Towards a multi-level exploration of human and computational re-representation in unified cognitive frameworks

Abstract

Re-representation is a critical ability to (i) understanding human creative problem solving, and (ii) modelling computational cognitive systems able to support or perform creative problem solving tasks on their own. This paper proposes a unified multi-level cognitive approach to investigating re-representation: the study of sensory-based, concept-based and problem template based possible forms of re-representations in an integrated manner. Descriptions and explanations of each level prepare the ground for further computational modelling. A study is run to explore this hypothesis. A significant correlation between the tasks we investigate is discovered. Two previous studies from the literature are replicated. A new strong and significant relationship between the Pattern Meanings Test and the Alternative Uses Test is observed.