A Neutrosophic Decision-Support Framework for Adaptive Learning
Pathways in Digital Education Platforms
Tanvir Mahmoud Hussein1, Priyanka Sharma2, Aastha Budhiraja2, Anshu Sharma2,
Tojiyev Rakhmatilla3, Sonia Setia4,∗
1College of Administrative & Financial Sciences, Gulf University, Bahrain
2Manav Rachna International Institute of Research and Studies, India
3Tashkent State University of Economics, Uzbekistan
4School of Computer Science Engineering, Galgotias University, Greater Noida, India
Emails: dr.tanvir@gulfuniversity.edu.bh, priyankasharmaiitd2@gmail.com; aasthakohli0410@gmail.com;
anshu.atri25@gmail.com; r.tojiyev@tsue.uz; setiasonia53@gmail.com
Abstract
Personalized learning pathways in digital education platforms have become essential for addressing the unique
needs and behaviors of individual learners. However, traditional adaptive systems often fail to account for the
uncertainty, ambiguity, and inconsistency inherent in educational data. This paper proposes a novel neutro-
sophic decision-support framework that models learner profiles using truth (T ), indeterminacy (I), and falsity
(F ) scores derived from student interaction and performance data. Utilizing the Open University Learning
Analytics Dataset (OULAD), we compute neutrosophic learner vectors based on assessment outcomes, en-
gagement patterns, and virtual learning environment (VLE) activity. A rule-based decision engine then rec-
ommends adaptive learning pathways—ranging from remedial to advanced—by interpreting the T/I/F distri-
butions through a neutrosophic logic framework. Experimental results demonstrate that the proposed model
enhances pathway assignment accuracy and provides better support for learners with incomplete or uncertain
data compared to traditional fuzzy and crisp models. The neutrosophic approach also ensures interpretability
and flexibility, making it well-suited for real-world educational platforms aiming to achieve adaptive learning
at scale.
Keywords: Neutrosophic logic; Adaptive learning; Decision support system; Educational data mining; Un-
certainty modeling; OULAD dataset