опрацювання природної мови

Data Set Formation Method for Checking the Quality of Learning Language Models of the Transitive Relation in the Logical Conclusion Problem Context

A method for data set formation has been developed to verify the ability of pre-trained models to learn transitivity dependencies. The generated data set was used to test the quality of learning the transitivity dependencies in the task of natural language inference (NLI). Testing of a data set with a size of 10,000 samples (MultiNLI) used to test the RoBerta model.

Construction of Dialogue Student-pc System on the Basis of Natural Language Exchange Using the Java Environment

The review and analysis of natural language processing methods (NLP) was carried out. Definitely a global criterion for evaluating the effectiveness NLP funds. Recommendations on the expediency of the use NLP funds are given

Розпізнавання багатослівних конструкцій

This paper surveys the problem of multiword expressions (MWE), which plays the important role in development of large-scale, linguistically sound natural language processing technology. Multiword expressions are expressions which are made up of at least 2 words and which can be syntactically and/or semantically idiosyncratic. This category includes such constructions as compound nouns, idioms and phrasal verbs.This paper deals with modern approaches to MWE stratification, extraction and identification.

Граматична корекція речень з використанням графу взаємовиключних гіпотез

A method for automatic correction of Ukrainian sentences is introduced. The method is based on dependency grammar and utilizes mutually exclusive hypothesis graph for word sense disambiguation. 37 % of ambiguous sentences which were correctly corrected as opposed to 14 % corrected by spell checker.