CoAnnotating: Uncertainty-Guided Work Allocation between Human and Large Language Models for Data Annotation

Minzhi Li, Taiwei Shi, Caleb Ziems, Min-Yen Kan, Nancy Chen, Zhengyuan Liu, Diyi Yang


Abstract
Annotated data plays a critical role in Natural Language Processing (NLP) in training models and evaluating their performance. Given recent developments in Large Language Models (LLMs), models such as ChatGPT demonstrate zero-shot capability on many text-annotation tasks, comparable with or even exceeding human annotators. Such LLMs can serve as alternatives for manual annotation, due to lower costs and higher scalability. However, limited work has leveraged LLMs as complementary annotators, nor explored how annotation work is best allocated among humans and LLMs to achieve both quality and cost objectives. We propose CoAnnotating, a novel paradigm for Human-LLM co-annotation of unstructured texts at scale. Under this framework, we utilize uncertainty to estimate LLMs’ annotation capability. Our empirical study shows CoAnnotating to be an effective means to allocate work from results on different datasets, with up to 21% performance improvement over random baseline. For code implementation, see https://github.com/SALT-NLP/CoAnnotating.
Anthology ID:
2023.emnlp-main.92
Volume:
Proceedings of the 2023 Conference on Empirical Methods in Natural Language Processing
Month:
December
Year:
2023
Address:
Singapore
Editors:
Houda Bouamor, Juan Pino, Kalika Bali
Venue:
EMNLP
SIG:
Publisher:
Association for Computational Linguistics
Note:
Pages:
1487–1505
Language:
URL:
https://aclanthology.org/2023.emnlp-main.92
DOI:
10.18653/v1/2023.emnlp-main.92
Bibkey:
Cite (ACL):
Minzhi Li, Taiwei Shi, Caleb Ziems, Min-Yen Kan, Nancy Chen, Zhengyuan Liu, and Diyi Yang. 2023. CoAnnotating: Uncertainty-Guided Work Allocation between Human and Large Language Models for Data Annotation. In Proceedings of the 2023 Conference on Empirical Methods in Natural Language Processing, pages 1487–1505, Singapore. Association for Computational Linguistics.
Cite (Informal):
CoAnnotating: Uncertainty-Guided Work Allocation between Human and Large Language Models for Data Annotation (Li et al., EMNLP 2023)
Copy Citation:
PDF:
https://aclanthology.org/2023.emnlp-main.92.pdf
Video:
 https://aclanthology.org/2023.emnlp-main.92.mp4