如何应对学生的个体差异，真正做到因材施教的个性化教学（学习），一直是摆在每个教育工作者面前的课题。本项目拟从课程、学生、教师三个视角出发，在项目组已有教育大数据分析与挖掘研究成果和教学经验的基础上展开研究，具体包括：①基于维基数据、ACM CCS分类树和微软学术图上下位关系知识库，利用深度学习方法引入用户反馈，实现结合学术与产业热点的代表性话题抽取，为课程的内容组织提供依据。同时将社会分析领域被广泛采用的因子与TAM基础模型结合，完成学生课程学习行为的定性分析。②通过结合静态和动态建模方法的学生个性化模型构建，实现学生学习内容的个性化推荐和学习效果的分类预测。③基于注意力模型的深度学习方法，通过对学生行为（反馈）文本数据的挖掘分析，实现对课堂教学效果的及时检测。本项目的研究工作并不局限于某特定学科的教育问题，拟取得的研究成果具有普适性，并将通过实际教学环节对成果的有效性和实用性进行检验。

Every educator has always been thinking about how to teach students according to their individual differences. Based on the research results and teaching experience of the research group, this research is planned to study from three perspectives: curriculum, students, and teachers. The specific contents are as follows. First, based on Wiki Data, ACM CCS Taxonomy Tree and Microsoft Academic Map (MAG) of hyponymy knowledge base, we intend to use deep learning methods to introduce user feedback to achieve the representative topic extraction which combines academic and industrial hotspots, and ultimately provide the basis for organizing course content. At the same time, we combine the factors that are widely used in the social analysis field with the TAM model to complete the qualitative analysis of the students' learning behavior. Second, we intend to build a student personalization model in combination with static and dynamic modeling methods, to realize the personalized recommendation of students' learning content and the prediction of students' performance. Third, the deep learning method based on attention model is used to mine and analyze the text data generated by student feedback, so as to achieve timely examination of classroom teaching effect. The work of this research is not limited to the education problem in a particular subject, and the research results to be obtained are universal. Finally, we will examine the effectiveness and practicality of the results by practical course.