DISINFORMATION, FAKES AND PROPAGANDA IDENTIFICATION METHODS IN MASS MEDIA BASED ON MACHINE LEARNING

This article provides a comprehensive research of modern approaches used to identify fakes and propaganda. The study begins by examining the prevalence and impact of misinformation, emphasising the need for advanced identification methods. Tracing the evolution of methods, the article delves into historical approaches and their adaptation to the dynamic media landscape. Central to the discussion are advanced techniques such as natural language processing, which analyses linguistic patterns to detect inconsistencies in textual content. The article explores the benefits of natural language processing, including scalability and real-time analysis, along with the limitations associated with contextual ambiguity and the development of linguistic nuances. Multimodal analysis takes centre stage, offering a holistic perspective through the textual and visual elements integration. The article explores the benefits of multimodal analysis, highlighting its potential for comprehensive understanding and increased accuracy, while acknowledging the challenges associated with analytical complexity and contextual nuance. Machine learning is emerging as a dynamic tool for pattern recognition and ada