Towards an intelligent rhythmic generator based on given examples : a memetic approach

In this paper we present an artificial intelligence system inspired by Richard Dawkin's theory of memes for the composition of rhythmic passages. Rhythmic composition is understood as the process of interconnecting sequences of basic elements (or "rhythmic memes"). The overall design of the system consists of two broad stages: the learning stage and the production stage. In the learning stage, which is the main focus of this paper, the system is trained with examples of musical pieces in order to evolve a “musical worldview”. We study the dynamics of this evolution by analyzing the behaviour of the memes logged during the learning and the interaction processes. A series of learning simulations is presented using rhythmic information taken from Brazilian composers of different styles.