The technology today has advanced to an extent that the computational algorithms and computer based data analysis can be used to make large and effective decisions. An essential role of such technologies in human life is felt when they provide ease in various domains and help make future plans for managing risks. Among all the various issues, one of the major issues that comes into picture is the disaster issue due to which every year the world loses many of its residents. Hereby, we propose a solution to this major issue.
The proposed work includes three main contributions for prediction based technique development. Firstly, we generate the test and training dataset using Google API. In this, data processed is filtered and transformed from an unstructured to structured dataset. Our next challenge is to develop an accurate and precise structure learning of the disasters, their places of origin and their nature. The methods to be used used would be k-means clustering and hidden Markov model. For the final stage, we will evaluate the probablities of occurrence of disasters in the future based on the dataset of the current scenario. We will implement this using Python.
In this proposed solution the supervised data mining technique would be used to evaluate the historical data and the natural events and predict the upcoming events. This task can be efficiently performed with the help of data mining algorithms by analysing the news contents from various online news sources after checking their reliability. Therefore, different processes of the data mining techniques like machine learning algorithms would be combined together for developing the enhanced scheme for the prediction. In order to prepare the proposed data model, there is a need to perform the collection of data, pre-processing of data to create a structured dataset, data model development and finally utilizing the data model for predicting the upcoming natural disasters. This would greatly reduce the number of lives that are compromised due to sudden encounter of natural events.