Twitter Sentiment Analysis of COVID-19 Vaccine Based on BiLSTM with Attention Mechanism

At the end of 2019, COVID-19 started to rage and the vaccine became the most effective solution. the COVID-19 vaccine triggered a high level of attention from social media. twitter became one of the main platforms for the public to post their attitudes and tendencies towards the COVID-19 vaccine, providing a text corpus for public opinion research based on public sentiment analysis. In this paper, we use Twitter posted between August 2020 and August 2021 as the information base. We use the BiLSTM with Attention Mechanism algorithm to calculate the sentiment value of the data, then we use the temporal and spatial dimensions to mine the topics users pay attention to. The results show that positive or negative public sentiment toward vaccines is influenced by external influences (e.g., vaccine safety events, increase in the number of patients receiving vaccines, etc.) and varies by region. The more severe the outbreak (e.g., India, the United States, etc.), the higher the user engagement. The lower the sentiment status and predicted value, the greater the impact on the trend of opinion evolution. In opinion-related networks, the level of engagement with the topic also varies by region and is closely related to the severity of the epidemic, the extent of vaccine promotion and the evolution of topical events. Understanding vaccination sentiment through Twitter can help institutions improve vaccination by keeping abreast of public sentiment about the new crown vaccine.