Detecting bursts of interest among user communities on social media towards the various publications during the era of remote work and interaction, which began with COVID-19, is more relevant than ever. The paper considers the problems of detecting activity bursts in various communities of an online social network. A survey of burst detection methods is presented, including real-time monitoring. The questions of terminology and classification of methods concerning the object of study are considered. Monitoring options are proposed and discussed. Particular attention is paid to the technical issues of developing a related information system, its modular architecture and software implementation tools. For the assessment of possible deviations (and further development of metrics) from the mean values, preliminary monitoring results are obtained using various partitioning methods for the interval of bursts (between a mean and maximum value). Conclusions are drawn about the applicability of the proposed method for detecting bursts and stability of a community audience interest burst characteristics while this audience doesn’t substantially change and the thematic direction of the community is not shifting.