Analysis of Propaganda Elements Detecting Algorithms in Text Data

Abstract

This article is devoted to the problem of identifying propaganda in text files. Developing methods and techniques that can be used for this analysis is an important task, since the amount of propaganda is enormous. Such amounts of information cannot be analyzed by specialists. A study of how agitation is changing over time is valuable for understanding which areas of our life are particularly covered by propaganda, how its rhetoric is changing, and what impact it will have. All of the above indicates the relevance of research in this area. The objects of the research are the content of electronic media news, users, and the interrelations between them. The purpose of this work is to improve the accuracy of the classification of textual information through the appropriate use of existing methods of data mining using the most effective methods of text preprocessing and powerful machine learning algorithms for classification problems. The methods for solving the problem are considered to solve the problem of classifying text information for spam filtration tasks, contextual advertising, news categorization, creation of subject catalogs. That is why it is necessary to automate the process of searching, filtering and structuring text data. To solve this problem, the automated classification of texts is used - the task of machine learning from the field of natural language processing. The task of text classification has practical application in many areas, for example, spam filtering, contextual advertising, news categorization, creation of thematic catalogs. Most methods of automatic classification of texts are based on the assumption that the texts of each thematic heading contain certain features, the presence or absence of which indicates the belonging of the text of a rubric. The task of classification methods is to select the best following characteristics and formulate rules, will decide whether to refer the text to a certain category and conduct interactive drilling.