Article Per Year (5 Year)
p-Index From 2020 - 2025
0.23
P-Index
This Author published in this journals
All Journal Sistemasi: Jurnal Sistem Informasi
Subandono, Anelta Tirta Putri
Unknown Affiliation
Computer Science & IT Electrical & Electronics Engineering

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

Found 1 Documents
Search
Journal : Sistemasi: Jurnal Sistem Informasi

 1 
Optimizing Feature Selection in Sentiment Analysis of Bank Saqu: A Comparative Study of SVM and Random Forest using Information Gain and Chi-Square Subandono, Anelta Tirta Putri; Ariatmanto, Dhani
Sistemasi: Jurnal Sistem Informasi Vol 14, No 3 (2025): Sistemasi: Jurnal Sistem Informasi
Publisher : Program Studi Sistem Informasi Fakultas Teknik dan Ilmu Komputer

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.32520/stmsi.v14i3.5106

Abstract

The selection of an optimal feature selection method is a crucial factor in improving the accuracy and efficiency of text classification models. Irrelevant features can degrade model performance, increase computational complexity, and lead to overfitting. Although various feature selection techniques have been employed in sentiment analysis, systematic studies comparing the effectiveness of Information Gain and Chi-Square in enhancing classification performance remain limited. This study aims to evaluate and optimize the impact of different feature selection methods on the performance of Support Vector Machine (SVM) and Random Forest (RF) in sentiment analysis. Experiments were conducted using eight testing schemes, including models without feature selection, with Information Gain, Chi-Square, and a combination of both. The results showed that SVM with Chi-Square achieved the highest accuracy at 93%, while Random Forest with Chi-Square achieved the best performance at 91%. These findings indicate that Chi-Square is more effective than Information Gain in improving accuracy, and that SVM outperforms Random Forest in text classification tasks. In conclusion, selecting the appropriate feature selection method significantly contributes to enhancing the accuracy of text classification models. This research can serve as a reference for optimizing feature selection techniques in the development of more accurate and efficient machine learning-based systems.
Co-Authors DHANI ARIATMANTO