Two Computational Methods for Detecting Meter in Hungarian Poetry

Abstract

This paper presents two methods for the machine recognition of meter in Hungarian poetry, one for quantitative and another for qualitative meter. The algorithm analysing the former classifies the poems as dactylic, anapestic, iambic or trochaic, and defines the degree of regularity of the rhythm in the poems. The algorithm detecting the Hungarian qualitative meter returns the number of syllables of each beat. The paper also demonstrates the applicability of the two algorithms by using a test corpus containing the poems of eleven authors written between 1838 and 1941.