Musicat: a computational model of creativity in a musical domain

点击翻译按钮获取中文摘要

This project takes a new approach to understanding human creativity, based on cognitive modeling of creative thought in a musical domain. Cognitive models developed by the Fluid Analogies Research Group (FARG) at the Center for Research on Concepts and Cognition (CRCC) in the past three decades have provided new insight into a wide variety of creative tasks such as solving analogy problems, extrapolating number sequences, and designing typefaces. The domain of music provides an opportunity to extend these techniques for the study of fundamental mechanisms of cognition and creativity. Specifically, the proposed research aims to develop a novel model of music cognition that simulates the creative processes involved in music ? not, however, in the composition or improvisation of music, but in the seemingly much simpler and more common act of music perception. How is music perception creative? Perception is an active process where the brain interprets sensory inputs, and transforms the raw input into a compressed format. Far from being a passive, robotic algorithm, this process of interpretation is quite dynamic and influenced by cognitive context. Such perception is not only subjective, but creative, as evidenced by its generation of novel internal representations of sensory data. A central tenet of the philosophy underlying this work is that the exact same processes that drive the creative perception of sensory input also give rise to the high-level generation of creative thoughts, ideas, and works of art.