Analyzing textual information using semantic methods has become a necessity in order to control the increasing amount of information and its condensed presentation on the small displays of mobile devices. A large knowledge base is a prerequisite for such algorithms. In this work, we present a new algorithm which uses the structured database Freebase and the entity extraction from the Ontonaut web service for classifying textual information and show the functioning of this algorithm as part of the Ontonaut smartphone app. The algorithm automatically extracts and transforms data from Freebase to a knowledge base and uses this knowledge base for classification. The weighted Freebase types used for the classification allow a detailed representation of the topic of the text, which is especially useful for applications, which use Freebase as a data source.

Original title:

Bachelor thesis written by Martin Jakobus, Mai 2012.

Second supervisor: Dipl.-Ing. Christoph Diefenthal, CTO Top21 GmbH, Stuttgart