In their own words

SNTMNT is an online tool that gives daily insights into online consumer sentiment surrounding 25 AEX bluechip funds and the index itself. Recent research (Sprenger and Welpe, 2010; Bollen, et al, 2010, Van Leeuwen, 2011) has shown that online sentiment can be used successfully as an early indicator of market movement (bullish / bearish) and trading volume due the fact that online media can transmit news quicker than traditional media does.

We use over 10,000 sources including Twitter, Facebook and leading news Websites and Blogs to determine the emotion of online news. Together with Clip It, we classify all online news in five different mood states with over 80% accuracy. We’ve aggregated these mood states into one indicator of online consumer sentiment behavior.

Next to insights into online sentiment we’re predictive trend analysis. We use data mining and machine learning algorithms in order to predict stock market movement between 1 and 7 days in the future through a confidence interval. This confidence interval is the percentage that explains the share price variance of a specific fund: to which extent can online media sentiment predict bullish (up) or bearish (down) share price movement. With the current (limited) dataset and algorithms we can explain up to ten percent of share price variance with a p-value under 0.01 (99% confident).