Let quantitativo represent an unknown document and let y represent verso random target author’s stylistic ‘profile’. During one hundred iterations, it will randomly select (a) fifty per cent of the available stylistic features available (di nuovo.g. word frequencies) and (b) thirty distractor authors, or ‘impostors’ from verso pool of similar texts. Con each iteration, the GI will compute whether interrogativo is closer to y than to any of the profiles by the thirty impostors, given the random selection of stylistic features in that iteration. Instead of basing the verification of the direct (first-order) distance between interrogativo and y, the GI proposes preciso record the proportion of iterations con which interrogativo was indeed closer puro y than puro one of the distractors sampled. This proportion can be considered verso second-order metric and will automatically be a probability between zero and one, indicating the robustness of the identification of the authors of interrogativo and y. Our previous work has already demonstrated that the GI system produces our teen network app gratuita excellent verification results for classical Latin prose.31 31 Padrino the setup sopra Stover, et al, ‘Computational authorship verification method’ (n. 27, above). Our verification code is publicly available from the following repository: This code is described per: M. Kestemont et al. ‘Authenticating the writings’ (n. 29, above).
For modern documents, Koppel and Winter were even able to report encouraging scores for document sizes as small as 500 words
We have applied per generic implementation of the GI preciso the HA as follows: we split the individual lives into consecutive samples of 1000 words (i.ed. space-free strings of alphabetic characters), after removing all punctuation.32 32 Previous research (see the publications mentioned per the previous two notes) suggests that 1,000 words is verso reasonable document size in this context. Each of these samples was analysed individually by pairing it with the profile of one of the HA’s six alleged authors, including the profile consisting of the rest of the samples from its own text. We represented the sample (the ‘anonymous’ document) by per vector comprising the incomplete frequencies of the 10,000 most frequent tokens per the entire HA. For each author’s profile, we did the same, although the profile’s vector comprises the average incomplete frequency of the 10,000 words. Thus, the profiles would be the so-called ‘mean centroid’ of all individual document vectors for per particular author (excluding, of course, the current anonymous document).33 33 Koppel and Seidman, ‘Automatically identifying’ (n. 30, above). Note that the use of per solo centroid per author aims puro reduce, at least partially, the skewed nature of our data, since some authors are much more strongly represented con the insieme or preparazione pool than others. If we were not using centroids but mere text segments, they would have been automaticallysampled more frequently than others during the imposter bootstrapping.
Esatto the left, a clustering has been added on apice of the rows, reflecting which groups of samples behave similarly
Next, we ran the verification approach. During one hundred iterations, we would randomly select 5,000 of the available word frequencies. We would also randomly sample thirty impostors from a large ‘impostor pool’ of documents by Latin authors, including historical writers such as Suetonius and Livy.34 34 See Appendix 2 for the authors sampled. The pool of impostor texts can be inspected sopra the code repository for this paper. Durante each iteration, we would check whether the anonymous document was closer esatto the current author’s profile than onesto any of the impostors sampled. Mediante this study, we use the ‘minmax’ metric, which was recently introduced per the context of the GI framework.35 35 See Koppel and Winter, ‘Determining if two documents’ (n. 26, above). For each combination of an anonymous text and one of the six target authors’ profiles, we would primato the proportion of iterations (i.ancora. verso probability between zero and one) sopra which the anonymous document would indeed be attributed to the target author. The resulting probability table is given mediante full sopra the appendix sicuro this paper. Although we present per more detailed colloque of this giorno below, we have added Figure 1 below as an intuitive visualization of the overall results of this approach. This is per heatmap visualisation of the result of the GI algorithm for 1,000 word samples from the lives con the HA. Cell values (darker colours mean higher values) represent the probability of each sample being attributed to one of the alleged HA authors, rather than an imposter from a random selection of distractors.