Few-NERD: A Few-shot Named Entity Recognition Dataset

By Ning Ding, Guangwei Xu, Yulin Chen, Xiaobin Wang, Xu Han, Pengjun Xie, Haitao Zheng, Zhiyuan Liu. 2021

303 citations recorded by the source corpus.

Recently, considerable literature has grown up around the theme of few-shot named entity recognition (NER), but little published benchmark data specifically focused on the practical and challenging task. Current approaches collect existing supervised NER datasets and re-organize them to the few-shot setting for empirical study. These strategies conventionally aim to recognize coarse-grained entity types with few examples, while in practice, most unseen entity types are fine-grained. In this paper, we present Few-NERD, a large-scale human-annotated few-shot NER dataset with a hierarchy of 8 coarse-grained and 66 fine-grained entity types. Few-NERD consists of 188,238 sentences from Wikipedia, 4,601,160 words are included and each is annotated as context or a part of the two-level entity type. To the best of our knowledge, this is the first few-shot NER dataset and the largest human-crafted NER dataset. We construct benchmark tasks with different emphases to comprehensively assess the generalization capability of models. Extensive empirical results and analysis show that Few-NERD is challenging and the problem requires further research. The Few-NERD dataset and the baselines will be publicly available to facilitate the research on this problem.

Subfields: Artificial Intelligence

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

1. Zhiyuan Liu
2. Maosong Sun
3. Ning Ding
4. Haitao Zheng
5. Shengding Hu
6. Yulin Chen
7. Weilin Zhao
8. Ricardo Henao

Citations in the corpus

• Few-Shot Class-Incremental Learning for Named Entity Recognition (2022, background): “Also, note that class-incremental learning is different meta-learning with episode training (Ding et al., 2021; Finn et al., 2017), since tasks/classes of meta-leaning may appear multiple times in episode training, while we assume the dataset of each class only appear once in class-incremental…”
• Long-context LLMs Struggle with Long In-context Learning (2024, unknown): “…across a variety of real-world domains such as emotion classification from text, named entity recognition, and biological function prediction, each requiring precise differentiation among vast label spaces (Zhang et al., 2017; Sileo et al., 2019; Demszky et al., 2020; Ding et al., 2021).”
• OpenPrompt: An Open-source Framework for Prompt-learning (2021, background): “, 2010), Few-NERD (Ding et al., 2021b) for information extraction; MNLI (Williams et al.”
• LongBench v2: Towards Deeper Understanding and Reasoning on Realistic Long-context Multitasks (2024, unknown): “(Ding et al., 2021), the relation classification dataset DocRED (Yao et al., 2019), the event detection dataset MAVEN (Wang et al., 2020), and the sentiment classification dataset GoEmotions (Demszky et al., 2020).”