Sebuah Analisis Sentimen Terhadap Transformasi Program Makan Siang Gratis menjadi Makan Bergizi Gratis di YouTube Menggunakan Logistik Regression

Studi Kasus Komentar YouTube Menggunakan Regresi Logistik

Abstract

Abstract- This study delves into sentiment analysis concerning the transformation from a simple free lunch program to a nutritious free lunch initiative, based on user comments on YouTube. Employing logistic regression analysis with varied test sizes of 10%, 20%, and 30%, the data was rigorously examined to discern public opinion. Initial data collection through crawling from May to July 2024 yielded 5,390 comments, reflecting a significant engagement primarily in June. Post-processing categorized the sentiments as 1,682 negative, 634 neutral, and 958 positives. The classification accuracy varied: 85.32% (10% test size), 82.90% (20% test size), and 81.41% (30% test size), indicating the logistic regression model's effectiveness despite increasing data volume. This research underscores the utility of sentiment analysis in evaluating public reception and potential areas for improvement in governmental nutrition programs.