Performance Analysis of Indian Stock Market via Sentiment Analysis and Historical Data

Abstract

Stock market has always generated interest among the people since the time of its initiation. It is a very complex and challenging system, where people invest to gain money in order to attain higher gains. But the scenario may be negative if the investment is made in the stocks without any proper analysis. The performance of any stock depends on many parameters and factors like historical prices, social media data, news, country economics, production of the company, etc. In our research, we consider two major factors like historical prices and social media data and will let the investors give an idea about the stock performance in the nearby future. Therefore, we combine the sentiments of the different stakeholders across the Internet with historic prices of the stock to predict the stock recital. For combining the above approaches, we are using the decision tree approach of machine learning for classification and prediction for more accurate prophecy. The proposed algorithm gives above 70% accuracy for the given data.