Article Per Year (5 Year)
p-Index From 2021 - 2026
0.23
P-Index
This Author published in this journals
All Journal RABIT: Jurnal Teknologi dan Sistem Informasi Univrab
Amanda Diyas Setiyoadi
Universitas Muria Kudus
Computer Science & IT Engineering

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

Found 1 Documents
Search

 1 
KOMPARASI KINERJA MACHINE LEARNING TEROPTIMASI SMOTE DAN PSO PADA KLASIFIKASI SENTIMEN ULASAN ROBLOX Amanda Diyas Setiyoadi; Yudie Irawan; Soni Adiyono
Rabit : Jurnal Teknologi dan Sistem Informasi Univrab Vol 11 No 2 (2026): Juli
Publisher : LPPM Universitas Abdurrab

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.36341/rabit.v11i2.8034

Abstract

The emergence of digital platforms like Roblox has led to an increase in the number of user reviews on the Google Play Store. These reviews contain important information regarding public perception, satisfaction levels, and user complaints about the app. However, the large volume of reviews and the unstructured nature of the text make manual analysis inefficient. Therefore, an automated solution in the form of machine learning-based sentiment classification is needed. This study was conducted to evaluate and compare the effectiveness of three machine learning algorithms, namely Logistic Regression, Support Vector Machine (SVM), and Random Forest, in classifying Roblox app review sentiment into three categories: positive, neutral, and negative. The research data consisted of 10,000 reviews collected through a crawling process from the Google Play Store. Synthetic Minority Oversampling Technique (SMOTE) was applied to address class imbalance, while Particle Swarm Optimization (PSO) was used to optimize model parameters. Experimental results show that Random Forest combined with SMOTE achieved the highest performance with an accuracy of 0.7219, a precision of 0.7241, a recall of 0.7219, an F1-score of 0.7228, and an AUC of 0.778. However, the accuracy of 72.19% is still a limitation for direct practical application, so further improvements are needed. This study also developed a Streamlit-based dashboard to monitor sentiment classification results in real-time. Based on these findings, the combination of Random Forest and SMOTE can be considered quite effective, although it still has limitations in the level of model accuracy.
Co-Authors Soni Adiyono Yudie Irawan