[Purpose/significance] Aiming at shortcomings of the current service quality of digital archives, such as insufficient intelligence and single service content, this paper proposed to build an overall framework of knowledge graph of digital archives, so as to achieve statistical analysis of big data for digital archive, integration and optimization of digital archiving resources, and improvement on the overall service level of digital archives. [Method/process] By collecting literatures, service deficiencies of digital archives were analyzed. This paper investigated the web page service level of digital archives in various provinces and cities, summarized deficiencies of the current service mode of each archive, and improved the framework process of digital archives’ knowledge graph, finally the framework process was presented in the form of flow chart. [Result/conclusion] Knowledge graph can transform multiple types of digital archive text into data that computers can understand, and improve the level of computer intelligent identification. At the same time, dynamic timing and pertinence of the graph can improve the degree of integration archives according to the time node, user data can be updated based on users’ retrieval trace to improve the service quality, enhance the cooperation and communication between digital archives to achieve the result of resources integration, and provide a useful reference to better optimize the digital archiving service.