With the rapid development of the Internet and the media, all kinds of information are coming, people are often difficult to distinguish the truth and falsehood of these information, confused by many false information, unable to get the required information from many information correctly. In particular, in the environment of the epidemic situation, information on epidemic rumors poses a huge challenge to epidemic prevention and control. Based on the epidemic information data, this paper uses web crawler crawling epidemic information as the data source of epidemic analysis. Based on data visualization technology, the time line distribution and the relationship of likes comments forwarding of epidemic information and epidemic rumors are analyzed in detail. Finally, on the basis of data visualization analysis, through Naive Bayes model to train and test the epidemic information. In this model, the prediction accuracy of epidemic cases is 0.91, and the prediction accuracy of epidemic rumors is 0.90. The results show that the identification model of epidemic rumor information has good effectiveness and can provide theoretical and technical support for the epidemic prevention work of relevant departments.