Big data and social media: analysis of key media data phenomena for predicting public opinion

in the era of digitalization, big data from social media has become a valuable resource for analyzing and predicting public opinion. This article is devoted to a comprehensive study of key phenomena inherent in media data and their role in building accurate predictive models. The focus is on phenomena such as content virality, the formation of information cascades, the dynamics of network discourses, and the identification of latent topics using thematic modeling methods. We examine in detail how these phenomena, extracted from massive unstructured data (texts, likes, reposts), serve as predictive indicators for assessing moods and trends in society. To this end, we apply a combination of machine learning methods (including sentiment analysis and clustering) and network analysis. In conclusion, the article discusses the potential ethical challenges and limitations associated with the use of big data and suggests possible avenues for the development of this interdisciplinary field of research.

Big data, social media, public opinion, forecasting, sentiment analysis, network analysis, machine learnin