Research on Recommendation Methods Based on Data Augmentation in a Heterogeneous View

In recent years, many domains contain different object and relationship types, forming Heterogeneous Information Networks (HINs). These HINs include richer semantic and structural information than homogeneous information networks. In this paper, a comprehensive survey of data mining and applications using HINs is presented. Specifically, we first review the current achievements in HIN research and outline the research trends of HINs in data mining. Next, we explore the applications of HINs in various fields, such as finance and education. Finally, through a series of experiments, we explore research hotspots and the development of HINs in educational applications and make recommendations for future areas of study in HIN applications. The experimental results demonstrate the effectiveness of HIN based techniques in real-world data mining tasks, particularly in educational applications such as course recommendation.