A Review of Algorithms and Challenges for Personalized Learning Path Recommendation in E-learning

Authors:
  • GJGapar Md Johar
    Gapar Md Johar

Abstract
A Review of Algorithms and Challenges for Personalized Learning Path Recommendation in E-learning-Advances in Computer and Communication-Hill Publishing Group The rapid expansion of e-learning has exposed a fundamental conflict: vast resource accessibility now clashes with rigid, one-size-fits-all instruction, failing to meet learners' personalization needs. Personalized Learning Path Recommendation (PLPR) emerges as the critical technology to resolve this conflict, a direction strongly endorsed by global education policies. This paper offers a structured review of the PLPR field. It first forges a clear classification of core algorithms into five major lineages: non-machine learning, traditional machine learning, graph-based, deep learning, and reinforcement learning. For each lineage, the analysis critically weighs documented advantages against inherent limitations, anchoring the comparison in empirical findings from representative literature to reveal their practical trade-offs. Building on this analysis, the paper dissects the field's persistent challenges—notably data sparsity, the cold-start problem, and the demands of maintaining accuracy at scale. Ultimately, this work charts a strategic path for both researchers and practitioners, aiming to connect PLPR's technical potential with the urgent, practical needs of today's digital learning environments.

This publication has 23 references indexed in Scilit: