ANALISIS SENTIMEN PERSEPSI MASYARAKAT TERHADAP EFEKTIVITAS EARLY WARNING SYSTEM BENCANA BANJIR DI KOTA MALANG MENGGUNAKAN PEMBOBOTAN TF-IDF

Abstract

This research aims to analyze the sentiment of public perception of the effectiveness of the Early Warning System (EWS) for  lood disasters in Malang City. Sentiment Analysis will be carried out using TF-IDF weighting on the initial data. Initial data is  obtained from personal interviews with people who have experienced flooding around the EWS area, obtained as many as 13  respondents. The data is then processed by preprocessing, labeling, and TF-IDF weighting. This research was conducted at 6  points of the EWS tool with purposive sampling based on the area that is likely to hear the warning sound of the EWS tool  with the condition that the respondent has experienced flooding and knows the existence of EWS in the area. From the  sentiment analysis results, it was found that the community has two main focuses in the effectiveness of Flood EWS, namely the sound aspect and the timeliness aspect. However, the community still has a negative perception of the effectiveness of flood EWS in Malang City. Suggestions for future research are that the number of respondents should be increased because the respondents in this study were quite small.