Currently, the digitalization of ancient Chinese texts primarily involves basic document scanning and coarse-grained file management. However, the development of generative artificial intelligence technologies offers new opportunities for advancing the depth of digitalization of ancient texts. Based on the theory of frame semantics, this article constructs a semantic logical structure framework for semantic tagging. It poses questions to large language models using generative logic, recursively extracts deep semantic content from ancient text corpora, and outputs it as structured data conforming to the semantic tagging framework. This enables ancient texts to be processed with a unified semantic logic at the basic semantic level. The automatic semantic tagging framework for ancient texts can achieve large-scale, automated generative representation of the content structure of ancient documents, providing an automatically viable technical solution for the intelligent transformation of ancient text organization.