Two-Headed Monster and Crossed Co-Attention Networks

Yaoyiran Li, Jing Jiang


Abstract
This paper investigates a new co-attention mechanism in neural transduction models for machine translation tasks. We propose a paradigm, termed Two-Headed Monster (THM), which consists of two symmetric encoder modules and one decoder module connected with co-attention. As a specific and concrete implementation of THM, Crossed Co-Attention Networks (CCNs) are designed based on the Transformer model. We test CCNs on WMT 2014 EN-DE and WMT 2016 EN-FI translation tasks and show both advantages and disadvantages of the proposed method. Our model outperforms the strong Transformer baseline by 0.51 (big) and 0.74 (base) BLEU points on EN-DE and by 0.17 (big) and 0.47 (base) BLEU points on EN-FI but the epoch time increases by circa 75%.
Anthology ID:
2020.aacl-srw.2
Volume:
Proceedings of the 1st Conference of the Asia-Pacific Chapter of the Association for Computational Linguistics and the 10th International Joint Conference on Natural Language Processing: Student Research Workshop
Month:
December
Year:
2020
Address:
Suzhou, China
Editors:
Boaz Shmueli, Yin Jou Huang
Venue:
AACL
SIG:
Publisher:
Association for Computational Linguistics
Note:
Pages:
8–15
Language:
URL:
https://aclanthology.org/2020.aacl-srw.2
DOI:
Bibkey:
Cite (ACL):
Yaoyiran Li and Jing Jiang. 2020. Two-Headed Monster and Crossed Co-Attention Networks. In Proceedings of the 1st Conference of the Asia-Pacific Chapter of the Association for Computational Linguistics and the 10th International Joint Conference on Natural Language Processing: Student Research Workshop, pages 8–15, Suzhou, China. Association for Computational Linguistics.
Cite (Informal):
Two-Headed Monster and Crossed Co-Attention Networks (Li & Jiang, AACL 2020)
Copy Citation:
PDF:
https://aclanthology.org/2020.aacl-srw.2.pdf