International Journal of Computer Applications
Foundation of Computer Science (FCS), NY, USA
|
Volume 186 - Issue 72 |
Published: March 2025 |
Authors: Delali Kwasi Dake |
![]() |
Delali Kwasi Dake . Post-pandemic recovery: investigating factors that affected students’ online engagement during the pandemic in Ghana – A machine learning approach. International Journal of Computer Applications. 186, 72 (March 2025), 19-33. DOI=10.5120/ijca2025924518
@article{ 10.5120/ijca2025924518, author = { Delali Kwasi Dake }, title = { Post-pandemic recovery: investigating factors that affected students’ online engagement during the pandemic in Ghana – A machine learning approach }, journal = { International Journal of Computer Applications }, year = { 2025 }, volume = { 186 }, number = { 72 }, pages = { 19-33 }, doi = { 10.5120/ijca2025924518 }, publisher = { Foundation of Computer Science (FCS), NY, USA } }
%0 Journal Article %D 2025 %A Delali Kwasi Dake %T Post-pandemic recovery: investigating factors that affected students’ online engagement during the pandemic in Ghana – A machine learning approach%T %J International Journal of Computer Applications %V 186 %N 72 %P 19-33 %R 10.5120/ijca2025924518 %I Foundation of Computer Science (FCS), NY, USA
The COVID-19 pandemic remains a scar on humans, with profound lessons for the future. The post-pandemic recovery in education is crucial for guiding proactive measures for educational stakeholders and governmental entities. During the pandemic, in-person teaching and learning was halted with hasty migration to online learning, especially in developing countries. Students in developing and underdeveloped nations riskily adapted to complete the academic semester online amid concerns of getting the infectious virus. Instructors' pedagogy abruptly shifted, with little or no training on professional approaches to support students during online learning in the face of a virus. This research employed a machine learning approach to examine online education during the pandemic in anticipation of unforeseen circumstances. K-modes clustering was applied to a categorical dataset to reveal latent cluster formations and variable correlations. The naïve bayes classifier was then used to build a predictive model for future cluster members. Four clusters were established, encompassing patterns of unsupportive parents, inadequate internet connectivity, supportive instructors, and average academic success. The naïve bayes classifier also outperformed the random forest algorithm with an accuracy of 84.47%.