ATTA: Automatic Time-Span Tree Analyzer Based on Extended GTTM

This paper describes a music analyzing system called the automatic time-span tree analyzer (ATTA), which we have developed. The ATTA derives a time-span tree that assigns a hierarchy of 'structural importance' to the notes of a piece of music based on the generative theory of tonal music (GTTM). Although the time-span tree has been applied with music summarization and collaborative music creation systems, these systems use time-span trees manually analyzed by experts in musicology. Previous systems based on GTTM cannot acquire a timespan tree without manual application of most of the rules, because GTTM does not resolve much of the ambiguity that exists with the application of the rules. To solve this problem, we propose a novel computational model of the GTTM that re-formalizes the rules with computer implementation. The main advantage of our approach is that we can introduce adjustable parameters, which enables us to assign priority to the rules. Our analyzer automatically acquires time-span trees by configuring the parameters that cover 26 rules out of 36 GTTM rules for constructing a time-span tree. Experimental results showed that after these parameters were tuned, our method outperformed a baseline performance. We hope to distribute the time-span tree as the content for various musical tasks, such as searching and arranging music.