Worst Case Scenario: Analyzing Big Data in Disaster Crisis

The industrial and information age has brought significant changes in the way we manage risk and response to disasters. In this context, Big Data analysis is becoming an important tool in improving the effectiveness of crisis communication. This approach enables faster and more informed decision-making in worst-case scenarios, which can ultimately minimize damage and loss of life. However, challenges such as data overload, misinterpretation, and the spread of disinformation remain obstacles that need to be overcome. By analyzing 14,083 tweets using Natural Language Processing (NLP), this study found that negative sentiment dominated public responses to the crisis (12,947 tweets), while positive sentiment was relatively small (1,136 tweets). Frequently occurring words, such as "gempa," "korban," "bencana," indicate that the public's attention is mostly focused on the consequences of the disaster, while the words "kecepatan," "informasi," and "pengolahan data" reflect high expectations of a fast and accurate response. The results of this study confirm that effective crisis communication should be data-driven and integrate big data analysis to understand communication trends during a crisis. The use of NLP enables real-time sentiment monitoring, misinformation detection, and the development of more adaptive communication strategies. From a crisis management perspective, this approach can help government agencies to reduce public uncertainty, quell panic, and rebuild public trust. The study also recommends the need for stricter regulations related to disaster communication to address disinformation that has the potential to worsen the situation. With technology-based approaches and data analysis, crisis communication strategies can be more effective in dealing with the dynamics of public opinion during emergency situations.