Article Per Year (5 Year)
p-Index From 2021 - 2026
0.408
P-Index
This Author published in this journals
All Journal JOURNAL OF SCIENCE AND SOCIAL RESEARCH Journal of Technology and Computer (JOTECHCOM)
Sinurat, Satria
Unknown Affiliation
Computer Science & IT Economics, Econometrics & Finance Education Social Sciences

Published : 2 Documents Claim Missing Document
Claim Missing Document
Check
Articles

Found 1 Documents
Search
Journal : Journal of Technology and Computer (JOTECHCOM)

 1 
Comparative Machine Learning Analysis for Sentiment Classification of Sumatra Disaster 2025 Alfarizi, Nauval; Lydia, Prima; Novelan, Muhammad Syahputra; Putra, Adi; Sinurat, Satria
Journal of Technology and Computer Vol. 3 No. 1 (2026): February 2026 - Journal of Technology and Computer
Publisher : PT. Technology Laboratories Indonesia (TechnoLabs)

Show Abstract | Download Original | Original Source | Check in Google Scholar

Abstract

Indonesia is highly vulnerable to natural disasters due to its geological position, resulting in extensive disaster-related news coverage that shapes public sentiment. This study presents a comparative machine learning analysis for sentiment classification of online news related to natural disasters in Sumatra during December 2025. The dataset was collected through web scraping from two major Indonesian news portals, like CNN Indonesia and Detik, and categorized into three sentiment classes: negative, neutral, and positive. Sentiment classification was conducted using Naive Bayes, Support Vector Machine (SVM), and k-Nearest Neighbors (KNN) algorithms. The results demonstrate that Naive Bayes achieved accuracy values of 0.57 on the CNN Indonesia dataset and 0.61 on the Detik dataset. However, its performance was highly biased toward the dominant negative class, as indicated by low macro-average F1-scores of (0.24) and (0.39). In contrast, SVM showed the most balanced performance by reducing class bias, achieving accuracies of (0.68) and (0.67) with macro-average F1-scores of (0.51) and (0.59), respectively. KNN demonstrated moderate performance, with accuracy values of 0.60 and 0.59, but remained less effective than SVM in handling minority sentiment classes.
Co-Authors Adi Putra Alfarizi, Nauval Lydia, Prima Muhammad Amin Muhammad Syahputra Novelan