Learning Performance Prediction and Alert Method in Hybrid Learning

In online learning, students’ learning data such as time and logs are commonly used to predict the student’s performance. a hybrid context, activities occur both offline. Thus, how integrate offline effectively for an accurate performance prediction becomes very challenging. This paper proposes “prediction alert” model in context. The is developed evaluated through analyzing 16-week (one semester) attributes of English 50 students eighth grade. Six significant variables were determined attributes, namely, qualified rate, excellent scores, number practice sessions, time, completion. proposed was put into actual four months application modification, which sample middle school participated. shows feasibility effectiveness analysis learning. It can support continuous more effectively.