Abstract: Continuous Authentication (CA) comprises of observing and checking over and again client conduct surrounded by a registering session keeping in mind the end goal to separate in the middle of real and impostor practices. Stylometry examination, which comprises of checking whether a target achieved was composed or not by a particular individual, could conceivably be utilized for Continuous Authentication. In this work, existing stylometric peculiarities and add to another origin check model relevant for consistent verification. Existing lexical, syntactic, and application particular peculiarities, and propose new feature in light of n-gram investigation. At first with a huge peculiarities set, and recognize a diminished number of client particular features by processing the data pick up. Moreover, the methodology incorporates a system to dodge issues with respect to uneven dataset which is an inalienable issue in stylometry investigation by utilizing Naive Bayes classifier for characterization. Test assessment taking into account the Enron email dataset including 76 creators yields extremely guaranteeing results comprising of an Equal Error Rate (EER) of 12.42% for message pieces of 500 characters.
Keywords: Continuous authentication; biometrics systems; security; creation check; classification; stylometry; n-gram features; short message confirmation; content mining; write print.
Title: A Framework for Continuous Authentication Utilizing Stylometry for E-Mail Author Identification and Authentication
Author: Ms. Sunita Jadhav, Mr. Ankeet Jadhav, Mr. Raunak Saraf, Mr. Dhananjay Dahadade
International Journal of Computer Science and Information Technology Research
ISSN 2348-120X (online), ISSN 2348-1196 (print)
Research Publish Journals
