THE FORMATION OF THE SEMANTIC FIELD "EMPIRE" ON THE BASE OF DISTRIBUTIONAL AND STATISTICAL ANALYSIS.

The paper presents a part of a research project which is aimed at the creation of corpusdriven semantic field of the concept of "empire" in Russian, English and Czech languages. The result will be a lexical semantic net (thesaurus) with frequency characteristics and usage examples for lexical units. This paper presents results of automatic creation of the semantic field in Russian. Achievement of the goal is provided with the solution of a number of tasks, in particular, the corpus research of semantic relations of the word "empire" with other words based on distributional and statistical methods. The main tool of the study is Sketch Engine system. Text material is represented by corpora of Russian texts of XVIII-XX centuries. The major thematic content is history, literature, journalism, and philosophy. Results obtained show that usage of text corpora and corpus tools makes it possible to detect syntagmatic and paradigmatic relations in the automated mode and to create broader and more adequate filling of the lexical-semantic system. [ABSTRACT FROM AUTHOR]