Detecting the Meaning of the Allowance Policy of the House of Representatives of the Republic of Indonesia Using Naïve Bayes Classifier and Lexicon Algorithms

Abstract

Social media serves as a source of information that can be used to gauge public opinion regarding government policies one topic frequently discussed by the public is the policy regarding allowances for members of the Indonesian House of Representatives. The objective of this study is to examine public sentiment regarding these policies using the Naïve Bayes Classifier and Lexicon Sentiment methods. The research approach applied is CRISP-DM (Cross Industry Standard Process for Data Mining), which encompasses the stages of business understanding, data understanding, data preparation, modeling, evaluation, and implementation. Data was collected from the social media platform X (Twitter) via scraping and processed through preprocessing steps and TF-IDF weighting. The findings of this study indicate that the Naïve Bayes Classifier method achieved an accuracy of 74%, while the Lexicon Sentiment method helped in understanding the emotional nuances present in the text. The combination of these two methods produces a more comprehensive and relevant sentiment analysis compared to using only one method alone. This study demonstrates that the combination of statistical and lexicon-based approaches is highly useful in analyzing sentiment regarding the opinions of the Indonesian-speaking public.