A Model Comparison for Chord Prediction on the Annotated Beethoven Corpus

Abstract

This paper models predictive processing of chords using a corpus of Ludwig van Beethoven’s string quartets. A recently published dataset consisting of expert harmonic analyses of all Beethoven string quartets was used to evaluate an $n$-gram language model as well as a recurrent neural network (RNN) architecture based on long-short-term memory (LSTM). We compare model performances over different periods of Beethoven’s creative activity and provide a baseline for future research on predictive processing of chords in full Roman numeral representation on this dataset

Publication
Proceedings of the 16th Sound & Music Computing Conference (SMC 2019)
Fabian C. Moss
Fabian C. Moss
Research Fellow in Cultural Analytics

Fabian C. Moss is a Research Fellow in Cultural Analytics at University of Amsterdam (UvA). He was born in Cologne, Germany, and studied Mathematics and Educational Studies at University of Cologne, and Music Education (Major Piano) and Musicology at Hochschule für Musik und Tanz, Köln. He obtained is PhD in Digital Humanities from École Polytechnique Fédérale de Lausanne (EPFL). Working with large symbolic datasets of musical scores and harmonic annotations, he is primarily interested in Computational Music Analysis, Music Theory, Music Cognition, and their mutual relationship.