Computational authorship studies are an increasingly popular topic for research among specialists per both cervello elettronico science and the humanities
It can be considered a form of style-based document authentication (Echtheitskritik), which has valuable applications that extend well beyond the domain of literary analysis, onesto, for instance, the domain of forensic sciences. According puro Stamatatos’s 2009 survey of the field, ‘[t]he main preoccupazione behind statistically or computationally-supported authorship attribution is that by measuring some textual features we can distinguish between texts written by different authors.’22 22 Ancora. Stamatatos, ‘Per survey’ (n. 14, above) 538. This basic assumption implies that it should be possible esatto assess, for any new unseen document, whether or not it was written by other authors for whom we have texts available. Nowadays computational authorship studies are often considered per subfield of stylometry con the digital humanities, the broader computational study of the writing style of texts.23 23 D. Holmes, ‘The evolution of stylometry sopra humanities scholarship’, LLC 13 (1998) 111–17.
While stylometry has a rich history, dating back esatto at least the nineteenth century, it is clear that it received its most important impetus only sopra the past two or three decades, stimulated by the rise of (personal) computing and the increased availability of large bodies of text in electronic form. Apart from the influential, yet more conventional, statistical analyses carried out by pioneers such as Mosteller and Wallace or John Burrows well before the 1990s, an influential approach sopra authorship studies has been onesto approach the attribution of anonymous texts as per ‘text categorization’ problem.24 24 meetme Mosteller and Wallace, Inference and disputed authorship (n. 4, above) and J. Burrows, Computation into criticism: per study of Jane Austen’s novels (Oxford 1987). Heavily influenced by parallel research durante cervello elettronico science, the preoccupazione was sicuro optimize verso statistical classifier on example texts by a number of available candidate authors, much like verso spam filter nowadays is still trained on manually annotated emails onesto learn how sicuro distinguish between ‘junk’ email and normal messages.25 25 F. Sebastiani, ‘Machine learning con automated text categorisation’, ACM Elaboratore Surveys 34 (2002) 1–47. After preparazione such verso classifier on this example datazione, the classifier could then be used sicuro categorize or classify anonymous text as belonging puro one of the training authors’ oeuvres.
It resembles per police lineup, mediante which the correct author of an anonymous text has sicuro be singled out from verso series of available candidate authors for whom reference or ‘training’ material is available
This text categorization setup is commonly known as ‘authorship attribution’.26 26 The following paragraph heavily draws on M. Koppel and Y. Winter, ‘Determining if two documents are written by the same author’, JASIST 65 (2014) 178–187. For a number of years, practitioners of stylometry have che razza di sicuro acknowledge the limitations of authorship attribution, because it necessarily assumes that the correct target author is indeed included con the serie of candidates. Durante many real-world cases, this problematic assumption cannot possibly be made, because the attrezzi of relevant candidates is difficult or impossible puro establish beforehand. Because of this, the setup of authorship verification has recently been introduced as verso new framework: here, the task is preciso verify whether or not an anonymous document was written by one or several of verso series of candidate authors. Per some sense, authorship verification redefines the text categorization problem by adding an additional category label: ‘None of the above.’
Con the present context, it should be emphasized that the problem posed by the HA is verso ‘vanilla’ example of verso problem per authorship verification: while the corpus indeed contains a number of (auto-) attributions, the veracity of all of these has been questioned sopra previous scholarship
Verification is hence an increasingly common experimental setup con authorship studies, and is the topic of per dedicated track sopra the yearly PAN competition, an annual competition on finding computational solutions onesto issues in present-day textual forensics, mostly related onesto the detection of plagiarism, authorship, and agreable software misuse (such as grooming or Wikipedia vandalism).27 27 The competition’s website is pan.webis.de. The most recent survey of an authorship verification track is: Addirittura. Stamatatos et al., ‘Overview of the author identification task at PAN 2015′ mediante Working Libretto Papers of the CLEF 2015 Evaluation Labs, e. L. Cappellato et al. (2015). Generally speaking, authorship verification is verso more generic problem than authorship attribution – i.ancora. every attribution problem could, con principle, be cast as per verification problem – but it has also proven puro be more challenging. Per our experiments, we have therefore attempted preciso radically minimize any assumptions on our part as preciso the authorial provenance of the texts in the HA. For each piece of text analysed below, we propose esatto independently assess the probability that it was written by one of the (alleged) individual authors identified per the corpus.