Abstract: In order to find the learning rules and trends of the MOOC course learners,this paper uses the background data of the MOOC course of the circuit analysis foundation of the University of Electronic Science and Technology of China,and makes detailed data mining and analysis from the types of participants,the correlation analysis of each knowledge point in the course,the prediction of the learning effect and so on.The types of participants,the association analysis of different topics in the course,as well as the anticipation of the final testing grades are addressed in the paper.Through the data mining work,the following interesting conclusions can be reached.First,the more the learners participate in the test and discussion,the higher the possibility of obtaining the MOOC certificate.Second,there is strong association between the mastery of the “Kirchhoff theorem test and the reference direction” knowledge point and the unit test results.Third,it can predict the learning effect very well like other MOOC courses,which also lack the enthusiasm to discuss.The data mining in this paper provides an important basis for further enhancing the learning effect and optimizing the structure of MOOC course.