Online Learning and Analytics

Fine-tuned language model provides automated question answering in large calculus courses, achieving 75% accuracy. Hybrid human-AI workflows maintain instructor oversight while scaling support.

Study of 96 undergraduates examining how relevance, clarity, detail, examples, and preferences influence perceived usefulness of supplementary materials in software engineering education.

Explore AI's transformative impact on higher education, examining opportunities in personalized learning alongside challenges in academic integrity, bias, and institutional governance.

Design-based research evaluates GenAI literacy training for teacher education students, showing improvements in self-efficacy and critical engagement with AI tools.

Mixed-methods evaluation of an AI learning assistant in engineering courses reveals high usability appeal but identifies ethical policy uncertainty as a key barrier to broader student engagement.

Survey of 583 Chinese undergraduates validates the UNESCO AI Competency Framework, revealing stronger performance in ethics and mindset than technical skills, contradicting national priorities.

Study applies Bayesian network modelling to predict employment outcomes and match vocational students with suitable majors, achieving 75% predictive accuracy and reducing skill-job mismatches.

Survey of 71 students reveals perceptions of generative AI as academically beneficial while raising concerns about privacy, reliability, and plagiarism, with strong commitment to integrity principles.

Deep learning framework classifies undergraduate GPA by integrating family background and psychological factors, achieving 0.798 accuracy in identifying academic risk among Chinese college students.

Multinomial logistic regression identified distinct predictors of perceived learning and grade achievement in online college courses, enabling differentiated student support strategies.

AlphaLearn proposes a multi-objective evolutionary framework for adaptive e-learning pathways that integrates fairness as a core optimization criterion alongside learning effectiveness and engagement.

Hybrid framework combining graph networks and transformers with synthetic data augmentation for automatic student engagement recognition from facial video in e-learning systems.

How educators reimagine course objectives to emphasize human cognition, critical judgment, and creativity in response to generative AI capabilities.

De-identified clickstream dataset from KU Leuven supporting Learning Analytics research while maintaining learner privacy through transparent anonymization.

Contextual Thompson sampling approach for personalizing exercise sequences in digital learning environments, optimizing skill advancement at scale using bandit-based algorithms.

Case study of data science education examining effects of interest-targeted instructional videos and self-directed data analysis exercises on student learning outcomes across 8,509 participants.

Gurukulya is an AI-powered smart campus platform combining LLMs, sentiment analysis, and automated assessment to boost student engagement, reduce evaluation time by 85%, and improve learning.

TC-MAG framework uses multi-agent AI to improve automated grading of open-ended student responses with explainable partial credit assessment and teacher oversight capabilities.

Linguistic analysis of generative AI discourse in higher education reveals how agency, evaluation, stance, and coherence constructions shape learner outcomes and disciplinary representation.

Systematic survey of 1,125 papers from flagship conferences identifying 172 open datasets in learning analytics, with PRACTICE guidelines for dataset publication and reproducibility practices.