Relationship discovery of price movements between sentiment analysis on social media data and stock market
Naqvi, Mohammed Moosa
A desire to make a profit on investment has been a prominent motivational factor in financial investments. The idea of growing with a blue chip firm or an emerging start-up has allured both individual investor(s) and large investing firms alike. One of the financial market areas that gives such opportunity to become part of something bigger is the stock market. Across the globe, stock exchanges become the medium through which billions of stocks are traded on daily basis. Nevertheless, stock market volatility always challenges a seasoned investor to find new ways to invest into stocks that will be profitable in near future. These challenges are equally important for financial firms that are building algorithms for creating profitable stock portfolio. With the advent of social media and similar resonance in digital news media, we have witnessed huge data explosion and this has also opened new opportunities to harvest these data into information for profitable stock trading. In this research, I have performed analysis of more than 8.5 million news article and twitter messages to determine relationship between stock price and media sentiments. Using novel data visualization and Natural Language Processing techniques, I have implemented novel data visualizations such as frequency of news items and other related events affecting the company share price.
Machine Learning, Sentiment Analysis, Computational Finance