Predicting Dow Jones Movement with Twitter

The use of machine learning in the realm of finance is becoming much more prevalent as algorithmic trading catches on. Similarly, online social networking data becomes more valuable with new research in mining. The goal of this project is to take the next steps in these directions. Using different methods to interpret Twitter content, we hope to predict movement in the Dow Jones Industrial Average. More specifically, we find the word counts of our corpus of tweets using different word lists, calculate high-level features, and use machine learning techniques to see whether a correlation with the stock market is likely to exist.