Abstract In today’s rapidly changing society, exploring the historical evolution of Marxist ideology in Civic Education can provide a more comprehensive understanding of the subject. In this paper, oriented to Civic Education, we use crawlers and other related technologies to obtain a Civic Text Corpus and complete the acquisition of Civic entities based on the BiLSTMCRF model for its annotation and collection. The acquired entities are divided into unstructured and semi-structured data, and the former is used to extract the relational triad by crawling technology, while the latter is obtained with the help of dependency syntactic analysis. Based on the analysis of the constructed cognitive map of Civic Education, the word “construction” in Marxist Civic Education in the 21st century has the highest degree of enthusiasm, reaching 0.191, followed by “student”, with a degree of enthusiasm of 0.150.