Instilling Type Knowledge in Language Models via Multi-Task QA

Shuyang Li, Mukund Sridhar, Chandana Satya Prakash, Jin Cao, Wael Hamza, Julian McAuley


Abstract
Understanding human language often necessitates understanding entities and their place in a taxonomy of knowledge—their types.Previous methods to learn entity types rely on training classifiers on datasets with coarse, noisy, and incomplete labels. We introduce a method to instill fine-grained type knowledge in language models with text-to-text pre-training on type-centric questions leveraging knowledge base documents and knowledge graphs.We create the WikiWiki dataset: entities and passages from 10M Wikipedia articles linked to the Wikidata knowledge graph with 41K types.Models trained on WikiWiki achieve state-of-the-art performance in zero-shot dialog state tracking benchmarks, accurately infer entity types in Wikipedia articles, and can discover new types deemed useful by human judges.
Anthology ID:
2022.findings-naacl.45
Volume:
Findings of the Association for Computational Linguistics: NAACL 2022
Month:
July
Year:
2022
Address:
Seattle, United States
Venue:
Findings
SIG:
Publisher:
Association for Computational Linguistics
Note:
Pages:
594–603
Language:
URL:
https://aclanthology.org/2022.findings-naacl.45
DOI:
10.18653/v1/2022.findings-naacl.45
Bibkey:
Cite (ACL):
Shuyang Li, Mukund Sridhar, Chandana Satya Prakash, Jin Cao, Wael Hamza, and Julian McAuley. 2022. Instilling Type Knowledge in Language Models via Multi-Task QA. In Findings of the Association for Computational Linguistics: NAACL 2022, pages 594–603, Seattle, United States. Association for Computational Linguistics.
Cite (Informal):
Instilling Type Knowledge in Language Models via Multi-Task QA (Li et al., Findings 2022)
Copy Citation:
PDF:
https://preview.aclanthology.org/nodalida-main-page/2022.findings-naacl.45.pdf
Video:
 https://preview.aclanthology.org/nodalida-main-page/2022.findings-naacl.45.mp4
Code
 amazon-research/wikiwiki-dataset
Data
WikiWiki