Don’t Change Me! User-Controllable Selective Paraphrase Generation

Mohan Zhang, Luchen Tan, Zihang Fu, Kun Xiong, Jimmy Lin, Ming Li, Zhengkai Tu


Abstract
In the paraphrase generation task, source sentences often contain phrases that should not be altered. Which phrases, however, can be context dependent and can vary by application. Our solution to this challenge is to provide the user with explicit tags that can be placed around any arbitrary segment of text to mean “don’t change me!” when generating a paraphrase; the model learns to explicitly copy these phrases to the output. The contribution of this work is a novel data generation technique using distant supervision that allows us to start with a pretrained sequence-to-sequence model and fine-tune a paraphrase generator that exhibits this behavior, allowing user-controllable paraphrase generation. Additionally, we modify the loss during fine-tuning to explicitly encourage diversity in model output. Our technique is language agnostic, and we report experiments in English and Chinese.
Anthology ID:
2021.eacl-main.307
Volume:
Proceedings of the 16th Conference of the European Chapter of the Association for Computational Linguistics: Main Volume
Month:
April
Year:
2021
Address:
Online
Editors:
Paola Merlo, Jorg Tiedemann, Reut Tsarfaty
Venue:
EACL
SIG:
Publisher:
Association for Computational Linguistics
Note:
Pages:
3522–3527
Language:
URL:
https://aclanthology.org/2021.eacl-main.307
DOI:
10.18653/v1/2021.eacl-main.307
Bibkey:
Cite (ACL):
Mohan Zhang, Luchen Tan, Zihang Fu, Kun Xiong, Jimmy Lin, Ming Li, and Zhengkai Tu. 2021. Don’t Change Me! User-Controllable Selective Paraphrase Generation. In Proceedings of the 16th Conference of the European Chapter of the Association for Computational Linguistics: Main Volume, pages 3522–3527, Online. Association for Computational Linguistics.
Cite (Informal):
Don’t Change Me! User-Controllable Selective Paraphrase Generation (Zhang et al., EACL 2021)
Copy Citation:
PDF:
https://aclanthology.org/2021.eacl-main.307.pdf
Data
MS COCOParaBank