Plagiarism and authorship analysis: introduction to the special issue

The Internet has facilitated both the dissemination of anonymous texts as well as easy ‘‘borrowing’’ of ideas and words of others. This has raised a number of important questions regarding authorship. Can we identify the anonymous author of a text by comparing the text with the writings of known authors? Can we determine if a text, or parts of it, has been plagiarized? Such questions are clearly of both academic and commercial importance. The task of determining or verifying the authorship of an anonymous text based solely on internal evidence is a very old one, dating back at least to the medieval scholastics, for whom the reliable attribution of a given text to a known ancient authority was essential to determining the text’s veracity. More recently, the problem of authorship attribution has gained greater prominence due to new applications in forensic analysis, humanities scholarship, and electronic commerce, and the development of computational methods for addressing the problem. Over the last century and more, a great variety of methods have been applied to authorship attribution problems of various sorts. One can roughly trace the evolution of methods through three main stages. In the earliest stage researchers sought a single numeric function of a text to discriminate between authors. In a later stage, statistical multivariate discriminant analysis was applied to word frequencies and related numerical features. Most recently, machine learning methods and highdimensional textual features have been applied to sets of training documents to