【关 键 词 】：
学习习惯； 作业习惯； 学业拖延； 学业勤奋； 智能学习系统； 建模
Homework habits are closely related to learners' academic achievement and the modeling of homework habits is an urgent problem in the design of intelligent learning system. Based on a multi-level homework model, this study adopts a hybrid research approach to identify "academic procrastination" and "academic diligence" as two important and measurable dimensions of homework habits, which represent the two core types of homework behaviors: homework time and homework effort. A sub-model of academic procrastination habits is constructed, including procrastination in doing and submitting. Clustering is applied to diagnose the habit of academic procrastination, and four types of learners are defined as "non-procrastinators", "severe procrastinators", "procrastinators in submitting" and " procrastinators in doing". Based on the theoretical framework of time investment and focus model, a sub-model of academic diligence habits including time investment and focus is constructed. The artificial neural network is used to diagnose the academic diligence habits, as a result, five levels of diligence are diagnosed, and the validity of the model is verified.