[Purpose/significance] Virtual digital human are becoming the new darling of various industries in the era of digital economy. This study deeply identifies and analyzes the driving factors of virtual digital human users' online participation behavior from user demand perspective, which provides reference for the theoretical research and optimi⁃
zation of operation practice of virtual digital human. [Method/process] This study collects the bullet screen text re⁃
leased by the users of Bilibili in the original videos of typical virtual digital human such as Hatsune Miku, Luo Tianyi
and Liu Yexi. Based on the computational grounded theory, combined with the uses and gratifications theory, this study
extracts the demand factors of virtual digital human user's online participation behavior, then analyzes the relationships
between virtual digital human user's demand and the influence paths on online participation behavior. [Result/conclu⁃
sion] Virtual digital human users’demand includes emotional demand, social demand, content demand and aesthetic
acceptance demand, which all affect their online participation behavior. Content demand, aesthetic acceptance demand
and social demand affect users'emotional demand. Several useful optimization strategies from users’demand perspec⁃
tive are accordingly put forward, such as shaping rich and three-dimensional character setting, creating substantial and
interesting personalized content, and providing real-time and continuous emotional relationship.