인공지능 알고리즘 거문고 산조의 실시간 연주 및 정간보 생성

Real-Time A.I. Algorithm Geomungo Sanjo Player with Jungganbo

Every country has her own culture and it s different from others even though the science is same. With continuous development of computer in 20th century, one of the things that many people has been thinking of, and wanted to implement is a computer-music interface. Since the properties of artificial intelligence such as the random process and non-linearity of the network structure are well matched to those of music which is non-deterministic, coincidental and probabilistic, many computer programs concerning music and sound have been developed. These programs are, however, for much part, only about western music and instruments, but rarely focus on Korean traditional music. Since the view of sound and music between the western and the Korean traditional is very different, it is hard to adopt existing methods to analyze and play Korean traditional music. This thesis is on an Artificial Intelligence System for Real-time A.I. Algorithm Geomungo Sanjo performance and simultaneous visualization of Jungganbo. It is based on Max/msp tools. Geomungo is a Korean traditional string instrument and Jungganbo is a Korean traditional score in Chosun Dynasty. Sanjo is a non-fixed form of Geomungo play, and its non-deterministic and non-patterned style is modeled with random process of A.I. Since Sanjo has large variation in playing, it was hard to be recorded with Jungganbo. Proposed program, however, solved this kind of problem by visualizing corresponding Jungganbo simultaneously.