PUBLIC SAFETY SYSTEM WITH ALARM TRIGGERING BASED ON J48 DECISION TREES CLASSIFICATION.

There is much research and work around the world on how to raise situation awareness for cases of natural disasters and catastrophes. It is often necessary to evaluate the extent of such events in advance with the intention to minimise the risks and the damage arising from such events. Our challenge comes from the public safety system. The goal is to use open data, and by using a classifier to recognise in advance when the extent of such events will be so large that it will be necessary to trigger an alarm within public safety system for the staff and all supporting services. We are using J48 decision tree as a classifier. Based on TP (True Positive), that is, of instances correctly classified as a given class for triggering alarm we can conclude that we can classify with a high probability when such events happen. We thus confirmed that it is possible to carry out classification based on open data. Such alarm triggering is an additional feature of public safety systems which can help to improve situational awareness further. [ABSTRACT FROM AUTHOR]