MULTIVARIATE STATISTICAL ANALYSIS OF SYNONYMS BASED ON A CORPUS: A CASE STUDY OF VERBS WITH THE MEANING OF “ALLOWANCE”

Abstract

This study presents a corpus-based multivariate statistical analysis of synonymous relationships using Russian verbs with the meaning of “allowance” as material. The objects of the study are five verbs: razreshat’, pozvolyat’, dozvolyat’, dopuskat’, and davat’. The empirical base consists of 1000 sentences selected from the National Corpus of the Russian Language. The research methodology is based on multivariate statistical analysis using logistic regression. During the work, linguistic factors characterizing the use of the studied verbs are identified and then integrated as variables into the regression model. Analysis of the obtained regression coefficients allowed determining the degree of influence of each linguistic factor on the choice of a specific verb and identifying general trends in their usage. The results of the study reveal the specifics of the use of each of the five verbs under study, highlighting their semantic nuances and combinatorial preferences. The application of multivariate statistical analysis based on corpus data provided an objective assessment of the semantic and syntactic potential of the studied verbs. The use of logistic regression, taking into account a number of linguistic factors, made it possible to identify statistically significant patterns in the use of the studied verbs with the meaning of “allowance”.