Retrieving musical information from neural data: how cognitive features enrich acoustic ones

Various features – from low-level acoustics, to higher-level statistical regularities, to memory associations – contribute to the experience of musical enjoyment and pleasure. Recent work suggests that musical surprisal, that is, the unexpectedness of a musical event given its context, may directly predict listeners' experiences of pleasure and enjoyment during music listening. Understanding how surprisal shapes listeners' preferences for certain musical pieces has implications for music recommender systems, which are typically content- (both acoustic or semantic) or metadata-based. Here we test a recently developed computational algorithm, called Dynamic-Regularity Extraction (D-REX), that uses Bayesian inference to predict the surprisal that humans experience while listening to music. We demonstrate that the brain tracks musical surprisal as modeled by D-REX by conducting a decoding analysis on the neural signal (collected through magnetoencephalography) of participants listening to music. Thus, we demonstrate the validity of a computational model of musical surprisal, which may remarkably inform the next generation of recommender systems. In addition, we present an open-source neural dataset which will be available for future research to foster approaches combining MIR with cognitive neuroscience, an approach we believe will be a key strategy in characterizing people's reactions to music