The constant growth in the number of users of the Internet and the improvement in technical capabilities of communications allow the use of various tools for the rapid notification of the population about the events occurring in the world. Depending on the type of source, models of information dissemination differ. When analyzing the information it is necessary to determine the relationship of signal distribution channels, determination of the primary source, etc. The article examines the dissemination of information messages in open networks using messages on religious topics using visual analytics. The paper specifies the ways to identify the messages of the required topic, as well as the visualization of the content. For topic modelling Latent Dirichlet Allocation (LDA) is used. The applicability of various dimensionality reduction and clustering algorithms for the interpretation of clustering results is considered. The developed methods can be scaled to analyze information events in different thematic areas.