On Generating of Proofs

Abstract

The paper introduces an approach to automated deduction. Described is a method for generating of proofs which are considered as finite sequences of formulas. Unlike the most common approaches to automated deduction which are based mainly on the resolution principle, our method resembles principles of generating of programs which are usually used in genetic programming. We present an overview of the method and present two case studies. We argue that even in its preliminary stage, our method can be helpful to experts who need to find proofs or counterexamples.