Semi-automatic labeling method and device for constructing process knowledge graph

The invention provides a semi-automatic labeling method and device for constructing a process knowledge graph, and aims to solve the problem that time and labor are wasted in current triple labeling. The method comprises the following steps: firstly, proposing a method for analyzing a semi-structured file knowledge information storage form, and defining a rule for automatically generating a text and a rule for subsequently labeling triples; 2, integrating the scattered knowledge information into a complete natural language description by using a preorder defined rule, and automatically labeling a relation triple stored in the structured information; thirdly, automatically generated texts and triples are manually audited in labeling software, and information with problems is modified and rewritten; and finally, other implicit relation triples are manually marked and exported and stored, and a corresponding file format is selected to construct a process knowledge graph or be used for training of other models. A