Article Per Year (5 Year)
p-Index From 2021 - 2026
0.23
P-Index
This Author published in this journals
All Journal Aviation Electronics, Information Technology, Telecommunications, Electricals, Controls (AVITEC)
Furqon, Ihtiar Nur
Unknown Affiliation
Aerospace Engineering Computer Science & IT Electrical & Electronics Engineering Engineering

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

Found 1 Documents
Search

 1 
The Role of VADER and SentiWordNet Labeling in Naïve Bayes Accuracy for Sentiment Analysis of Rice Price Increases Furqon, Ihtiar Nur; Soyusiawaty, Dewi
Aviation Electronics, Information Technology, Telecommunications, Electricals, and Controls (AVITEC) Vol 7, No 1 (2025): February
Publisher : Institut Teknologi Dirgantara Adisutjipto

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.28989/avitec.v7i1.2806

Abstract

The surge in rice prices in Indonesia in 2024 is a critical issue affecting social welfare and national food security, particularly amid rising rice imports. This study evaluates public sentiment on Twitter using the Naïve Bayes method and compares the effectiveness of two automated labeling methods, VADER and SentiWordNet, in improving sentiment analysis accuracy. The research is significant due to the limited literature on automated labeling comparisons, especially in food price crises. The methodology includes data collection, preprocessing, translation, sentiment labeling using VADER and SentiWordNet, TF-IDF feature extraction, Naïve Bayes classification, and performance evaluation across different data split ratios: 60% training and 40% testing, 70% training and 30% testing, 80% training and 20% testing, and 90% training and 10% testing. Results show that VADER excels in detecting positive sentiments, achieving 74.42% accuracy at a 90:10 split but struggles with negative sentiment identification, with a highest F1-score of 56.58%. SentiWordNet performs better for positive sentiment detection, reaching 77.86% accuracy and 96.22% recall at an 80:20 split but yields a low F1-score of 32.15% for negative sentiments. In conclusion, VADER is suitable for balanced sentiment detection, while SentiWordNet is more effective for identifying positive sentiments.
Co-Authors Dewi Soyusiawaty