Extreme Event Detection and Management using Twitter Data Analysis

Events like the natural disasters are impossible to predict and are inevitable. The best course of action in that scenario is to effectively detect the event and respond to it accordingly. With quick detection and tangible information retrieval, casualties can be greatly reduced. Traditionally, this has been a tiring task as the data was siphoned from human observers or expensive equipment. In countries with scarce economies, the acquisition of these equipment can prove herculean and relying on individual human observers may lead to the spread of misinformation. An inexpensive and reliable solution to this problem can be found in the form of a system which facilitates the detection of extreme real-world events through the reactions of people on online social networks like Twitter. Unusual bursts in traffic from Twitter can be detected and this can be associated with an event. Moreover, the system can monitor the Twittersphere for rescue requests that appear on Twitter at the aftermath of an extreme event. Thanks to the containerized designed and hence its scalability, this model can undergo worldwide deployment with ease.