Mining Logical Event Schemas From Pre-Trained Language Models

Lane Lawley, Lenhart Schubert


Abstract
We present NESL (the Neuro-Episodic Schema Learner), an event schema learning system that combines large language models, FrameNet parsing, a powerful logical representation of language, and a set of simple behavioral schemas meant to bootstrap the learning process. In lieu of a pre-made corpus of stories, our dataset is a continuous feed of “situation samples” from a pre-trained language model, which are then parsed into FrameNet frames, mapped into simple behavioral schemas, and combined and generalized into complex, hierarchical schemas for a variety of everyday scenarios. We show that careful sampling from the language model can help emphasize stereotypical properties of situations and de-emphasize irrelevant details, and that the resulting schemas specify situations more comprehensively than those learned by other systems.
Anthology ID:
2022.acl-srw.25
Volume:
Proceedings of the 60th Annual Meeting of the Association for Computational Linguistics: Student Research Workshop
Month:
May
Year:
2022
Address:
Dublin, Ireland
Venue:
ACL
SIG:
Publisher:
Association for Computational Linguistics
Note:
Pages:
332–345
Language:
URL:
https://aclanthology.org/2022.acl-srw.25
DOI:
10.18653/v1/2022.acl-srw.25
Bibkey:
Cite (ACL):
Lane Lawley and Lenhart Schubert. 2022. Mining Logical Event Schemas From Pre-Trained Language Models. In Proceedings of the 60th Annual Meeting of the Association for Computational Linguistics: Student Research Workshop, pages 332–345, Dublin, Ireland. Association for Computational Linguistics.
Cite (Informal):
Mining Logical Event Schemas From Pre-Trained Language Models (Lawley & Schubert, ACL 2022)
Copy Citation:
PDF:
https://aclanthology.org/2022.acl-srw.25.pdf
Video:
 https://aclanthology.org/2022.acl-srw.25.mp4
Data
FrameNet