In response to the current trend of data archiving, this article studies the archiving of electronic documents from the perspective of knowledge engineering. Based on the four stages of document collection, document arrangement, document handover and file reception in the Electronic Document Archiving Specification for Party and Government Organs, it is transformed into three tasks of Information extraction, information classification and information matching. The key of Information extraction is to extract its metadata and content information; Information classification is based on the idea of ontology to build the association between the contents of the documents; Information matching is the process of determining the ownership relationship of completed electronic documents and categorizing them into their respective directories. The documents in the State Council Policy Document Library on the Central People's Government website were used as the data source for empirical evidence to implement the feasibility of the program proposed in this article.