The Application of Machine Learning to Educational Process Data Analysis

A Systematic Review

Bibliographic Data

ID22045154
AuthorsJing Huang (0000-0002-6205-0025, Department of Educational Studies, College of Education, Purdue University, 100 North University Street, West Lafayette, IN 47906, USA), Yan Ping Xin (0000-0002-0035-9203, Department of Educational Studies, College of Education, Purdue University, 100 North University Street, West Lafayette, IN 47906, USA), Hua-Hua Chang (0000-0002-2949-1902, Department of Educational Studies, College of Education, Purdue University, 100 North University Street, West Lafayette, IN 47906, USA)
Year2025
Volume15
Issue7
Pages888
Publication date2025-07-11
Open AccessYes
TypeARTICLE
VenueEducation Sciences (JOURNAL)
Journal identifiersISSN: 2227-7102 • E-ISSN: 2227-7102
PublisherMDPI AG (PUBLISHER • IT)
DOI10.3390/educsci15070888
LanguageEN
Citations received5
References cited35

Educational process data offers valuable opportunities to enhance teaching and learning by providing more detailed insights into students’ learning and problem-solving processes. However, its large size, unstructured format, and inherent noise pose significant challenges for effective analysis. Machine learning (ML) has emerged as a powerful tool for tackling such complexities. Despite growing interest, a comprehensive review of ML applications in process data analysis remains lacking. This study contributes to the literature by systematically reviewing 38 peer-reviewed publications, dated from 2013 to 2024, following PRISMA 2020 guidelines. The findings of this review indicate that (1) clickstream data is the most widely used processing data type, (2) process data analysis offers actionable insights to support differentiated instruction and address diverse student needs, and (3) ML typically serves as a tool for coding process data or estimating student ability. Persistent challenges, including feature extraction and interpreting results for practical applications, are also discussed. Finally, implications for future research and practice are discussed with a focus on enhancing personalized learning, improving assessment accuracy, and promoting test fairness

Tools

Open DOI