Musical Phrase Segmentation via Grammatical Induction

TL;DR

This paper presents a novel approach to musical phrase segmentation using grammatical induction algorithms, demonstrating that the LONGESTFIRST algorithm with duration viewpoints outperforms others across multiple datasets.

Contribution

It introduces the application of grammatical induction algorithms to musical phrase segmentation and identifies the most effective algorithm and encoding for this task.

Findings

LONGESTFIRST achieves the highest F1 scores

Including duration viewpoints improves performance

Performance varies across datasets

Abstract

We outline a solution to the challenge of musical phrase segmentation that uses grammatical induction algorithms, a class of algorithms which infer a context-free grammar from an input sequence. We analyze the performance of five grammatical induction algorithms on three datasets using various musical viewpoint combinations. Our experiments show that the LONGESTFIRST algorithm achieves the best F1 scores across all three datasets and that input encodings that include the duration viewpoint result in the best performance.