EmpTwit gauges market sentiment via Twitter streams. With the aid of the MIT common sense knowledge base, Open Mind, and affect extension of Wordnet, EmpTwit sets out to abstract variation in public mood and investor sentiment via Twitter updates. By using natural language parsing techniques to extract significant concepts in tweets, EmpTwit sets out to capture more accurate sentiment beyond simple keyword matching techniques employed in the field.
The goal of EmpTwit is to correlate the variation in public mood with stock market behavior. The idea is that EmpTwit will measure investor sentiment before the market fully adjusts to it. The motivation of this project stems from various psychological evidences of a strong link between weather and sports outcomes with investor sentiment and market behavior.