The article justifies the feasibility of using the social network X (Twitter) as a sensor of the behavior of cryptocurrency market participants. Based on the analysis of more than 4 million tweets published between July 2022 and September 2023, the key text and metadata of the publications were examined, such as hashtags, language, emotional coloring, user mentions, audience reactions, as well as sources and methods of publication. Data analysis and visualization methods were applied to identify information trends, identify popular topics, and assess the level of engagement. It is shown that the frequency of mentions, the emotional background of publications, and structural features of content can be integrated into machine learning models to predict the dynamics of virtual asset prices. Social signals from the X platform reflect collective sentiments, behavior patterns, and reactions to events in real time, which opens up new opportunities for building informative models within the framework of behavioral economics. The proposed scientific and practical approach can be used to monitor changes in the sentiment of cryptocurrency market participants and increase the accuracy of market dynamics forecasts.