Analysis of music similarity based on Pearson correlation coefficient
Zhiqun Li, Lusheng Wang, Dandan Wang · 2021
In order to understand the role of music in our human lives, it is very important to develop a method that can quantify the evolution of music. First, we selected seven indicators of music characteristics, used Pearson's correlation coefficient to construct a music similarity measurement model, and performed a cluster analysis of artists within the genre to solve the Pearson correlation coefficient within and between genres. It is concluded that the music similarity within genres is higher than that between genres. Next, we selected some indicators through principal component analysis, observed the change curve of indicators over time, and analyzed how the genre changes over time. Finally, we use the coefficient of variation method to analyze the musical characteristics of influencers and followers, and obtain the relationship between the appeal of musical characteristics.