Towards a Computational Case-Based Model for Creative Planning

Abstract

This paper describes a computational case-based model for the creative planning process. Our approach is inspired in Wallas’ model for the creative process in that we consider that creativity involves a sequence of four stages: preparation, incubation, illumination and verification. Preparation includes problem acquisition and assimilation of background knowledge, which is represented by cases, i.e., documented past experiences. With the aim of achieving a flexible knowledge representation, as a means to potentiate specific creative abilities like Fluency, Synthesis and Analysis, we structure each case as a network of hierarchically and temporally related case pieces. These case pieces can be considered individually, providing better recombinations of them. These recombinations, rather than made by chance, are guided by those hierarchical and temporal case piece relations (or explanations). We explain the role of opportunistic knowledge acquisition at the incubation stage. We sustain that illumination may comprise recursive calls of the sequence of the first three stages. This computational model is implemented in the system INSPIRER (ImagiNation1 taking as Source Past and Imperfectly RElated Reasonings). An application in musical composition domain is presented. We also show how a musical composition task may be cognitively modelled and treated as a planning task. We also present a short example illustrating how INSPIRER generates music.