Text Attribute Control via Closed-Loop Disentanglement

Lei Sha, Thomas Lukasiewicz


Abstract
Changing an attribute of a text without changing the content usually requires first disentangling the text into irrelevant attributes and content representations. After that, in the inference phase, the representation of one attribute is tuned to a different value, expecting that the corresponding attribute of the text can also be changed accordingly. The usual way of disentanglement is to add some constraints on the latent space of an encoder-decoder architecture, including adversarial-based constraints and mutual-information-based constraints. However, previous semi-supervised processes of attribute change are usually not enough to guarantee the success of attribute change and content preservation. In this paper, we propose a novel approach to achieve a robust control of attributes while enhancing content preservation. In this approach, we use a semi-supervised contrastive learning method to encourage the disentanglement of attributes in latent spaces. Differently from previous works, we re-disentangle the reconstructed sentence and compare the re-disentangled latent space with the original latent space, which makes a closed-loop disentanglement process. This also helps content preservation. In addition, the contrastive learning method is also able to replace the role of minimizing mutual information and adversarial training in the disentanglement process, which alleviates the computation cost. We conducted experiments on three text datasets, including the Yelp Service review dataset, the Amazon Product review dataset, and the GoEmotions dataset. The experimental results show the effectiveness of our model.
Anthology ID:
2024.tacl-1.11
Volume:
Transactions of the Association for Computational Linguistics, Volume 12
Month:
Year:
2024
Address:
Cambridge, MA
Venue:
TACL
SIG:
Publisher:
MIT Press
Note:
Pages:
190–209
Language:
URL:
https://aclanthology.org/2024.tacl-1.11
DOI:
10.1162/tacl_a_00640
Bibkey:
Cite (ACL):
Lei Sha and Thomas Lukasiewicz. 2024. Text Attribute Control via Closed-Loop Disentanglement. Transactions of the Association for Computational Linguistics, 12:190–209.
Cite (Informal):
Text Attribute Control via Closed-Loop Disentanglement (Sha & Lukasiewicz, TACL 2024)
Copy Citation:
PDF:
https://aclanthology.org/2024.tacl-1.11.pdf