Лексика как классифицирующий признак современной поэзии [Vocabulary as a Classifying Feature оf Russian Postmodern Poetry]

Abstract

The article discusses the possibility of classifying poetic books on the basis of their vocabulary. The distance between 190 poem collections is calculated as the Euclidean distance between books’ vocabularies, for each element of which the value of TF-IDF (term frequency – inverse document frequency) is calculated (each book has 190 measurements with this method of calculation). Using t-SNE (t-distributed stochastic neighbor embedding), these measurements are reduced to two, and the K-means clustering method is applied to the resulting structure. With such a classification method, poets are grouped on the basis of their originality / similarity, which in turn helps to overcome more traditional classifications based on poets’ generations or literary schools.