We have recently started a project to explore the information content of microblogs and study in how far microblogging platforms such as Twitter can be used as an indicator of events in the offline world.

I am fascinated by the use of electronic platforms to leverage collective intelligence. This includes both tools specifically esigned for that purpose (such as prediction markets), but also social media (e.g., microblogging platforms) where the content may also be interpreted as an information market.

With a recent focus on the latter of these two areas, we have started a project to explore in how far microblogging platforms such as Twitter can be used as an indicator of events in the offline world. We have already shown that Twitter messages can be used to predict election results and reflect the political landscape quite well. Due to my professional background in the financial industry, I would like to examine whether Twitter can serve as an indicator of financial market activity. Other research areas that interest me are the use of microblogging as a channel for corporate communication and the reflection of brand sentiment on Twitter.

Methodologically, we are planning to use dictionary-based and machine learning tools to classify message board content automatically and so extract the sentiment contained in the postings.