EUVIMEDEUVIMED entdecken
Crossref

A Context-Aware Hybrid Deep Learning and Federated Reinforcement Framework for Personalized E-Learning Course Recommendation

Aditya Garg, M.S. Satish Babu

International Journal of Drug Delivery Technology · 29.05.2026 · 16 (40s)

DOI: 10.25258/ijddt.16.40s.20 ↗

Kurz gefasst

Fachartikel in der Zeitschrift International Journal of Drug Delivery Technology, online erschienen am 29.05.2026. Verfasst von Aditya Garg und 1 weiteren Autor:innen.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Federated Hybrid and Contextual Reinforcement Learning-Based E-learning Recommendation System. The exponential growth of digital platforms for education has revolutionized knowledge sharing by making it ubiquitous, flexible, and selfpaced. Yet, the unprecedented amount of available courses, modules, and multimedia resources today is causing a severe cognitive burst, decision fatigue, and hence, lack of optimal learning paths. Existing E-learning Recommendation Systems (ERSs) face challenges like dynamic preference of learners, data sparsity, cold-start, semantic misalignment, and data privacy. In this paper, we propose FedHybrid-RLERS, a context-aware and privacy-preserving E-learning recommendation platform that harnesses the power of hybrid filtering (content-based + collaborative), deep representation learning, reinforcement-based sequential adaptation, and federated architecture for decentralized model training. Learner-course interactions are modeled as a dynamic hypergraph, where multi-modal features from content metadata, sequential behavioral activities, and contextual signals (time + device + learning pace + affective proxies) are extracted and aggregated. Recommendation policies are optimized using a deep Q-network with prioritized experience replay. To overcome the data silo problem while preserving the privacy of students, we train this model in a federated setting with differential privacy via a federated averaging (FedAvg) protocol. Extensive experiments on four real datasets, including Coursera, EdX, MOOC-Cube, and a synthetically augmented contextual dataset, reveal that our system (FedHybrid-RLERS) outperforms the state-of-the-art methods on accuracy, diversity, coverage, and fairness metrics. FedHybrid-RLERS outperforms strong baselines (LightGCN, AutoRec, BPR-MF) by improving NDCG@10 by 14.7%, Recall@20 by 12.3%, and reducing cold-start degradation by 31.5%. Ablation studies are provided to ascertain the contribution of individual components of the system. The framework is implemented using PyTorch, FastAPI, Neo4j, and Docker, ensuring scalability and real-time inference (

Zitieren

APA

Garg, A., & Babu, M. S. (2026). A Context-Aware Hybrid Deep Learning and Federated Reinforcement Framework for Personalized E-Learning Course Recommendation. International Journal of Drug Delivery Technology, 16(40s). https://doi.org/10.25258/ijddt.16.40s.20

Vancouver

Garg A, Babu MS. A Context-Aware Hybrid Deep Learning and Federated Reinforcement Framework for Personalized E-Learning Course Recommendation. International Journal of Drug Delivery Technology. 2026;16(40s). doi:10.25258/ijddt.16.40s.20

BibTeX
@article{Garg2026,
  author = {Garg, Aditya and Babu, M.S. Satish},
  title = {A Context-Aware Hybrid Deep Learning and Federated Reinforcement Framework for Personalized E-Learning Course Recommendation},
  journal = {International Journal of Drug Delivery Technology},
  year = {2026},
  volume = {16},
  number = {40s},
  doi = {10.25258/ijddt.16.40s.20}
}
Kostenlos starten →