Analisis Sentimen Sosial Media Twitter Terkait Tabungan Perumahan Rakyat (Tapera) Dengan Naïve Bayes

The advancement of technology has brought significant changes in various aspects of life, including how people interact and express their opinions on social media. This study aims to analyze the sentiment of conversations on Twitter regarding the government-initiated Housing Savings Program (TAPERA). Using the Naïve Bayes classification method, data was collected through the Google Collab platform and processed using the tweet-harvest library, successfully gathering 1,800 tweets that met the search criteria. The data preprocessing stage included cleaning and manual sentiment labeling to obtain accurate training data. The analysis results show that out of 720 tested tweets, 473 tweets had negative sentiment (65.69%) and 247 tweets had positive sentiment (34.31%). The Naïve Bayes model used in this study demonstrated good performance with an accuracy of 84.17%, precision for the negative class at 88.71%, recall for the negative class at 88.60%, precision for the positive class at 78.54%, and recall for the positive class at 76.08%. The conclusion of this study is that the Naïve Bayes method is effective in classifying public sentiment towards the TAPERA program, providing important insights for relevant parties in evaluating and improving the program's effectiveness.