😆 Must-read papers on relation extraction especifically for low-resource setting.
Dataset | #train | #dev | #test | #rel | no_relation | entity type |
---|---|---|---|---|---|---|
TACRED [link | paper] | 68,124 | 22,631 | 15,509 | 42 | ✔ | ✔ |
TACREV [link | paper] | 68,124 | 22,631 | 15,509 | 42 | ✔ | ✔ |
Re-TACRED [link | paper] | 58,465 | 19,584 | 13,418 | 40 | ✔ | ✔ |
Wiki80 [link | paper] | 50,400 | 5,600 | -- | 80 | ✘ | ✘ |
FewRel 1.0 [link | paper] | 44,800 | 11,200 | 14,000* | 100 (64/16/20) | ✘ | ✘ |
FewRel 2.0 [link | paper] | 44,800 | 2,500 | (64 / 25) | ✘ | ✘ |
(*--> unpublic)
Dataset | #train | #dev | #test | #rel | #tuples (train | test) | entity overlap type (NEO/EPO/SEO) |
---|---|---|---|---|---|---|
NYT24 [link | paper] | 56,196 | 5,000 | 24 | 88,366 | 8,120 | 37,371 / 15,124 / 18,825 | 3,289 / 1,410 / 1,711 | |
NYT29 [link | paper] | 63,306 | 4,006 | 29 | 78,973 | 5,859 | 53,444 / 8,379 / 9,862 | 2,963 / 898 / 1,043 | |
WebNLG [link | paper ] | 5,019 | 500 | 703 | 216 | ||
ACE05 [link] | ||||||
ACE04 [link] | ||||||
SciERC [link | paper ] | 1,861 | 275 | 551 | 7 |
-
FewRel: A Large-Scale Supervised Few-Shot Relation Classification Dataset with State-of-the-Art Evaluation. [pdf], [project]
-
FewRel 2.0: Towards More Challenging Few-Shot Relation Classification. [pdf], [project]
-
Hybrid Attention-Based Prototypical Networks for Noisy Few-Shot Relation Classification. [pdf], [project]
-
Multi-Level Matching and Aggregation Network for Few-Shot Relation Classification. [pdf], [project]
-
Matching the Blanks: Distributional Similarity for Relation Learning. [pdf], [project]
-
Hierarchical Attention Prototypical Networks for Few-Shot Text Classification. [pdf ]
-
Few-shot Relation Extraction via Bayesian Meta-learning on Relation Graphs. [pdf], [project]
-
Enhance Prototypical Network with Text Descriptions for Few-shot Relation Classification. [pdf]
-
Learning from Context or Names? An Empirical Study on Neural Relation Extraction. [pdf], [project]
-
Bridging Text and Knowledge with Multi-Prototype Embedding for Few-Shot Relational Triple Extraction. [pdf]
-
Entity Concept-enhanced Few-shot Relation Extraction. [pdf], [project]
-
Learning Discriminative and Unbiased Representations for Few-Shot Relation Extraction. [pdf]]
-
Zero-shot Relation Classification from Side Information. [pdf], [project]
-
MapRE: An Effective Semantic Mapping Approach for Low-resource Relation Extraction [pdf]
-
Exploring Task Difficulty for Few-Shot Relation Extraction. [pdf], [project]
-
Towards Realistic Few-Shot Relation Extraction. [pdf], [project]
- KnowPrompt: Knowledge-aware Prompt-tuning with Synergistic Optimization for Relation Extraction. [pdf], [project]
- PTR: Prompt Tuning with Rules for Text Classification. [pdf], [project]
- GradLRE: Gradient Imitation Reinforcement Learning for Low resource Relation Extraction. [pdf], [project]
- Label Verbalization and Entailment for Effective Zero and Few-Shot Relation Extraction. [pdf], [project]
Joint Extraction of Entities and Relations