints
Loading recent arXiv preprints…
Who publishes on it
Institutions on the most papers of this approach since 2023 (of 33 papers; from OpenAlex and the papers' affiliations).
- Alibaba Group · 3 papers
- Zhejiang University · 3 papers
- Donghai Laboratory · 3 papers
- Beijing University of Posts and Telecommunications · 3 papers
- Korea Advanced Institute of Science and Technology · 3 papers
- Huawei · 2 papers
- University of Edinburgh · 2 papers
- University of Science and Technology of China · 2 papers
- Renmin University of China · 2 papers
- University of California, Riverside · 2 papers
Use it
Download BibTeX of 40 key papersOpen the topic in the interactive map
Get alerts. Describe your idea in scime atlas and save it (free scime account): you get a weekly email when new preprints land close to it.
Related approaches
Same keyword cluster
- Temporal knowledge graph reasoning (84 papers)
- Logical query answering and rule reasoning (68 papers)
- Inductive knowledge graph completion (38 papers)
Elsewhere in the topic
- Continuous-time dynamic graph learning (132 papers)
- Graph reasoning for visual recognition (96 papers)
- Graph-based dialog systems (94 papers)
- Social and behavior graphs (87 papers)
- Scene graph generation (70 papers)
- Temporal graph neural architectures (69 papers)
- Heterogeneous graph learning (66 papers)
- Spatio-temporal graph traffic forecasting (60 papers)
Benchmark numbers, ablations and baseline numbers are extracted by AI from each paper's full text and can be wrong; the paper is the authority. Data: results index of 2026-11-02, citation counts from OpenAlex (2026-10-09).