Korean Court Music

Korean court music, a rich tradition with historical significance, is undergoing renewed scrutiny through computational analysis. Researchers are employing machine learning techniques, such as transformer models and convolutional neural networks, to analyze existing musical notations and large datasets of folk songs, aiming to understand and reconstruct historical pieces and uncover underlying compositional structures. This work utilizes novel encoding schemes and self-supervised learning to address challenges posed by limited data and diverse musical styles, revealing insights into melodic structures and compositional principles. These computational approaches offer valuable tools for preserving and interpreting this important cultural heritage.