Mining Unstructured Data in Social Media for Natural Disaster Management in Indonesia

This paper proposed a model system for unstructured mining data in social media for natural disaster management in Indonesia. The model system of natural disaster management will be tested using real data where the application will be run from the stage crawl social media, tokenization, filtering, stemming, similarity measure, and Name Entity Recognizer so as to ascertain whether the software is built is in conformity with the rules of data collection events natural disasters that can be reliable. The proposed model system of natural disaster management can help the Indonesian government to calculate the impact of floods, landslides, and tornados that it could decide to focus fixes in the correct fields. If the government has made improvements by the mapping of disaster impact it will automatically proclamation of floods, landslides, and tornados in social media and news websites will decrease, and the value graph will change impacts linkages so that the government can focus subsequent repair.