AFRIDATACYBERJOHN: An Africa-Centric Mobile Money and USSD Cybersecurity Dataset for Behavioural Intrusion Detection and Transaction Security
Year: 2026
Type: dataset
Abstract: AFRIDATACYBERJOHN is an Africa-centric synthetic cybersecurity dataset designed for research into behavioural intrusion detection, mobile-money security, USSD transaction security, session analysis, and cyberattack detection in African financial-service contexts. The dataset contains session-oriented mobile-money and USSD event records together with prediction-window data designed to support behavioural cybersecurity research. It provides both an authoritative hard-negative event layer and a corresponding current event layer, enabling forensic comparison, reproducibility analysis, and evaluation of security analytics workflows. The release contains 1,029 authoritative event records, 1,060 current event records, 100 authoritative prediction windows, and 100 current prediction windows. The prediction windows cover horizons of 5, 10, 15, 30, and 60 minutes. The dataset includes information describing users, sessions, USSD interactions, authentication activity, transaction events, mobile-money platforms, session characteristics, device observations, network characteristics, response behaviour, and prediction-window matching information. AFRIDATACYBERJOHN is synthetic and programmatically generated using reproducible Python-based procedures. It is not presented as a direct extract of production banking, telecommunications, or customer records. The release includes comprehensive documentation covering the dataset description, methodology, data dictionary, forensic validation, provenance, and semantic recovery of prediction-window fields. The frozen CSV files have been independently integrity-verified using SHA-256 checksums.
Author: AJAYI, JOHN OJO
Institution: National Open University of Nigeria
Language: English
Subfield: Information Systems
Field: Computer Science
Domain: Physical Sciences
Open Access status: green
OpenAlex ID: W7214162312