This study utilizes clustering algorithms, based on subject knowledge mapping, to construct a learning community that can support the comprehensive development of group wisdom, taking into account the learner portrait, initial learning community, and the learning path. In order to verify the validity of the E-GPPE-C model of wisdom learning, University X was selected to conduct a one-semester quasi-experimental study. Attitude data and outcome data from the pre- and post-teaching portraits were collected to analyze the effectiveness of teaching based on the amount of change and significance test. Using the E-GPPE-C model of smart learning proposed in this study, the number of learning achievers before and after teaching increased to 8, which can improve the learners’ academic performance in civics and politics (P<0.05) and their’ satisfaction with the learning environment. Many students believed that smart teaching had changed their ideal beliefs and knowledge base in civics and politics.