Software Authorship Analysis as a Part of Software Forensics
The general field of software forensics subsumes quite a few areas and copyright infringement forensics is only one such area. Another area, related to copyright infringement forensics and yet different from it, is software authorship analysis. Although the objectives of authorship analysis procedures are different from those of software copyright infringement forensics, their functional areas can sometimes overlap, as proof of authorship can sometimes confirm copyright infringement. It is these overlapping areas that make them attractive to experts involved software copyright infringement forensics. This chapter primarily focuses on giving an overview of authorship analysis and then explaining how authorship analysis can be of help, although in a limited way, in software copyright infringement forensics.
Software authorship analysis is often carried out with the help of several mathematical models. Each of these models has a set of forensic procedures to implement the model. All these models try primarily to find features that help identify the author of a software and to decide if any given program has been written by a particular author (or suspect author, in case of a dispute). In addition to this primary use and also a couple of other secondary uses, many of these models are capable of identifying similarities between two sets of source codes that can be attributed to the same author in order to detect plagiarism, piracy and sometimes, even copyright infringement. Before proceeding with software authorship analysis, a preliminary discussion on its roots in natural languages would add value.