This paper introduces the cognitive computing model into Civic Education in colleges and universities and collects the multi-dimensional cognitive data of the learners through methods such as eye movement data measurement sequence data mining, which accurately reflects the individual differences in the interaction in the educational environment. Using algorithms such as personalized learning to achieve personalized learning path design based on learners. Using 42 students from a university as the research subjects, it was found that students prefer video-type learning resources with approximately 62% of them choosing them. Cognitive characteristics are portrayed through students’ behavioral data: some students are very active in exchanging learning activities, while others are relatively independent. Students’ behavior showed a multi-linear pattern. There was a significant difference between the high and low wind groups in terms of assigning learning activities with a p-value of less than 0.05. There were outliers and inattentiveness in the eye movement trajectories of students 3, 12 and 5. Personalized teaching based on the cognitive computing model is significantly better than the traditional teaching model’s teaching effect.