Indah Aida Sapitri, - (2023) PENGKLASIFIKASIAN SENTIMEN ULASAN APLIKASI WHATSAPP PADA GOOGLE PLAY STORE MENGGUNAKAN SUPPORT VECTOR MACHINE. PENGKLASIFIKASIAN SENTIMEN ULASAN APLIKASI WHATSAPP PADA GOOGLE PLAY STORE MENGGUNAKAN SUPPORT VECTOR MACHINE, 6 (1). ISSN 2621-3079

Text (JURNAL) INDAH-1.pdf - Published Version Restricted to Repository staff only Download (1MB)

Abstract

Abstract The Google Play Store is a platform commonly used to download applications, one of which is WhatsApp. The Google Play Store also provides a feature so that users can provide reviews in the form of comments containing both positive and negative points of view. The method used in this study is the Support Vector Machine method. The purpose of this study is to apply the SVM method in classifying sentiments and knowing the accuracy test of the method. This study uses 1000 reviews collected from the scrapping process and uses two comparisons, namely 90:10 and 80:20. A comparison of 90:10 produces an accuracy of 82%, while a comparison of 80:20 produces an accuracy of 81%, a comparison of 90:10 produces a precision value of 58%, 35% recall, f1-score 44% for the negative class and a precision value of 85 %, 94% recall, 89% f1-score for the positive class, while the 80:20 ratio produces 62% precision, 34% recall, 44% f1-score for the negative class and 84% precision value, 94% recall, f1- score 89% for the positive class. The best parameter pairs are at C=1.0 and γ = 1.0 with an accuracy of 68% at a ratio of 90:10, while at a comparison of 80:20 the best parameter pairs are at C=0.9 and γ=0.7 with an accuracy of 67%. Keywords: Support Vector Machine, Classification, Sentiment Analysis, Play Store, WhatsApp Application.

Actions (login required)

View Item