Abstract: In this paper, we present a multi-task learning (MTL) model to classify sentiment and emotion in Bengali and English languages. For this multi-task learning work, different Bengali and English datasets were collected from publicly available sources and developed two MTL models by utilizing pre-trained mBERT and MuRIL models. Our proposed MTL model outperforms their corresponding standalone classifiers with an average F1-score of 0.5728 (+0.041) and 0.7590 (+0.046) for Bengali sentiment, emotion and English sentiment, emotion classification tasks respectively.
Paper Type: Short
Research Area: Machine Learning for NLP
Research Area Keywords: multi-task learning
Contribution Types: NLP engineering experiment
Languages Studied: Bengali, English
Submission Number: 5018
Loading