Automated Planning and Music Composition: an Efficient Approach for Generating Musical Structures

Automatic music composition is a fascinating field within computational creativity. While different Artificial Intelligence techniques have been used for tackling this task, Planning – an approach for solving complex combinatorial problems which can count on a large number of high-performance systems and an expressive language for describing problems – has never been exploited. In this paper, we propose two different techniques that rely on automated planning for generating musical structures. The structures are then filled from the bottom with “raw” musical materials, and turned into melodies. Music experts evaluated the creative output of the system, acknowledging an overall human-enjoyable trait of the melodies produced, which showed a solid hierarchical structure and a strong musical directionality. The techniques proposed not only have high relevance for the musical domain, but also suggest unexplored ways of using planning for dealing with non-deterministic creative domains.