Sentiment Analysis of Public Health Concerns of Tokyo 2020 Olympics Using LSTM

In recent years, in this age of social media, the Tokyo Olympic Games 2020 has received massive online sentiments and feedback from individuals, despite the Tokyo Olympic Games 2020’s motto of “united by emotions.” In this paper, a Recurrent Neural Network model was trained and implemented for the Tokyo 2020 Olympics public health sentiment analysis. A total of ten thousand (10,000) tweets were collected and trained, with tweets classified as positive, negative, or neutral. The RNN model implemented was the LSTM model. The model's performance was evaluated using performance metrics such as accuracy, precision, recall, and F1-score. According to the results, the LSTM model performed effectively, with a final validation accuracy of 88.2% and precision values of 0.86, 0.90, and 0.89 for negative, neutral, and positive predictions, respectively. The results of the LSTM model's high accuracy indicate that the model successfully determined the public health sentiments of the Tokyo 2020 Olympics, which were found to be positive at 40.39%.