The Convergence of Blockchain, Elections, and Data Science
Ebiringa, Divine (2025)
2025
Julkaisun pysyvä osoite on
https://urn.fi/URN:NBN:fi:amk-2025101626070
https://urn.fi/URN:NBN:fi:amk-2025101626070
Tiivistelmä
This thesis investigates how blockchain technology and data science methods can jointly improve the security, transparency, and trustworthiness of electoral systems. Against a backdrop of rising concerns over election integrity in traditional voting systems, scholars and nations have begun to integrate blockchain systems. However, adoption appears to remain low.
This study systematically reviews 116 peer-reviewed articles and analyzes five national case studies (Estonia, Switzerland, the United States, Russia, and Romania). It examines (1) blockchain’s potential to secure voter verification, prevent fraud, and ensure immutable vote records; (2) data-science methods (e.g. machine learning, zero-knowledge proofs, homomorphic encryption) for anomaly detection, performance optimization, and privacy preservation; and (3) the ethical, political, and societal implications of digital voting, including the digital divide and regulatory compliance. A mixed-methods approach was adopted. First, SLR was used to identify, screen, and synthesize 28 core studies. Then multiple case studies were used to analyse real-world blockchain-voting system implementations to highlight practical successes and setbacks. From this, a conceptual framework for a hybrid blockchain voting system that integrates smart contracts, layered consensus model, and an off-chain data-science layer for real-time monitoring was developed.
Findings show consensus that blockchain can enhance election integrity, and data-science techniques further strengthen authentication, detect intrusions, and enable privacy-preserving analytics. However, it was also found that legal and regulatory gaps, infrastructure and literacy barriers, lack of scalability, and the need to build public trust remain huge hindrances to widespread adoption. The thesis recommended a hybrid voting architecture that combines public and private blockchains and off-chain data science monitoring, and its interface is user-friendly. Policymakers and election administrators are urged to pilot such integrated systems, refine identity-management protocols, and invest in voter education to ensure both technical robustness and broad societal acceptance, thereby paving the way toward more secure, transparent, and efficient democratic processes.
This study systematically reviews 116 peer-reviewed articles and analyzes five national case studies (Estonia, Switzerland, the United States, Russia, and Romania). It examines (1) blockchain’s potential to secure voter verification, prevent fraud, and ensure immutable vote records; (2) data-science methods (e.g. machine learning, zero-knowledge proofs, homomorphic encryption) for anomaly detection, performance optimization, and privacy preservation; and (3) the ethical, political, and societal implications of digital voting, including the digital divide and regulatory compliance. A mixed-methods approach was adopted. First, SLR was used to identify, screen, and synthesize 28 core studies. Then multiple case studies were used to analyse real-world blockchain-voting system implementations to highlight practical successes and setbacks. From this, a conceptual framework for a hybrid blockchain voting system that integrates smart contracts, layered consensus model, and an off-chain data-science layer for real-time monitoring was developed.
Findings show consensus that blockchain can enhance election integrity, and data-science techniques further strengthen authentication, detect intrusions, and enable privacy-preserving analytics. However, it was also found that legal and regulatory gaps, infrastructure and literacy barriers, lack of scalability, and the need to build public trust remain huge hindrances to widespread adoption. The thesis recommended a hybrid voting architecture that combines public and private blockchains and off-chain data science monitoring, and its interface is user-friendly. Policymakers and election administrators are urged to pilot such integrated systems, refine identity-management protocols, and invest in voter education to ensure both technical robustness and broad societal acceptance, thereby paving the way toward more secure, transparent, and efficient democratic processes.