Genetic Algorithms for Extension Search in Default Logic

Abstract

A default theory can be characterized by its sets of plausible conclusions, called its extensions. But, due to the theoretical complexity of Default Logic (Sigma_2p-complete), the problem of finding such an extension is very difficult if one wants to deal with non trivial knowledge bases. Based on the principle of natural selection, Genetic Algorithms have been quite successfully applied to combinatorial problems and seem useful for problems with huge search spaces and when no tractable algorithm is available. The purpose of this paper is to show that techniques issued from Genetic Algorithms can be used in order to build an efficient default reasoning system. After providing a formal description of the components required for an extension search based on Genetic Algorithms principles, we exhibit some experimental results.