Implementation of Support Vector Machine for Sentiment Analysis on Blibli App Reviews on Play Store

Santika, Bima (2025) Implementation of Support Vector Machine for Sentiment Analysis on Blibli App Reviews on Play Store. Implementation of Support Vector Machine for Sentiment Analysis on Blibli App Reviews on Play Store, Vol.13 (No. 2). pp. 133-138. ISSN 2716-3989

	Text (Artikel) 133666-109573-1-PB (1).pdf - Published Version Restricted to Repository staff only Download (846kB) \| Request a copy
	Text (Turnitin) Bima Santika Artikel ASC1.pdf - Published Version Restricted to Repository staff only Download (3MB) \| Request a copy

Official URL: http://ejournal.unp.ac.id/index.php/voteknika

Abstract

Blibli is one of the leading e-commerce platforms in Indonesia with a 4.7-star rating on the Google Play Store. Although it has many positive reviews, automatic sentiment analysis is needed to understand user perceptions objectively and efficiently. This study aims to classify the sentiment of user reviews of the Blibli application using the Support Vector Machine (SVM) algorithm. A total of 2,000 recent reviews were collected through web scraping and underwent preprocessing processes including case folding, cleaning, normalization, filtering, stemming, and tokenization. To address class imbalance, the SMOTE method was applied so that positive and negative sentiment data became balanced, with 1,413 reviews each. The data were then divided into three training and testing ratio scenarios: 90:10, 80:20, and 70:30. Text transformation was carried out using the TF-IDF method. Experiments were conducted by comparing four SVM kernels: Linear, RBF, Polynomial, and Sigmoid. The best results were obtained in the 90:10 scenario using the Linear kernel, achieving an accuracy of 96.82%, precision of 96.93%, recall of 96.82%, and F1-score of 96.82%. These findings indicate that SVM, particularly with the Linear kernel, is highly effective and balanced in classifying user review sentiments. The results of this study are expected to contribute to application developers in improving service quality based on user feedback obtained automatically and in a structured manner.

Item Type:	Article
Additional Information:	Artikel ini sudah terbit pada Jurnal Vocational Teknik Elektronika dan Informatika Vol. 13, No. 2, 2025
Uncontrolled Keywords:	Keywords— Sentiment analysis, Blibli, Play Store, SVM, SMOTE
Subjects:	Universitas Muhammadiyah Bengkulu > 02-Fakultas Teknik > 55201-(S1) Teknik Informatika 02-Fakultas Teknik > 55201-(S1) Teknik Informatika
Divisions:	02-Fakultas Teknik > 55201-(S1) Teknik Informatika Subjek Terkait > 02-Fakultas Teknik > 55201-(S1) Teknik Informatika
Depositing User:	Mrs Atika Ledia Anggraini
Date Deposited:	20 May 2026 02:12
Last Modified:	20 May 2026 02:12
URI:	http://repository.umb.ac.id/id/eprint/2348

Actions (login required)

View Item