Select and Attend: Towards Controllable Content Selection in Text Generation

Xiaoyu Shen, Jun Suzuki, Kentaro Inui, Hui Su, Dietrich Klakow, Satoshi Sekine


Abstract
Many text generation tasks naturally contain two steps: content selection and surface realization. Current neural encoder-decoder models conflate both steps into a black-box architecture. As a result, the content to be described in the text cannot be explicitly controlled. This paper tackles this problem by decoupling content selection from the decoder. The decoupled content selection is human interpretable, whose value can be manually manipulated to control the content of generated text. The model can be trained end-to-end without human annotations by maximizing a lower bound of the marginal likelihood. We further propose an effective way to trade-off between performance and controllability with a single adjustable hyperparameter. In both data-to-text and headline generation tasks, our model achieves promising results, paving the way for controllable content selection in text generation.
Anthology ID:
D19-1054
Volume:
Proceedings of the 2019 Conference on Empirical Methods in Natural Language Processing and the 9th International Joint Conference on Natural Language Processing (EMNLP-IJCNLP)
Month:
November
Year:
2019
Address:
Hong Kong, China
Venues:
EMNLP | IJCNLP
SIG:
SIGDAT
Publisher:
Association for Computational Linguistics
Note:
Pages:
579–590
Language:
URL:
https://aclanthology.org/D19-1054
DOI:
10.18653/v1/D19-1054
Bibkey:
Cite (ACL):
Xiaoyu Shen, Jun Suzuki, Kentaro Inui, Hui Su, Dietrich Klakow, and Satoshi Sekine. 2019. Select and Attend: Towards Controllable Content Selection in Text Generation. In Proceedings of the 2019 Conference on Empirical Methods in Natural Language Processing and the 9th International Joint Conference on Natural Language Processing (EMNLP-IJCNLP), pages 579–590, Hong Kong, China. Association for Computational Linguistics.
Cite (Informal):
Select and Attend: Towards Controllable Content Selection in Text Generation (Shen et al., EMNLP 2019)
Copy Citation:
PDF:
https://aclanthology.org/D19-1054.pdf
Attachment:
 D19-1054.Attachment.pdf
Code
 chin-gyou/controllable-selection
Data
WikiBio