Gender Factor in Associative Links of Words: Dictionary and Distributive-Semantic Model Data

Abstract

The aim of the work is to make a comparative analysis of the associative links of full-meaning words from the upper zone of the frequency list, compiled on the basis of a research corpus of blog texts in Russian, in a psycholinguistic experiment and the distributivesemantic model Global Vectors (GloVe), trained on this corpus. The relevance of the work is due to the need for a comprehensive study of the psychologically relevant meaning of the word. The novelty of the study lies in the fact that such an analysis is carried out taking into account the gender factor of the respondent / author of the text. The use of a set of methods for data mining (clustering, classification) and visualization of its results made it possible to establish the influence of gender on the composition of the semantic associates of the analyzed words (that is, words with close vectors in the distributive-semantic model) and the absence of such an effect in their associates recorded in the associative dictionary. As the study showed, distributivesemantic models and dictionary associative norms reflect different aspects of the psychologically relevant content of the word and should be used as complementary sources when modeling the psychologically relevant meaning of the word, taking into account the individual characteristics of the speaker, while conducting such an analysis it is advisable to use data mining methods.

Keywords

• ассоциативный эксперимент
• дистрибутивная семантика
• дистрибутивно-семантическая модель
• корпус текстов
• интеллектуальный анализ данных