Supervisor: Martin Böhringer, Student: Erik Hosang

Aim: Monitoring of information streams (like microblogging) can be a time-consuming task. Therefore automated supporting agents would be of great help for users. However, due to missing semantics such text-only postings are often hard to interpret automatically. Natural Language Processing (NLP) and similar disciplines try to solve that problem. The thesis answers the question, if they could help in the information stream / microblogging case.

Content: The thesis consists of 2 parts: (1) evaluates the basics text analysis (i.e. NLP) and presents current solutions (an example is the service akibot.com) (2) systematically answers the question, which algorithms and/or services (there are services like Open Calais and others) could support especially small-text information streams like we found them in microblogging. For this purpose, theoretical and experimental argumentation should be used.

The thesis can be written either in German or English.