MGBEMENA, CHIEMELIE STANLY (2026) DESIGN AND DEVELOPMENT OF A SECURE WEB-BASED DIASPORA VOTING SYSTEM (VOTE.NG). Other thesis, Godfrey Okoye University, Enugu.
|
Text
Mgbemena Chiemelie.pdf Download (2MB) | Preview |
Abstract
Nigeria's diaspora population, estimated at over 17 million eligible citizens resident abroad, currently has no legally operational mechanism through which to exercise the constitutional right to vote in presidential elections. The Independent National Electoral Commission (INEC) has repeatedly cited the technical challenges of identity verification, ballot secrecy, double-vote prevention, and secure result transmission as barriers to diaspora enfranchisement. This study presents the design and development of Vote.ng, a Secure Web-Based Diaspora Voting System. Vote.ng is a kiosk-based polling unit platform that enables Nigerian diaspora citizens to cast their votes at INEC-designated embassy and consulate locations worldwide. The system is architected as a three-tier distributed application comprising a React single-page application frontend, a Node.js/Express REST API backend, and a PostgreSQL relational database, with an AI-assisted anomaly detection layer that monitors session behaviour for fraudulent access patterns. The system implements a multi-factor voter authentication pipeline: Permanent Voter's Card (PVC) number validation against a simulated INEC voter registry, National Identification Number (NIN) cross-verification, and real-time webcam-based facial recognition powered by the face-api.js library running in the browser. Vote secrecy is enforced through cryptographic identity-vote severance at the database layer: the votes table stores no voter identifier, making it structurally impossible to link a vote to a voter even with full database access. The AI component manifests in three domains: the face-api.js facial recognition engine for biometric identity verification, a rule-based anomaly detection module that flags suspicious session patterns, and an AI-powered audit log analysis assistant. All twelve functional requirements were satisfied in testing and all fifteen backend test cases passed.
| Item Type: | Thesis (Other) |
|---|---|
| Subjects: | Q Science > Q Science (General) |
| Divisions: | Faculty of Engineering, Science and Mathematics > School of Electronics and Computer Science |
| Depositing User: | MICHAEL MADUBUKO |
| Date Deposited: | 22 Jul 2026 15:04 |
| Last Modified: | 22 Jul 2026 15:04 |
| URI: | http://eprints.gouni.edu.ng/id/eprint/6059 |
Actions (login required)
 |
View Item |
