Computational archival science is a data intensive academic research paradigm of archives born in the context of big data. It is a technical tool for archival governance at the micro level. It is an academic field formed by the intersection of computational science, data science and archival science. Analyzing and studying the causes and development prospects of computational archival science, and analyzing its contradictions with traditional archival science and related disciplines, can point out the direction of the overall development of computational archival science. The paradigm transformation of scientific research is the fundamental, internal, and inevitable reason for the emergence of the concept of computational archival science; The involution of natural science and the small volume of archival science are the direct, external, and accidental reasons for the emergence of the concept of computational archival science. Computational archival science focuses on the analysis of data, which contradicts the research object of traditional archival science; And it is difficult to delimit the boundary between computational archival science based on data analysis and other disciplines. The macro adjustment of computational archival science from the perspective of system, data and algorithm should first solve the problems of data privacy and ethics; The quality and structure of archival data should be improved to make the archival data become big data; It also solves the problem of scholar's information asymmetry and algorithm black box under the condition of double information asymmetry, so that computational archival science can truly become a technical tool that empowers archive governance and ultimately serves national governance.