Analysis of Adoption of a Mobile e-Payment System in State Universities and Colleges using Random Forest Classifier
Year: 2026
Type: article
Abstract: The rapid digital transformation in higher education institutions necessitates secure and efficient payment systems, a shift significantly accelerated by the COVID-19 pandemic; however, persistent security concerns continue to hinder widespread adoption. This study analyzes the adoption of a mobile e-payment system in State Universities and Colleges (SUCs) by integrating a Random Forest classifier for real-time fraud detection. Employing a multidisciplinary approach, the research utilized a convergent mixed-methods design for initial data gathering from a sample of 150 stakeholders, alongside a prototyping model for system development. Statistical analyses of the respondent data (n=150), including ANOVA, Cronbach’s Alpha, and Pearson correlation, revealed significant relationships between adoption readiness and three key determinants: usability/efficiency (r=0.684), insufficient security (r=−0.623), and user advantages (r=0.741). The developed system was evaluated using the ISO/IEC 25010 framework, achieving a "Very Good" aggregate score of 4.24/5.00. Furthermore, the Random Forest classifier demonstrated robust security capabilities, achieving 94.7% accuracy and an F1 score of 0.929. The findings indicate that embedding predictive machine learning significantly enhances fraud detection and transaction security, directly addressing stakeholder apprehensions and facilitating the adoption of digital payment systems in SUCs.
Authors: Peji, Joy, Castro, Flynn, Mission, Roger
Language: English
Subfield: Information Systems and Management
Field: Decision Sciences
Domain: Social Sciences
SDG: Quality education
Open Access status: diamond
OpenAlex ID: W7219980182