Noisy Text Data: Achilles’ Heel of BERT

Ankit Kumar, Piyush Makhija, Anuj Gupta


Abstract
Owing to the phenomenal success of BERT on various NLP tasks and benchmark datasets, industry practitioners are actively experimenting with fine-tuning BERT to build NLP applications for solving industry use cases. For most datasets that are used by practitioners to build industrial NLP applications, it is hard to guarantee absence of any noise in the data. While BERT has performed exceedingly well for transferring the learnings from one use case to another, it remains unclear how BERT performs when fine-tuned on noisy text. In this work, we explore the sensitivity of BERT to noise in the data. We work with most commonly occurring noise (spelling mistakes, typos) and show that this results in significant degradation in the performance of BERT. We present experimental results to show that BERT’s performance on fundamental NLP tasks like sentiment analysis and textual similarity drops significantly in the presence of (simulated) noise on benchmark datasets viz. IMDB Movie Review, STS-B, SST-2. Further, we identify shortcomings in the existing BERT pipeline that are responsible for this drop in performance. Our findings suggest that practitioners need to be vary of presence of noise in their datasets while fine-tuning BERT to solve industry use cases.
Anthology ID:
2020.wnut-1.3
Volume:
Proceedings of the Sixth Workshop on Noisy User-generated Text (W-NUT 2020)
Month:
November
Year:
2020
Address:
Online
Editors:
Wei Xu, Alan Ritter, Tim Baldwin, Afshin Rahimi
Venue:
WNUT
SIG:
Publisher:
Association for Computational Linguistics
Note:
Pages:
16–21
Language:
URL:
https://aclanthology.org/2020.wnut-1.3
DOI:
10.18653/v1/2020.wnut-1.3
Bibkey:
Cite (ACL):
Ankit Kumar, Piyush Makhija, and Anuj Gupta. 2020. Noisy Text Data: Achilles’ Heel of BERT. In Proceedings of the Sixth Workshop on Noisy User-generated Text (W-NUT 2020), pages 16–21, Online. Association for Computational Linguistics.
Cite (Informal):
Noisy Text Data: Achilles’ Heel of BERT (Kumar et al., WNUT 2020)
Copy Citation:
PDF:
https://aclanthology.org/2020.wnut-1.3.pdf
Optional supplementary material:
 2020.wnut-1.3.OptionalSupplementaryMaterial.zip
Data
GLUEIMDb Movie ReviewsSSTSST-2