On Plagiarism and Software Plagiarism

TL;DR

This paper discusses the challenges of detecting software plagiarism, introduces Project Martial as an open-source solution, and surveys existing techniques and legal considerations in software similarity detection.

Contribution

It presents Project Martial, an open-source tool for code similarity detection, and provides a comprehensive survey of existing methods and legal aspects of software plagiarism detection.

Findings

Project Martial demonstrates effective code similarity detection.

Survey of fingerprinting, birthmarks, and embeddings techniques.

Legal analysis of software copyright cases.

Abstract

This paper explores the complexities of automatic detection of software similarities, in relation to the unique challenges of digital artifacts, and introduces Project Martial, an open-source software solution for detecting code similarity. This research enumerates some of the existing approaches to counter software plagiarism by examining both the academia and legal landscape, including notable lawsuits and court rulings that have shaped the understanding of software copyright infringements in commercial applications. Furthermore, we categorize the classes of detection challenges based on the available artifacts, and we provide a survey of the previously studied techniques in the literature, including solutions based on fingerprinting, software birthmarks, or code embeddings, and exemplify how a subset of them can be applied in the context of Project Martial.