Computational modeling and analysis of music style evolution

In recent years, there has been growing interest in analyzing temporal changes in musical features using large-scale music datasets. In this talk, I will discuss methods and findings related to the evolutionary analysis of music, grounded in techniques from music information processing and mathematical modeling. Several key challenges are addressed, including the extraction of trends in high-dimensional feature spaces, understanding the temporal evolution of cluster structures, and analyzing the driving forces of evolution and the processes of knowledge reference. Results from analyses of the stylistic evolution of Japanese popular music and other datasets are presented, demonstrating that computational approaches can offer insights complementary to those obtained through traditional humanities-based methods in music history research.