Information extraction dataset zoo.
| Dataset | #Rel. | #Inst. | Feature | Source | Resource | Origin |
|---|---|---|---|---|---|---|
| Fewrel | 100 | 44,800 | Supervised | Wikipedia+Wikidata | url | url |
| TACRED | 42 | 68,120 | Supervised | Newswire+web | - | url |
| Semeval | 19 | 8,000 | Supervised | Web | url | url |
| Wikidata | 352 | 495,883 | Distent-supervision | Wikipedia+Wikidata | url | url |
| NYT10(tsinghua) | 53 | 522,043 | Distent-supervision | NYT+Freebase | url | url |
| NYT10-large(tsinghua) | 53 | 570,088 | Distent-supervision | NYT+Freebase | url | url |
| NYT-Wikidata | 100 | 882,177 | Distent-supervision | NYT+Wikidata | url | url |
| NYT10-29 | 29 | 70,339 | Distent-supervision | NYT+Freebase | url | url |
| NYT11-12 | 12 | 62,648 | DS+supervised | NYT+Freebase | url | url |
| NYT-manual | 24 | 235,982 | Distent-supervision | NYT+Freebase | url | url |
| NYT-Wiki(zju) | 73 | 1,989,377 | Distent-supervision | NYT-Wikipedia-Wikidata | url | url |
| Wiki-KBP | 19 | 23,784 | Distent-supervision | Wikipedia+KBP+Freebase | url | url |
| PubMed-BioInfer | 94 | 1,580 | Distent-supervision | PubMed+NESH | - | url |
| WebNLG | 14 | 75,325 | Supervised | Web | - | url |
| SKE | 50 | 173,108 | Supervised | Web | url | url |
| KBP37 | 37 | 15,916 | Supervised | Web | url | url |
| T-REx | 642 | 6.3M | Distent-supervision | Wikipedia+Wikidata | - | url |
| Google-RE | 5 | 59,576 | Supervised | Wikipedia | - | url |
| ADE | 3 | 23,516 | Supervised | Medical Report | url | url |
Other Datasets
FewRel : A Large-Scale Supervised Few-Shot Relation Classification Dataset with State-of-the-Art Evaluation
Matching the Blanks : Distributional Similarity for Relation Learning ACL2019
Multi-Level Matching and Aggregation Network for Few-Shot Relation Classification ACL2019
Position-aware attention and supervised data improve slot filling.
Matching the Blanks : Distributional Similarity for Relation Learning ACL2019
Matching the Blanks : Distributional Similarity for Relation Learning
Context-Aware Representations for Knowledge Base Relation Extraction
Attention-Based Capsule Networks with Dynamic Routing for Relation Extraction
Distant Supervision for Relation Extraction via Piecewise Convolutional Neural Networks
Self-Attention Enhanced Selective Gate with Entity-Aware Embedding for Distantly Supervised Relation Extraction
Attention-Based Capsule Networks with Dynamic Routing for Relation Extraction EMNLP2018
Hierarchical Relation Extraction with Coarse-to-Fine Grained Attention EMNLP2018
Long-tail Relation Extraction via Knowledge Graph Embeddings and Graph Convolution Networks NAACL2019
Incorporating Relation Paths in Neural Relation Extraction EMNLP2017
A Hierarchical Framework for Relation Extraction with Reinforcement Learning
Joint Extraction of Entities and Relations with a Hierarchical Multi-task Tagging Model
A Hierarchical Framework for Relation Extraction with Reinforcement Learning
Joint Extraction of Entities and Relations with a Hierarchical Multi-task Tagging Model
Indirect Supervision for Relation Extraction Using Question-Answer Pairs
CoType: Joint Extraction of Typed Entities and Relations with Knowledge Bases.
Joint Extraction of Entities and Relations Based on a Novel Decomposition Strategy
Extracting Relational Facts by an End-to-End Neural Model with Copy Mechanism
Joint Extraction of Entities and Relations Based on a Novel Tagging Scheme
Relation Adversarial Network for Low Resource Knowledge Graph Completion
CoType: Joint Extraction of Typed Entities and Relations with Knowledge Bases.
Indirect Supervision for Relation Extraction Using Question-Answer Pairs
CoType: Joint Extraction of Typed Entities and Relations with Knowledge Bases.
Extracting relational facts by an end-to-end neural model with copy mechanism
MrMep: joint extraction of multiple relations and multiple entity pairs based on triplet attention
Matching the Blanks : Distributional Similarity for Relation Learning ACL2019
Relation classification via recurrent neural network
T-Rex : A Large Scale Alignment of Natural Language with Knowledge Base Triples
K-ADAPTER: Infusing Knowledge into Pre-Trained Models with Adapters
| Dataset | # Inst. | Feature | Source | Resource | Origin |
|---|---|---|---|---|---|
| ACE05 | 599 | Supervised | Web | - | url |
| FewEvent(zju) | 71,385 | Supervised | ACE05+_TAC-KBP17 | url | url |
| CCKS2019_Event | 17,815 | Supervised | Financial Announcements | url | url |
| Doc2EDAG | 32,040 | Supervised | Financial Announcements | url | url |
too many papers
Meta-Learning with Dynamic-Memory-Based Prototypical Network for Few-Shot Event Detection
Doc2EDAG: An End-to-End Document-level Framework for Chinese Financial Event Extraction