Identification of Disaster Prone Areas-A Machine Learning Approach

摘要

In the earlier days when natural disasters are occurred that information is communicated through phone calls, telegram, direct observations or personal interview due to which relief operations used to get late; thus human lives and animal mortality will get increases and sufferings eventually increased. The internet technologies developed are used now days to some extent to control the rate of the sufferings. Tweets are the fast and real-time sources for information. We perform various approaches to identify which process can work faster compare to others. First we collect data from social media which is considered as fastest medium to reach vast number of people. Then we classify according to our needs and make clusters using algorithm. For better understanding, we use various tools to visualize and to identify the specified region to provide necessities like food, shelter and medicines in an earliest possible time.