Don’t Take This Out of Context!: On the Need for Contextual Models and Evaluations for Stylistic Rewriting

Akhila Yerukola, Xuhui Zhou, Elizabeth Clark, Maarten Sap


Abstract
Most existing stylistic text rewriting methods and evaluation metrics operate on a sentence level, but ignoring the broader context of the text can lead to preferring generic, ambiguous, and incoherent rewrites. In this paper, we investigate integrating the preceding textual context into both the rewriting and evaluation stages of stylistic text rewriting, and introduce a new composite contextual evaluation metric CtxSimFit that combines similarity to the original sentence with contextual cohesiveness. We comparatively evaluate non-contextual and contextual rewrites in formality, toxicity, and sentiment transfer tasks. Our experiments show that humans significantly prefer contextual rewrites as more fitting and natural over non-contextual ones, yet existing sentence-level automatic metrics (e.g., ROUGE, SBERT) correlate poorly with human preferences (𝜌=0–0.3). In contrast, human preferences are much better reflected by both our novel CtxSimFit (𝜌=0.7–0.9) as well as proposed context-infused versions of common metrics (𝜌=0.4–0.7). Overall, our findings highlight the importance of integrating context into the generation and especially the evaluation stages of stylistic text rewriting.
Anthology ID:
2023.emnlp-main.701
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:
11419–11444
Language:
URL:
https://aclanthology.org/2023.emnlp-main.701
DOI:
10.18653/v1/2023.emnlp-main.701
Bibkey:
Cite (ACL):
Akhila Yerukola, Xuhui Zhou, Elizabeth Clark, and Maarten Sap. 2023. Don’t Take This Out of Context!: On the Need for Contextual Models and Evaluations for Stylistic Rewriting. In Proceedings of the 2023 Conference on Empirical Methods in Natural Language Processing, pages 11419–11444, Singapore. Association for Computational Linguistics.
Cite (Informal):
Don’t Take This Out of Context!: On the Need for Contextual Models and Evaluations for Stylistic Rewriting (Yerukola et al., EMNLP 2023)
Copy Citation:
PDF:
https://aclanthology.org/2023.emnlp-main.701.pdf
Video:
 https://aclanthology.org/2023.emnlp-main.701.mp4