Generation of creative game scene patterns by the neutrosophic genetic CoCoSo method

Abstract

Creative procedural generation requires, by its nature, a nondeterministic result. Each set of outputs has to satisfy the required criteria and, at the same time, has to show unique results to be considered creative. One of the keys to increasing nondeterminism is to use fuzzy neutrosophic sets. A combination of a CoCoSo multi-criteria decision algorithm with a genetic algorithm to generate game scenes was proposed. The algorithm satisfies the required game rules and improves the aesthetics value of the visual game scene. The proposed approach seeks to produce a creative solution which satisfies the replayability principle, when the solution is not optimal in the optimization procedure sense. In this paper, improvements of the CoCoSo method were proposed in order to work with iterative genetic algorithm.