Artificial Intelligence Approaches to Music Composition

Artificial Intelligence (AI) techniques have been successfully applied to a wide range of problems that perform problem solving such as diagnosis, decision making and optimization problems. However, any AI algorithm applied to a creative problem requires some mechanism to substitute for the creative spark found in the human, as the computer has no creative capacity. Randomness, as supplied by a random number generator, cannot be the sole mechanism to bring about a creative composition. This paper examines three AI approaches applied to music composition. Specifically, the paper introduces MAGMA, a knowledge-based system that uses three different AI algorithms to generate music. Sample songs generated by MAGMA are compared.