The impact of social media on artificial intelligence forecasting models: An application to high-frequency financial data

Predicting future market movements is a goal of many investors and researchers. The fact that there are many factors affecting the behavior of financial markets and the complex interrelationships between these factors requires the joint modeling of many variables obtained from different data sources. The number of studies using social media data together with financial data is increasing. Today, social media is one of the most important communication tools and is influencing many areas. Financial markets are no exception. Five technology companies listed on the NASDAQ Stock Market were selected for this study. Social media data was collected through X and analyzed by calculating many attributes along with stock data. To include the influence of the person who shared the X post and the impact size reached by the post in the algorithm, author influence value and post influence value attributes were calculated using post statistics and these attributes were included in the analysis along with the sentiment values of the posts. Four different machine learning algorithms, namely decision tree, random forest, support vector machine and logistic regression, and LSTM networks were used in the binary and tri-class classification analyses. The algorithms are used to predict the direction of movement of the stock market and these predictions are used in the market simulation to calculate total profitability values. As a result of the analysis, both the effect of social media data was shown, and the most profitable parameter combinations were determined.