Abstract: A method and system for detecting negative learning behaviors associated with a student system in an education program environment. Activity information is collected that corresponds to the performance of learning events of the education program. Based on the activity information, a student profile corresponding to the student is generated. One or more features are extracted from the student profile and used by a machine learning model to classify the student profile as a negative learning indicator associated with the student. An alarm notification is generated that correspond to the negative learning behavior indicator.