Implementation of Text Processing Techniques on Citizen Opinions Regarding Floods in Surabaya

Abstract

An analysis of residents opinions on flooding in Surabaya is essential to identify their perceptions and aspirations towards this disaster. This can facilitate the making of appropriate flood mitigation policies. The current problem is that it is difficult to retrieve data from X in the form of both users and tweets automatically. So that in some studies that use tweet data becomes less efficient in the data collection process. This research confirmed the use of messages to determine highly impactful disaster zones and showed how tweets can be used to identify oscillations in disaster intensity over time. The topic of this research is to apply the Data Crawling method to obtain datasets on social media X. Then the next method is text preprocessing using Wordcloud, Matplotlib, (NTLK) Natural Language Toolkit, and Sastrawi libraries. In natural language processing, the data to be extracted includes unstructured or “arbitrary” data. In normal dialect preparing (NLP), the information to be extracted includes unstructured or "self-assertive" information. For future purposes (assumption examination, subject modeling, etc.), such information must be changed over into organized data. The discoveries of the think about can help the organization in comprehending the necessities and inclinations of the people with respect to surges. This article demonstrates how Artificial Intelegence may be applied to text data analysis in order to provide insightful findings. the outcomes of this research can help the government in making more effective policies to overcome flooding in Surabaya