Tweeting the Market: Leveraging LSTM for Stock Price Prediction

Investors and traders constantly seek an edge, a source of insight that can help them predict market movements more accurately. Over the past several years, the emergence of social media and the increase in use of platforms such as Twitter, Facebook, etc. has given birth to a new dimension of market analysis - sentiment analysis and its potential effect on stock prices. The mood dimensions that correlate with market movements can be complex and multifaceted. Only a few researches have studied the effect of social media on stock price predictions. Prior investigations have examined discussions surrounding the COVID-19 virus, its societal repercussions, and its potential sway over financial markets. This research has specifically scrutinized the effect of negative sentiments and their potential repercussions on stock prices of major tech companies like Microsoft, Google, and Tesla. The paper aims to identify the most effective model among ARIMA and LSTM for predicting stock prices by correlating them with sentiments expressed in social media posts. The motive of this paper is to help investors better analyze the market and adjust their strategies accordingly to earn more profits.