RECOGNITION AND CLASSIFICATION OF DISCOURSE STRATEGIES IMPLEMENTED BY ARTIFICIAL INTELLIGENCE

This research is a study of the potential of artificial intelligence, namely – the mechanisms of machine learning in terms of both recognition and classification of speech strategies and global discursive strategies on the example of public communication in social media. The research is aimed at conducting a comparative analysis of the results of strategic analysis implemented by artificial intelligence with those implemented by a human in order to define the possibilities and limitations of machine learning algorithms. The article consists of a theoretical part, including both a brief introduction and a review of some existing studies of machine learning mechanisms conducted in the field of computational linguistics, as well as fundamental research in pragmatic linguistics, namely critical discourse analysis. The practical research part includes fragments of illustrative material collected on the research topic. Each of the examples of the factual material collected and presented in the article is accompanied by two described variants of analysis of speech strategies and global discursive strategies implemented in each fragment. Linguistic and global discourse strategies, in turn, were recognized and classified both by the mechanisms of machine learning (with a brief description of their algorithms) and by the "living" human intellect of the author of the article in each of the examples, with the subsequent comparative analysis of the results. In addition, the research includes the graphical representation of its results, reflecting the statistical data of recognition and classification of discursive strategies realized by both artificial and human intelligence. The research conclusion includes both the conclusion regarding the achievement of the set research goals, as well as the formulated scientific novelty, theoretical and practical significance of the research