AUTOMATED MACHINE LEARNING TECHNOLOGIES FOR NATURAL LANGUAGE PROCESSING SYSTEMS

Building high-quality artificial intelligence systems depends on the knowledge of machine learning experts and requires a large amount of work, which prevents the widespread use of these technologies in various fields of human activity.This paper discusses modern approaches and systems of automatic machine learning.Automatic machine learning tools simplify the process of data collection, feature development, model creation and evaluation, and therefore facilitate the implementation of artificial intelligence systems, including natural language processing systems.The article describes the peculiarities of applying machine learning technologies to various natural language processing tasks.Purpose.Analyzing the state of development of the methodological framework for the development of artificial intelligence systems and automatic machine learning systems, in particular, for the construction of natural language processing systems.Method.The study was conducted using a systematic approach, methods of abstraction, system analysis, comparison and synthesis.Results.We have reviewed the stages of creating artificial intelligence systems and the approaches used to build systems using automatic machine learning technology.As a result of this work, an overview of automatic machine learning approaches and systems is presented.Scientific novelty.We have solved an urgent scientific task, which is to review modern automatic machine learning systems and approaches.Practical importance.It is the possibility of applying the scientific provisions and conclusions of the study for the development and implementation of artificial intelligence systems; in the process of teaching natural language processing disciplines in higher education institutions; in writing manuals on natural language processing; in conducting applied research.