M-ISFCM:A Semisupervised Method for Anomaly Detection of MOOC Learning Behavior

Massive online courses(MOOCs)are becoming increasingly vital in the modern era,yet tools to track and detect MOOC learners’progress are inadequate.In reality,labeled MOOC data are difficult to acquire,whereas unlabeled data make up the majority of the data,and these massive unlabeled data are difficult to analyze,resulting in data waste.This paper tackles this issue by presenting a MOOC learning behavior anomaly detection model(M-ISFCM)for the supervision and inspection of MOOC learners’learning that combines semisupervised fuzzy C-mean clustering(SFCM)and an isolated forest algorithm.To optimize MOOC data usage,the model leverages unlabeled and labeledMOOCdata as prior assumptions.The MOOC detection runtimes are enhanced by integrating the outliers of the isolated forest approach in SFCM.The results show that the model has a higher precision rate,recall rate,andAUC than the traditional anomaly models in MOOC data.Therefore,the model is effective for recognizing anomalous MOOC learning behaviors.