Blockchain-Based Convolutional Neural Network E-Voting Scheme

Abstract

SQL injection attack is now the most common server-side attack in web applications whereby malicious codes are injected into the database through user input fields by unauthorized users, and this could lead to data loss or in the worst case, to database hijacking. The utilization of blockchain technology in e-voting applications is not a new thing. Many systems have been proposed using cryptography and other security techniques. In those systems, minimal involvement of the third party is observed; a problem of coercion resistance and transparency maintenance at the same time is observed and most processes have not been implemented to evaluate the systems further. This paper applies the cryptographic signatures to validate the origin and integrity of the votes by preserving the voters’ choices during the election process. Then, the convolutional neural network (CNN), a regularized version of the multilayer perceptron in the system, is also applied to analyze visual imagery upon registration. Furthermore, authors provide several possible extensions and improvements that can be made as an addition to the scope of this research.