AUTOMATIC EXTRACTION OF COLLOCATIONS FROM A TEXT CORPUS
Світлана Валентинівна Петрасова, Maria Kuzmina · Bulletin of the National Technical University KhPI Series Actual problems of Ukrainian society development · 2018
The article deals with the method for automatic extraction of collocations from the text corpus of the Ukrainian language. Definitions of the term "collocation" have been considered from the viewpoint of approaches to its analysis in modern corpus linguistics. Existing statistical methods and systems that use statistical measures for extraction of collocations have been analyzed, and their features have been described. The corpus of technical documentation has been developed and its structure has been described. To extract collocations from the texts of the Ukrainian language that consist of the instructions of technical documentation, the developed software implementation of MI measure has been described.