Inspired Symbol-based Authentication against Dictionary attacks Using ML Algorithm

Abstract

Verifying an individual's identity before granting them access to a linked device, an internet service, or another resource is known as user authentication. The significance of authentication lies in its ability to safeguard data, apps, and networks for companies by limiting access to their protected resources to only authorized individuals or processes. This paper used the popular Big data technology Apache Spark for storing, and processing large data and proposed a novel authentication framework. A viable replacement for conventional alphanumeric passwords, bio-metric and facial authentications with dynamic symbol selection as an authentication. This authentication is tested in the Apache spark cluster which is the most distributed system. In these methods, SHA512 cryptography is used in several ways and comparison is done using existing authentication and machine learning algorithms. The straightforward authentication scheme and applied Apache Spark distributed system with 10 nodes produced the best results.