Patterns of Design Cognition in Back Propagation Neural Networks

Design cognition patterns play an important role as a measure of design expertise and cognitive informatics. In this paper, we have applied content-oriented protocol analysis to assess the differences between the cognitive patterns for a given design task between novice and expert designers. Investigating the key cognitive actions that lead to productivity in design, we have developed neural networks to explore their correlations. Our findings indicate that design cognition patterns can be modeled in a back propagation network in which the explicit relations among individual cognitive actions can be investigated. We aim at training the neural network of the novice using the expert´s