Performance Comparison of CEPL and CEPL-BERT for Real-Time Personaized Content Recommendation

DOI

10.2991/978-94-6239-616-6_3How to use a DOI?

Keywords

Personalized Learning; CEPL; CEPL-BERT; Deep Learning; and Transformer Embedding; Emotion Detection

Abstract

Personalized Learning (PL) has evolved to address the unique needs of individual learners by adapting content delivery based on preferences, prior knowledge and contextual factors. This paper presents a comparative study between two personalized learning models-CEPL (Context Emotion Personalized Learner) and enhances CEPL-BERT system. While CEPL utilizes emotion detection (via VADER), contextual inputs (like time and weather), and traditional content features, CEPL-BERT incorporates advanced semantic understanding using BERT based embedding for deeper personalization. The study evaluates these models across multiple models across multiple performance metrics including accuracy, F1-Score and AUC. Experimental results reveal that CEPL-BERT consistently outperforms CEPL in recommendation relevance, demonstrating the advantage of transformer-based insights into the role of emotion, context and semantic depth in enhancing learner engagement and system adaptability.

Unlike prior PL systems that integrate emotion, context, and content at a shallow level, CEPL-BERT introduces a transformer-based semantic layer and a weighted fusion attention mechanism, which dynamically balances emotion–context influence for each user in real time. This approach allows the system to provide more precise and adaptive content recommendations based on both user emotions and contextual factors Both models employ feature fusion strategies and deep learning classifiers for adaptive content recommendations. The novelty of our work lies not merely in replacing embedding’s, but in integrating semantic-aware content understanding with contextual gating and dynamic fusion of emotion and context, enabling real-time, user-adaptive personalized learning.

Copyright

© 2026 The Author(s)

Open Access

Open Access This chapter is licensed under the terms of the Creative Commons Attribution-NonCommercial 4.0 International License (http://creativecommons.org/licenses/by-nc/4.0/), which permits any noncommercial use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license and indicate if changes were made.

Download article (PDF)

TY  - CONF
AU  - Priyadharshini Gunasekaran
AU  - Bhuvaneswari Subbaraman
PY  - 2026
DA  - 2026/03/31
TI  - Performance Comparison of CEPL and CEPL-BERT for Real-Time Personaized Content Recommendation
BT  - Proceedings of the International Conference on Artificial Intelligence and Secure Data Analytics (ICAISDA 2025)
PB  - Atlantis Press
SP  - 16
EP  - 39
SN  - 1951-6851
UR  - https://doi.org/10.2991/978-94-6239-616-6_3
DO  - 10.2991/978-94-6239-616-6_3
ID  - Gunasekaran2026
ER  -