Method of automated detection of punctuation asymmetry in parallel texts

S. D. Ignatova^a, A. A. Goncharov^a, N. V. Buntman<sup>b

a</sup> Federal Research Center "Computer Science and Control" of the Russian Academy of Sciences, 44-2 Vavilov Str., Moscow 119333, Russian Federation
^b M. V. Lomonosov Moscow State University, 1-52 Leninskie Gory, GSP-1, Moscow 119991, Russian Federation

Abstract: The paper explores the method of automated detection of interlingual punctuation asymmetry in parallel texts. Analyzing the functioning of punctuation marks requires a large scale of empirical data, which determines the use of parallel text corpora. The study outlines the potential of using search with exclusion in a parallel text database to automate detection of punctuation asymmetry in two languages. A search with exclusion involves identifying pairs of text fragments that contain certain language units in one language but do not contain any units from a defined set in another language. The feasibility of automated detection of punctuation asymmetry was tested by examining the use of the exclamation mark in Russian and French. Throughout the study, seven types of language substitutions were identified and quantitively analyzed.

Keywords: punctuation, interlingual asymmetry, parallel texts, search with exclusion, databases.

Received: 12.03.2025
Accepted: 15.04.2025

DOI: 10.14357/08696527250207