Restructuring shapes in terms of emergent subshapes: a computational and cognitive model

In order to interpret the current state of a design, designers possess the powerful, spontaneous ability to restructure shapes in terms of emergent subshapes. In this paper a theoretical model of restructuring shapes is presented, one that aims to recognize explicit and implicit emergent subshapes computationally and to explain cognitively some critical phenomena of designers' visual behaviors. A computing procedure using two neuron-like connectionist networks and two attentional techniques serves as the vehicle to mimic some critical empirical phenomena. By a comparison of the computational and empirical variables embedded in the subshape viewing processes, the internal behavior of the model has been further clarified, and some hypothetical viewpoints have been cross-validated.