Abstract
In this paper, we present Tetris, a new task of Goal-Oriented Script Completion. Unlike previous work, it considers a more realistic and general setting, where the input includes not only the goal but also additional user context, including preferences and history. To address this problem, we propose a novel approach, which uses two techniques to improve performance: (1) concept prompting, and (2) script-oriented contrastive learning that addresses step repetition and hallucination problems. On our WikiHow-based dataset, we find that both methods improve performance.- Anthology ID:
- 2023.eacl-main.220
- Volume:
- Proceedings of the 17th Conference of the European Chapter of the Association for Computational Linguistics
- Month:
- May
- Year:
- 2023
- Address:
- Dubrovnik, Croatia
- Editors:
- Andreas Vlachos, Isabelle Augenstein
- Venue:
- EACL
- SIG:
- Publisher:
- Association for Computational Linguistics
- Note:
- Pages:
- 3026–3040
- Language:
- URL:
- https://aclanthology.org/2023.eacl-main.220
- DOI:
- 10.18653/v1/2023.eacl-main.220
- Cite (ACL):
- Chenkai Sun, Tie Xu, ChengXiang Zhai, and Heng Ji. 2023. Incorporating Task-Specific Concept Knowledge into Script Learning. In Proceedings of the 17th Conference of the European Chapter of the Association for Computational Linguistics, pages 3026–3040, Dubrovnik, Croatia. Association for Computational Linguistics.
- Cite (Informal):
- Incorporating Task-Specific Concept Knowledge into Script Learning (Sun et al., EACL 2023)
- PDF:
- https://preview.aclanthology.org/proper-vol2-ingestion/2023.eacl-main.220.pdf