In this article, taking the anti-terrorism system of the train as an example, and integrating of the current resources and the data, author establish the warning platform, and the new model of purchasing the railway ticket scientifically, in order to warn the terrorism. The existing anti-terrorism model is imprecise, and it is very single model which we use to hunt for terrorist. Especially in the periods of security. By analyzing the existing tickets purchasing system, and the writer build a proper data analysis model on the ticket purchasing system, in order to check the ticket purchasers, and listing the suspicious person, so that we can take effective measures before terrorist acts. After years of development, the big data technology has become increasingly mature, and played a very important role in the commercial field. The big data warning technology relies on the big data technology. The purpose is forecasting and alarming the potential terrorist threats,[1] and analysis of intelligence of terrorism.[2] However in terms of crime warning and criminal investigation, it is necessary to research further how the technology serve the anti-terrorism and anti-riot work. Actually, due to the wide range of criminology research,[3] at present, there are data barriers and inaccurate prediction in most countries. In this research, by using the existing train ticket database system, the author tries to set up a new data warning model, which can provide timely warning and judgment for terrorist activists.