Article Per Year (5 Year)
p-Index From 2021 - 2026
0.23
P-Index
This Author published in this journals
All Journal Jurnal Ilmiah Sistem Informasi dan Ilmu Komputer
Rivana Delfalizanty
Unknown Affiliation
Industrial & Manufacturing Engineering

Published : 1 Documents Claim Missing Document
Claim Missing Document
Check
Articles

Found 1 Documents
Search

 1 
Analisis Sentimen Persepsi Masyarakat pada Media Sosial X Menggunakan Metode K-Means, SVM, dan Teknik Smote : Studi Kasus: Program Makan Bergizi Gratis Rivana Delfalizanty; Yulhendri Yulhendri
Jurnal ilmiah Sistem Informasi dan Ilmu Komputer Vol. 6 No. 1 (2026): Maret: Jurnal ilmiah Sistem Informasi dan Ilmu Komputer
Publisher : Lembaga Pengembangan Kinerja Dosen

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.55606/juisik.v6i1.2071

Abstract

The Free Nutritious Meal (MBG) Program has emerged as a major public policy issue frequently discussed on social media platform X (Twitter). Public discourse surrounding this program reflects diverse perspectives, including support, criticism, and broader debates on national nutrition policies. This study seeks to identify the dominant topics within these conversations through K-Means clustering and to classify sentiments using the Support Vector Machine (SVM) algorithm, enhanced with SMOTE to address class imbalance. A total of 3,053 tweets were collected through crawling. The clustering results revealed three main themes: (1) nutrition fulfillment and food access for priority groups, highlighted by keywords such as nutrition, milk, fish, toddlers, schools, and pesantren; (2) political dynamics and program legitimacy, represented by terms like stunting, support, president, Prabowo, society, and national; and (3) state financial concerns, reflected in words including tax, funds, state budget, cost, people, and expenditure. In sentiment analysis, SVM without SMOTE produced accuracies of 65.69%, 67.76%, and 68.12% under 90:10, 80:20, and 70:30 split ratios. After applying SMOTE, accuracy increased to 77.94%, 76.38%, and 75.06%, with an F1-Score of 78%. These results confirm that K-Means is effective in identifying discussion topics, while SMOTE enhances SVM performance in sentiment classification.
Co-Authors Yulhendri Yulhendri