Pro-russian propaganda recognition and analytics system based on text classification model and statistical data processing methods

Abstract

The object of the research is the processing of textual data in the context of recognizing and analyzing pro-Russian propaganda. This article reviews the main methods and existing approaches to text classification and features of propaganda classification. The aim of the research is to develop a new system based on neural networks, which can both determine the political polarity of transmitted message sources and build various analytics based on collected data related to the Ukrainian-Russian war. To achieve this goal, a classification model based on the BERT text transformer is proposed, which is capable of classifying texts with higher accuracy than classical analogues and a number of algorithms that use the developed classification model to build various representations. Ref. 19, pic. 4, tabl. 1

Keywords

• neural networks
• text classification
• text transformers
• news
• propaganda
• hashtags
• networks
• telegram
• twitter