MOOC comprehensive feature fusion-based learning prediction method and device

The invention discloses a learning prediction method and device based on MOOC comprehensive feature fusion, and relates to the technical field of data prediction identification, and the method comprises the steps: obtaining a learning behavior log, basic attribute information and course structure data; respectively extracting sequence features from the learning behavior information sequence, the learner attribute information sequence and the course information sequence, and converting each sequence feature into a feature vector with a unified dimension; mutual information among different feature vectors is calculated, and the mutual information and original features are fused into a comprehensive feature vector; internal relation weights among the comprehensive feature vectors are analyzed, weighted summation is carried out on each comprehensive feature vector according to the weights, and a fusion feature vector is obtained; and carrying out learning risk dichotomy prediction on the fusion feature vector, an