TY - JOUR
T1 - Clustering-based knowledge graphs and entity-relation representation improves the detection of at risk students
AU - Albreiki, Balqis
AU - Habuza, Tetiana
AU - Palakkal, Nishi
AU - Zaki, Nazar
N1 - Publisher Copyright:
© 2023, The Author(s), under exclusive licence to Springer Science+Business Media, LLC, part of Springer Nature.
PY - 2023
Y1 - 2023
N2 - The nature of education has been transformed by technological advances and online learning platforms, providing educational institutions with more options than ever to thrive in a complex and competitive environment. However, they still face challenges such as academic underachievement, graduation delays, and student dropouts. Fortunately, by harnessing student data from institution databases and online platforms, it becomes possible to predict the academic performance of individual students at an early stage. In this study, we utilized knowledge graphs (KG), clustering, and machine learning (ML) techniques on data related to students in the College of Information Technology at UAEU. To construct knowledge graphs and visualize students’ performance at various checkpoints, we employed Neo4j-a high-performance NoSQL graph database. The findings demonstrate that incorporating clustered knowledge graphs with machine learning reduces predictive errors, enhances classification accuracy, and effectively identifies students at risk of course failure. Additionally, the utilization of visualization methods facilitates communication and decision-making within educational institutions. The combination of KGs and ML empowers course instructors to rank students and provide personalized learning interventions based on individual performance and capabilities, allowing them to develop tailored remedial actions for at-risk students according to their unique profiles.
AB - The nature of education has been transformed by technological advances and online learning platforms, providing educational institutions with more options than ever to thrive in a complex and competitive environment. However, they still face challenges such as academic underachievement, graduation delays, and student dropouts. Fortunately, by harnessing student data from institution databases and online platforms, it becomes possible to predict the academic performance of individual students at an early stage. In this study, we utilized knowledge graphs (KG), clustering, and machine learning (ML) techniques on data related to students in the College of Information Technology at UAEU. To construct knowledge graphs and visualize students’ performance at various checkpoints, we employed Neo4j-a high-performance NoSQL graph database. The findings demonstrate that incorporating clustered knowledge graphs with machine learning reduces predictive errors, enhances classification accuracy, and effectively identifies students at risk of course failure. Additionally, the utilization of visualization methods facilitates communication and decision-making within educational institutions. The combination of KGs and ML empowers course instructors to rank students and provide personalized learning interventions based on individual performance and capabilities, allowing them to develop tailored remedial actions for at-risk students according to their unique profiles.
KW - Clustering
KW - Educational ranking
KW - Knowledge graphs
KW - Machine learning
KW - Personalized learning
KW - Students’ performance
UR - http://www.scopus.com/inward/record.url?scp=85167346506&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=85167346506&partnerID=8YFLogxK
U2 - 10.1007/s10639-023-11938-8
DO - 10.1007/s10639-023-11938-8
M3 - Article
AN - SCOPUS:85167346506
SN - 1360-2357
JO - Education and Information Technologies
JF - Education and Information Technologies
ER -