Scholarlink

<p>This study investigates the key factors shaping the effectiveness of virtual education at PayameNoor University, Isfahan Province, Iran, with a particular focus on the potential of AI-driven personalization. A mixed-method quantitative approach was applied, combining Partial Least Squares Structural Equation Modeling (PLS-SEM) with advanced machine learning models, including Random Forest (RF) and Artificial Neural Networks (ANN). Data were collected from a stratified random sample of 190 faculty members and 8,680 students using a validated questionnaire encompassing 9 core dimensions: mission and objectives, organizational structure, student satisfaction, faculty engagement, research, evaluation, teaching–learning, e-content, and technical competence. The instrument demonstrated strong reliability and validity (Cronbach’s <i>α</i> = 0.89; CR &gt; 0.7; AVE &gt; 0.5). The PLS-SEM model explained 60% of the variance in e-learning effectiveness. Mission and objectives (β = 0.264), organizational structure (β = 0.231), student satisfaction (<i>β</i> = 0.218), and faculty engagement (β = 0.205) emerged as significant positive predictors. In contrast, research activities (<i>β</i> = –0.183), evaluation systems (<i>β</i> = –&#xa0;0.152), and teaching–learning methods (β = –0.135) showed adverse effects, reflecting misalignments with digital learning requirements. Complementary ML analyses highlighted curriculum design, clarity of course objectives, and faculty mastery of virtual tools as the most decisive indicators. Overall, the findings emphasize that human and organizational factors outweigh purely technical ones in determining virtual learning success. Strategic investment in curriculum design, faculty upskilling, and AI-driven personalization is essential for developing adaptive and effective virtual learning environments. The study contributes to both theory and practice by offering a data-driven framework that can guide policy and implementation in higher education institutions, particularly in resource-constrained contexts.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Artificial intelligence and machine learning for personalized virtual learning: a case-based analysis in higher education

  • 期刊论文

  • 发表日期：

    2026年2月14日

查看全文
  • Majid Javari

摘要

This study investigates the key factors shaping the effectiveness of virtual education at PayameNoor University, Isfahan Province, Iran, with a particular focus on the potential of AI-driven personalization. A mixed-method quantitative approach was applied, combining Partial Least Squares Structural Equation Modeling (PLS-SEM) with advanced machine learning models, including Random Forest (RF) and Artificial Neural Networks (ANN). Data were collected from a stratified random sample of 190 faculty members and 8,680 students using a validated questionnaire encompassing 9 core dimensions: mission and objectives, organizational structure, student satisfaction, faculty engagement, research, evaluation, teaching–learning, e-content, and technical competence. The instrument demonstrated strong reliability and validity (Cronbach’s α = 0.89; CR > 0.7; AVE > 0.5). The PLS-SEM model explained 60% of the variance in e-learning effectiveness. Mission and objectives (β = 0.264), organizational structure (β = 0.231), student satisfaction (β = 0.218), and faculty engagement (β = 0.205) emerged as significant positive predictors. In contrast, research activities (β = –0.183), evaluation systems (β = – 0.152), and teaching–learning methods (β = –0.135) showed adverse effects, reflecting misalignments with digital learning requirements. Complementary ML analyses highlighted curriculum design, clarity of course objectives, and faculty mastery of virtual tools as the most decisive indicators. Overall, the findings emphasize that human and organizational factors outweigh purely technical ones in determining virtual learning success. Strategic investment in curriculum design, faculty upskilling, and AI-driven personalization is essential for developing adaptive and effective virtual learning environments. The study contributes to both theory and practice by offering a data-driven framework that can guide policy and implementation in higher education institutions, particularly in resource-constrained contexts.