Automatic Extraction of Rhythm Figures and Analysis of Their Dynamics in Prose of 19th-21st Centuries

Abstract

The paper is devoted to automatic detection of rhythm in fiction and investigation of how rhythm of prosaic texts changed over 19th–21st centuries, based on results of such detection. The authors developed algorithms, which extract rhythm figures related to word repetitions (anaphora, epiphora, polysyndeton, etc.), and visualized their statistical features in plots and heat maps by decades on the material of British and Russian literature. The experiments allowed to find rhythm changes over periods and give interpretation of their reasons from a linguistic point of view.