article · Computers and Education Artificial Intelligence
Data generated by intelligent tutoring systems provides valuable insights when analysed effectively, particularly regarding student performance. Predicting whether a learner will answer future questions correctly relies heavily on historical interaction logs, helping to support individualised learning and improve educational outcomes. Incorporating the XGBoost machine learning algorithm into existing logistic regression-based predictive frameworks, specifically Item Response Theory, Performance Factor Analysis, and DAS3H, boosts predictive accuracy across diverse settings. An evaluation using eight real-world datasets from various online intelligent tutoring systems, including a newly introduced log dataset from Moodle Morocco, showed clear improvements. The integration of XGBoost raised the predictive performance of Performance Factor Analysis across seven datasets, achieving an area under the curve of up to 0.88. Additionally, the approach improved the performance of the DAS3H model on the ASSISTment17 dataset while maintaining comparable predictive results for Item Response Theory across several benchmarks.
Accurate forecasting of student performance allows intelligent educational platforms to identify when learners struggle and tailor support accordingly. Demonstrating that machine learning methods like XGBoost can upgrade established assessment models across diverse datasets, including data from North Africa, helps developers build more reliable systems to maximise student learning and achievement.
This applied research shows potential for developers and operators of intelligent tutoring systems and online learning management platforms seeking more accurate student tracking engines. Because the algorithmic enhancements have been evaluated against eight real-world educational datasets, the underlying models are in an applied and tested state, ready for integration into adaptive learning software to anticipate learner outcomes and adjust instruction.
AI-generated from the published abstract. Always read the original work before citing.
The huge amount of data generated by an Intelligent Tutoring System becomes useful when analyzed in an appropriate way to provide significant insights about learners, especially his or her performance. Performance data retrieved from historical interactions is the main engine for learner performance prediction, where the likelihood of the learner answering correctly future questions is calculated. Modeling learner performance can provide significant insights into individual students to promote successful learning and maximize educational achievement. This study aims to enhance the learner performance prediction of some logistic regression-based models, namely Item Response Theory, Performance Factor Analysis, and DAS3H using XGBoost, including an empirical comparison of eight real-world datasets, containing performance log data collected from different online intelligent tutoring systems, involving the first time a new dataset from Moodle Morocco. The results have demonstrated that the XGBoost has enhanced PFA predictive performance on seven datasets with an AUC of up 0.88 and improved the DAS3H AUC on the ASSISTment17 dataset while conserving almost the same predictive results for Item Response Theory on some datasets.
This page summarises published work. The authoritative version sits with the publisher.
DOI: 10.1016/j.caeai.2024.100254
Is something wrong with this record? Report it or request removal.
Discussion
Have you built on this work, tried to replicate it, or seen it applied in practice? Share what you know. Verified researchers and MARATTO™ domain experts can open a discussion, and any member can reply. Contributions are reviewed before they appear.
No discussion yet. Open the first thread.
New to MARATTO™? Create a free account.
