The Performance Evaluation of Textual Analysis Tools in Financial Markets

The rapid development of textual analysis techniques has paved the way for automatic sentiment analysis. Many studies have been focusing on using sentiments disclosed from social media to predict the financial market performance. However, the underlying value of the messages collected from social media is still not fully understood due to lack of comprehensive apprehending of the performance for different textual analysis methods. In this research, we shed light on this problem by comparing the different classification accuracy of dictionaries and machine-learning techniques in the financial context