Optimized Hybrid Model for COVID-19 Vaccine Sentiment Analysis for Hindi Text

A new virus disease called COVID-19, which first appeared in 2019, caused a global economic and health disaster. Millions of people infected with COVID-19 worldwide since March 2020. Vaccination is a potentially powerful tool for establishing herd immunity. This research aims to understand the opinions of Hindi-speaking public about the COVID-19 vaccine. A deep learning-based hybrid model with a metaheuristic-based whale optimization algorithm is proposed in this work. Two datasets of Hindi tweets about COVID-19 vaccination during the epidemic are analyzed. In the first dataset, total of 21,433 tweets are collected from January 2021 to August 2021. The second dataset contains total 14,689 tweets collected between March 2022 to July 2022. Original tweets are pre-processed by applying natural language processing techniques. In addition, the whale optimization approach has been utilized to pick the best features. Finally, CNN, GRU, and hybrid model(CNN+GRU) is utilized to classify the sentiments. The classification results are analyzed with all extracted features and optimal features both. The proposed model gives the highest 92.69%, 94.25%, 94.74%, 95.47% of precision, F-score, recall, and accuracy, respectively, on the first data set with optimal features. It gives highest 92.73%, 93.50%, 92.11%, 94.41% of precision, F-score, recall, and accuracy, respectively, on the second dataset. © 2023, The Author(s), under exclusive licence to Springer Nature Singapore Pte Ltd.