HinglishNLP at SemEval-2020 Task 9: Fine-tuned Language Models for Hinglish Sentiment Detection

Meghana Bhange, Nirant Kasliwal


Abstract
Sentiment analysis for code-mixed social media text continues to be an under-explored area. This work adds two common approaches: fine-tuning large transformer models and sample efficient methods like ULMFiT. Prior work demonstrates the efficacy of classical ML methods for polarity detection. Fine-tuned general-purpose language representation models, such as those of the BERT family are benchmarked along with classical machine learning and ensemble methods. We show that NB-SVM beats RoBERTa by 6.2% (relative) F1. The best performing model is a majority-vote ensemble which achieves an F1 of 0.707. The leaderboard submission was made under the codalab username nirantk, with F1 of 0.689.
Anthology ID:
2020.semeval-1.119
Volume:
Proceedings of the Fourteenth Workshop on Semantic Evaluation
Month:
December
Year:
2020
Address:
Barcelona (online)
Venue:
SemEval
SIGs:
SIGLEX | SIGSEM
Publisher:
International Committee for Computational Linguistics
Note:
Pages:
934–939
Language:
URL:
https://aclanthology.org/2020.semeval-1.119
DOI:
10.18653/v1/2020.semeval-1.119
Bibkey:
Cite (ACL):
Meghana Bhange and Nirant Kasliwal. 2020. HinglishNLP at SemEval-2020 Task 9: Fine-tuned Language Models for Hinglish Sentiment Detection. In Proceedings of the Fourteenth Workshop on Semantic Evaluation, pages 934–939, Barcelona (online). International Committee for Computational Linguistics.
Cite (Informal):
HinglishNLP at SemEval-2020 Task 9: Fine-tuned Language Models for Hinglish Sentiment Detection (Bhange & Kasliwal, SemEval 2020)
Copy Citation:
PDF:
https://aclanthology.org/2020.semeval-1.119.pdf
Code
 NirantK/Hinglish +  additional community code
Data
SentiMix