Open AccessSemantic Studies of Organisation and Functioning of Mental Lexicon
Author(s) -
Vadim A. Belov
Publication year - 2020
Publication title -
naučnyĭ dialog/naučnyj dialog
Language(s) - English
Resource type - Journals
eISSN - 2227-1295
pISSN - 2225-756X
DOI - 10.24224/2227-1295-2020-8-29-51
Subject(s) - mental lexicon , lexicon , computer science , natural language processing , connectionism , semantic similarity , semantic network , artificial intelligence , vocabulary , semantic compression , semantics (computer science) , linguistics , semantic computing , artificial neural network , semantic technology , semantic web , philosophy , programming language
The review article discusses the key problems of semantic organization of the mental lexicon. It is understood as a dynamic, cognitively organized semantic network of lexical units. The paper presents the characteristics of the main models of semantic organization of mental vocabulary, highlights the advantages and disadvantages of each approach. It is noted that currently connectionist models are developing most actively. Among them there are the following: the small world network, which considers a certain fragment of the lexicon; thesaurus models that combine all units of the lexicon; computational and distributive models that build relationships between a large number of units based on corpus data about shared usage. The author analyses the sources of information about the organization of the mental lexicon, among which the leading position is occupied by the results of associative experiments and priming. It is revealed that research is also carried out on the subjects’ intuitive assessments of word similarity, analysis of unintentional speech errors. Neuropsychological technologies are used to study the mental lexicon. The problem of semantic relations in the mental lexicon, which are described using a “spatial” metaphor, is discussed separately: semantic proximity is considered as a distance or a vector.