Abstract
To solve problems with emergencies, you need to have an idea in which regions more different emergencies occurred, in order to identify some kind of report and analysis. It is also not enough to have only the location of the emergency itself, you need to have different types of information about the emergency itself. This work describes a system that displays different types on a map with the necessary information about from sources of news sites. RandomForest and xgboost machine learning methods were used for forecasting. With the help of the geographic information system, types of emergencies are visualized on the maps of the city of Almaty. All visualization algorithms and technologies are implemented in Python.