Big Data Implementation of Natural Disaster Monitoring and Alerting System in Real Time Social Network using Hadoop Technology

Background/Objectives: The information generated by the social networks is exponentially higher and demand effective systems to yield effective results. In conventional techniques stay unqualified because they ignore the social related data. Methods/Statistical Analysis: The existing system doesn’t provide any proper methodology to report people about any natural disasters rapidly or to notify the rescue agencies for taking immediate actions to carry out the rescue process. The only existing methodology of reporting people is the media, i.e. news, radio, etc. The use of a particle filter fetches the necessary keywords from tweets by the use of Stemming along with the location and time. When the system encounters keywords related to natural disasters, an auto alert is sent to the people in the nearby locations and the rescue teams based on a proper verification algorithm. Findings: The method finding the two data set representations: one is considering the two directional social relations, and the other considering the one directional social relation. It is seen that the performance of the recommender system can be greatly boosted by the mentioned contextual factors. Application/ Improvements: On validation of the authenticity user, the user is allowed to enter the application. The user is allowed to tweet using the same application.