<![CDATA[Some of the natural disasters in Indonesia are floods, earthquakes, eruptions and others. In dealing with this, the classification of types of disasters is very crucial to determine appropriate steps and plans. Technology can be used to facilitate the grouping process, one of which is by utilizing text mining techniques. Clustering of information is done by entering into several clusters on the basis of interrelationships between words using the K-Means algorithm. In this study, it aims to produce a model for grouping natural disasters in Indonesia by applying the K-Means algorithm. The analysis is based on data from community comments about natural disasters on social media Twitter. The use of the text mining method with the RStudio application succeeded in grouping natural disasters based on their potential and type from community commentary data on Twitter social media. After carrying out text cleaning, text processing, and the TF-IDF method, it is known that floods and earthquakes are the highest natural disaster topics from data mining. The unsupervised method with the K-Means algorithm is used to build topic groups based on the distance between words. The evaluation was carried out using the Sum of Square Error and Silhoutte Coefficient methods, and obtained an accuracy of 75.0% and 96.7%. The conclusion is that the K-Means algorithm is successful in building topic groups based on the distance between words in the community's commentary data about natural disasters on Twitter.]]>
