This thesis deals with the use of Twitter for musical sonification. Based on reviews of previous works in sociology, psychology, musicology, and computer music, musical potential of Twitter as a real-time social network service (SNS) and large-scale database is presented with the idea of structural analogy with music. Several sonification mapping strategies to reflect relationships in the data space in the auditory domain are then suggested and analyzed through detailed evaluations.
As a result, Twitmusic - sonification software which utilizes both Twitter"s accumulated data as well as its real-time messages - is presented as a tool for general users to design their own mappings and generate music from auditory display of tweets. Result of user tests shows that sonified results with tweet-improvised phrases generated by Twitmusic using proposed mappings are not only appreciable as a musical piece but also potentially powerful for representing the characteristics of specific features of Twitter.