| Summary: | This article demonstrates that metre is a privileged indicator of authorial style in classical Latin hexameter poetry. Using only metrical features, classification experiments are performed between the works of six authors using four different machine-learning models. The results showed a pairwise classification accuracy of at least 90% with samples as small as ten lines and no greater than seventy-five lines (up to around 500 words). In a multiclass setting, classification accuracy exceeded 95% for all four algorithms when using eighty-one-line chunks. These sample sizes are an order of magnitude smaller than those typically recommended for BOW ('bag of words') or n -gram approaches, and the reported accuracy is outstanding. Additionally, this article explores the potential for outlier (forgery) detection, or 'one-class classification'. As an example, analysis of the disputed Aldine Additamentum (Sil. Ital. Pun. 8:144–223) concludes (P < 0.0001) that the metrical style differs significantly from that of the rest of the poem.
|
|---|