Nurul Asma As-shidiq, Donny Maulana, M. Zubair Abdurrohman

Sentiment analysis of public opinion on the non-cash food assistance (BPNT) program on platform x using naive bayes

Introduction

Sentiment analysis of public opinion on the non-cash food assistance (bpnt) program on platform x using naive bayes. Analyze public opinion on Indonesia's BPNT non-cash food assistance program on Platform X using Naive Bayes. Reveals public sentiment for social assistance policy evaluation.

Abstract

The Non-Cash Food Assistance (BPNT) program is one of the Indonesian government's social assistance policies that frequently receives public attention and discussion on social media. This study aims to analyze public sentiment toward the implementation of the BPNT program based on comments posted on Platform X. The dataset consisted of 1,758 Indonesian-language comments collected through web crawling between October 2025 and January 2026. The collected data were processed through several preprocessing stages, including case folding, cleaning, tokenization, normalization, stopword removal, and stemming. Furthermore, TF-IDF was applied to transform textual data into numerical features, and sentiment classification was performed using the Multinomial Naive Bayes algorithm. The dataset was divided into training and testing data using an 80:20 ratio. The results showed that neutral sentiment dominated public discussions with 51.48%, followed by positive sentiment with 33.90% and negative sentiment with 14.62%. Performance evaluation using a Confusion Matrix achieved an accuracy of 79.545%. These findings indicate that the Naive Bayes approach can effectively classify public sentiment regarding the BPNT program and provide useful insights for evaluating social assistance policies.

Full Text

Comments

You need to be logged in to post a comment.