Diversified integration, rationalized allocation and high-efficiency utilization of various practical teaching resources are the prerequisite and foundation for realizing the goal of monolithic education of Civic and Political Science teaching. This paper proposes a local two-way adaptive degree function of a complex network of Civic and Political Science teaching resources cluster and constructs a discrete regression-based resource diversification integration model based on the adaptive degree function by considering the characteristics of a complex network of Civic and Political Science teaching resources cluster with local worldliness (cluster structure), complexity, incomplete and asymmetric information obtained. The research results show that the discrete regression model Cronbach’s α coefficient value is 0.903, and the cumulative explained variance reaches 81.69%, which can effectively and highly fit the integration of diverse resources of Civics teaching. And compared with the traditional BA integration model, the discrete regression model increases the integration speed by 36.21% and the breadth by 41.39%. The discrete regression-based resource diversification integration model proposed in this paper effectively improves the level of practical teaching of Civics in colleges and universities and provides some reference and help for related researchers.