<![CDATA[This study aims to classify public opinion regarding the determination of national disaster status in Sumatra using data obtained from social media platform X (formerly Twitter). Public opinion data were collected from user-generated text posts related to flood and landslide events that occurred in November 2025 through a crawling process utilizing Node.js and the tweet-harvest library. The collected data underwent preprocessing stages, including noise removal, text normalization, and duplicate elimination. Subsequently, the textual data were transformed into numerical representations using the Term Frequency–Inverse Document Frequency (TF-IDF) weighting method through the Process Documents from Data operator in RapidMiner. Clustering analysis was conducted using the K-Means algorithm with five clusters. The results indicate that public opinion is classified into several dominant themes, namely government policy decisions, humanitarian assistance, disaster causation, housing needs for affected communities, as well as security and field monitoring activities. These findings demonstrate that the K-Means algorithm is effective in identifying patterns of public opinion from unstructured textual data and can support data-driven evaluation of disaster management policies.]]>
Copyrights © 2026
