Interactive Visualization and Manipulation of Attention-based Neural Machine Translation

Jaesong Lee, Joong-Hwi Shin, Jun-Seok Kim


Abstract
While neural machine translation (NMT) provides high-quality translation, it is still hard to interpret and analyze its behavior. We present an interactive interface for visualizing and intervening behavior of NMT, specifically concentrating on the behavior of beam search mechanism and attention component. The tool (1) visualizes search tree and attention and (2) provides interface to adjust search tree and attention weight (manually or automatically) at real-time. We show the tool gives various methods to understand NMT.
Anthology ID:
D17-2021
Volume:
Proceedings of the 2017 Conference on Empirical Methods in Natural Language Processing: System Demonstrations
Month:
September
Year:
2017
Address:
Copenhagen, Denmark
Editors:
Lucia Specia, Matt Post, Michael Paul
Venue:
EMNLP
SIG:
SIGDAT
Publisher:
Association for Computational Linguistics
Note:
Pages:
121–126
Language:
URL:
https://aclanthology.org/D17-2021
DOI:
10.18653/v1/D17-2021
Bibkey:
Cite (ACL):
Jaesong Lee, Joong-Hwi Shin, and Jun-Seok Kim. 2017. Interactive Visualization and Manipulation of Attention-based Neural Machine Translation. In Proceedings of the 2017 Conference on Empirical Methods in Natural Language Processing: System Demonstrations, pages 121–126, Copenhagen, Denmark. Association for Computational Linguistics.
Cite (Informal):
Interactive Visualization and Manipulation of Attention-based Neural Machine Translation (Lee et al., EMNLP 2017)
Copy Citation:
PDF:
https://aclanthology.org/D17-2021.pdf