Deep learning driven personalized learning: CNN, DNN, and Keras Tuner for activity recommendation

Bensalem Assri, Amine Mrhari, Moulay Youssef Hadi, Mohamed Boutaib

Abstract


Personalized learning for individual needs is one of the most important research areas in e-learning, aiming primarily to improve students’ outcomes. This paper proposes a novel framework for personalized learning based on students’ learning styles (LS) and behavior data. The prediction of student performance is crucial to identify students who need personalization interventions. Identifying LS is a major step to provide each student with a pedagogical experience tailored to their needs. The LS is identified based on the Felder-Silverman learning style model (FSLSM) and student interaction with virtual learning environments (VLE). A learning activity recommendation is used to recommend the most suitable activities for each student. Open university learning analytics dataset (OULAD) is used, which contains various data about students’ interaction with VLE. We employed a methodological core combining convolutional neural network (CNN) and deep neural network (DNN) models using Keras Tuner for hyperparameter tuning to predict student performance and identify learners’ LS. The results indicate that DNN and CNN achieved improved outcomes for 51.37% and 40.82% of students, respectively, after the recommendations were made. This study demonstrates the effectiveness of deep learning (DL) with hyperparameter tuning for personalized learning in distance learning.

Keywords


Convolutional neural network; Deep neural network; Felder-Silverman learning style model; Open university learning analytics dataset; Recommendation; Student outcome; Virtual learning environment

Full Text:

PDF


DOI: https://doi.org/10.11591/eei.v15i6.11936

Refbacks

  • There are currently no refbacks.


Creative Commons License
This work is licensed under a Creative Commons Attribution-ShareAlike 4.0 International License.

Bulletin of EEI Statistics

Bulletin of Electrical Engineering and Informatics (BEEI)
ISSN: 2089-3191 , e-ISSN: 2302-9285
This journal is published by the Institute of Advanced Engineering and Science (IAES) .