Traditional culture and civic education are two important parts of the higher education system. Based on knowledge graph technology, this paper explores a strategy to promote the intelligent integration of the two. Firstly, we construct a knowledge map for the field of civics and traditional culture and optimize the traditional named entity recognition algorithm by using the BERT module for word feature extraction and pre-training. Through the search and recognition of stored domain knowledge, a smart Q&A system that integrates ideology, politics, and traditional culture can be constructed and used for teaching. The F1 value of the constructed BERT-BiLSTM-CRF algorithm is 76.7, which outperforms the comparison algorithm, and the ablation experiments show that the F1 value of the algorithm decreases by 7.9% when the BERT module is missing, which verifies the reasonableness of the design of this paper. The results of the teaching experiments on the performance of Civics and the interest in learning traditional culture show that the application of an intelligent Q&A system for integrated teaching effectively improves the students’ performance in Civics and the interest in learning traditional culture.