Authorship Verification for Hired Plagiarism Detection

Abstract

Plagiarism detection is an important tool in modern academia. With growing class sizes and the modernization of the internet, there have been more ways that allow plagiarism to excel in modern culture. Methods such as patchwriting – where an individual may copy, paste and possibly modify the content – and commissioned writing – where an individual hires another person to do the work for them – are not considered by modern plagiarism detectors. This work aims to give instructors a way to identify and detect plagiarism in student writing that addresses these difficult-to-detect issues using artificial intelligence. We introduce a tool to aid instructors in detecting plagiarism that adapts to each students’ individual writing style as they submit writing assignments. This work incorporates artificial intelligence and natural language processing that identifies the ways in which a student writes based on a collection of their essays. The proposed Authorship Verification for Hired Plagiarism Detection (AVHPD) tool includes document storage, a clean user interface, and intuitive break-downs of how a given writing sample differs from prior samples.