Educational Data Mining: A Systematic Review on the Applications of Classical Methods and Deep Learning Until 2022

• Published in: IEEE Symposium on Industrial Electronics and Applications pp. 1 - 15
• Main Authors: Trung, Bui Duc, Son, Ngo Tung, Tung, Nguyen Duy, Son, Kieu Anh, Anh, Bui Ngoc, Lam, Phan Truong
• Format: Conference Proceeding
• Language: English
• Published: IEEE 15.07.2023
• Subjects: Data Mining; Deep learning; Education; Educational Data Mining; Ethics; Learning systems; Machine Learning; Neural networks; Policies; Regulation; Stakeholder; Surveys; Systematics
• ISSN: 2472-7660
• DOI: 10.1109/ISIEA58478.2023.10212273

Educational Data Mining (EDM) is a research field that focuses on extracting valuable insights and knowledge from data in the education sector. EDM exploits Data Mining (DM) techniques such as Clustering, Regression to analyze data and make predictions that help answer questions in education. Meanwhile, Deep Learning (DL) is a subfield of machine learning that uses neural networks to solve complex problems related to natural language processing and computer vision. DL is highly effective at processing textual and digital data such as images, audio, and video. DL can be applied to build automatic grading, attendance, monitoring systems, or intelligent learning systems in education. However, the applicability of DL in education still needs to be solved due to strict education regulations and policies with the many challenges involved in collecting and utilizing data processes. Additionally, education is subject to many ethical and legal controversies. This survey aims to clarify the concepts and issues related to EDM, such as technology, users, and data. It also explores the technologies and applications of DL in the educational environment, including the barriers and difficulties encountered when applying these technologies. Finally, the survey offers some comments on the future development direction of this field.