Application of density based clustering of disaster location in realtime social media

Abstract

Online life is used as a method of finding information, one of which is Twitter as the medium. The occurrence of natural disasters is very detrimental. Therefore, the application is needed to see natural disasters through social media Twitter. A small number of studies using clustering methods based on Twitter user data density are the beginning of this research. With the availability of data in certain areas makes it easy to group. After that, the data is grouped based on a high degree of similarity. One result of applying this method is the location of the disaster. NER-based rules are used to discover out the area of the disaster. Data accuracy testing is performed using the Silhouette coefficient.