One of the challenges of today's information society is to cope with the immense amount of digitally stored texts, as for instance provided by the internet. It would already be useful to have a tool, which allows verifying or refuting a hypothesis from the information given in a text. The goal of this thesis is to develop a prototype of such a tool, which functions by mapping the sentences of the text to PROLOG-like rules. This work is done in co-operation with the Computational Linguistics Department of Saarland University. Furthermore, in order to allow common sense inferences, world knowledge taken from the Suggested Upper Model Ontology (SUMO) is added. These data is given to a first-order theorem-prover (KRHyper), which will then try to verify or refute the hypothesis.