My PhD research is in the field of sentiment analysis and recognizing economic entities. How do people feel about certain products, companies, or services? This is in line with my interest in natural language processing applications. How can we make a computer understand our language and make practical use of that capability?
In the context of the COMMIT project Information Retrieval for Information Services (COMMIT/Infiniti), I am investigating the Mining of Economic Entities. As in the Web 2.0 age people easily share their sentiment regarding companies and products, reputation modeling has become increasingly important for commercial and non-commercial organizations. Due to the huge amount of opinions available through the Web, a manual approach for social media analysis for sentiment mining seems unfeasible, while there seems to be a huge potential for content-driven automated approaches.
For an excellent introduction to sentiment analysis and its (current) applications, I'd like to refer to Ronen Feldman's article in Communications of the ACM.