Corpus and AI-powered constructing of databases for determining connotative characteristics of news texts.

Abstract

Aim. To develop a database structure for determining connotative characteristics of news text.  Methodology. The key research methods are modeling, experimentation, and the author’s method of professionally oriented programming. The procedure includes design, writing, and debugging the programming code. Results. Experimental software for constructing three databases was developed. Using German news texts, an experiment was successfully conducted to collect news feeds, construct databases of potentially connotative phrases, mark them with numerical markers, and optimize them. Research implications. The study expands the methodological basis of applied linguistics and also makes it possible to use the created algorithms for the construction and author’s markup of databases for their application in software systems for determining the connotative characteristics of given news flows.