Semantic Label Smoothing for Sequence to Sequence Problems

Michal Lukasik, Himanshu Jain, Aditya Menon, Seungyeon Kim, Srinadh Bhojanapalli, Felix Yu, Sanjiv Kumar


Abstract
Label smoothing has been shown to be an effective regularization strategy in classification, that prevents overfitting and helps in label de-noising. However, extending such methods directly to seq2seq settings, such as Machine Translation, is challenging: the large target output space of such problems makes it intractable to apply label smoothing over all possible outputs. Most existing approaches for seq2seq settings either do token level smoothing, or smooth over sequences generated by randomly substituting tokens in the target sequence. Unlike these works, in this paper, we propose a technique that smooths over well formed relevant sequences that not only have sufficient n-gram overlap with the target sequence, but are also semantically similar. Our method shows a consistent and significant improvement over the state-of-the-art techniques on different datasets.
Anthology ID:
2020.emnlp-main.405
Volume:
Proceedings of the 2020 Conference on Empirical Methods in Natural Language Processing (EMNLP)
Month:
November
Year:
2020
Address:
Online
Venue:
EMNLP
SIG:
Publisher:
Association for Computational Linguistics
Note:
Pages:
4992–4998
Language:
URL:
https://aclanthology.org/2020.emnlp-main.405
DOI:
10.18653/v1/2020.emnlp-main.405
Bibkey:
Cite (ACL):
Michal Lukasik, Himanshu Jain, Aditya Menon, Seungyeon Kim, Srinadh Bhojanapalli, Felix Yu, and Sanjiv Kumar. 2020. Semantic Label Smoothing for Sequence to Sequence Problems. In Proceedings of the 2020 Conference on Empirical Methods in Natural Language Processing (EMNLP), pages 4992–4998, Online. Association for Computational Linguistics.
Cite (Informal):
Semantic Label Smoothing for Sequence to Sequence Problems (Lukasik et al., EMNLP 2020)
Copy Citation:
PDF:
https://aclanthology.org/2020.emnlp-main.405.pdf
Video:
 https://slideslive.com/38939196