Analisis Sentimen Berbasis Aspek pada Ulasan Mobile Banking Menggunakan Support Vector Machine dan Long Short-Term Memory

Diana, Annissa Hasna (2026) Analisis Sentimen Berbasis Aspek pada Ulasan Mobile Banking Menggunakan Support Vector Machine dan Long Short-Term Memory. Other thesis, Institut Teknologi Sepuluh Nopember.

Text 5003221057_Annissa Hasna Diana_Laporan.pdf - Accepted Version Restricted to Repository staff only Download (7MB) | Request a copy

Abstract

Perkembangan pesat teknologi informasi telah mendorong peningkatan penggunaan layanan mobile banking, yang memungkinkan nasabah melakukan transaksi keuangan secara cepat, fleksibel, dan aman. Penelitian ini bertujuan untuk menganalisis persepsi pengguna terhadap aplikasi mobile banking di Indonesia, yaitu BCA Mobile, BRImo, Livin’ by Mandiri, dan Wondr by BNI, melalui pendekatan Aspect-Based Sentiment Analysis (ABSA) dan visualisasi perceptual map berdasarkan ulasan pengguna pada Google Play Store. Proses ABSA dilakukan melalui dua tahap, yaitu klasifikasi aspek berdasarkan hasil LDA dan klasifikasi sentimen ke dalam kategori positif dan negatif. Kedua tahap dilakukan menggunakan metode Support Vector Machine (SVM) dan Long Short-Term Memory (LSTM). Hasil penelitian menunjukkan bahwa keempat aplikasi memiliki pola serupa, yaitu aspek kemudahan dan kepuasan layanan didominasi sentimen positif, sedangkan aspek keuangan dan pembayaran, akses dan verifikasi akun, serta performa dan stabilitas aplikasi, didominasi sentimen negatif, yang menunjukkan masih adanya kendala pada aspek layanan tersebut. Berdasarkan hasil perbandingan metode, SVM menunjukkan performa yang lebih baik dibandingkan LSTM, dengan akurasi sebesar 87,15% untuk klasifikasi aspek dan 95,72% untuk klasifikasi sentimen. Selain itu, hasil perceptual map menunjukkan bahwa BRImo memiliki skor sentimen tertinggi pada aspek kemudahan dan kepuasan layanan sebesar 0,7393. Pada aspek keuangan dan pembayaran, BRImo memperoleh skor sentimen tertinggi sebesar -0,6669. Pada aspek akses dan verifikasi akun, BRImo juga memperoleh skor sentimen tertinggi sebesar -0,8804. Sementara itu, pada aspek performa dan stabilitas aplikasi, Livin’ by Mandiri memperoleh skor sentimen tertinggi sebesar -0,8864. Namun demikian, semakin tinggi skor sentimen mengindikasikan bahwa tingkat permasalahan pada aspek tersebut relatif lebih rendah dibandingkan aplikasi mobile banking lainnya.
=====================================================================================================================================
The rapid development of information technology has driven the increasing adoption of mobile banking services, enabling users to conduct financial transactions quickly, flexibly, and securely. This study aims to analyze user perceptions of mobile banking applications in Indonesia, namely BCA Mobile, BRImo, Livin’ by Mandiri, and Wondr by BNI, using an Aspect-Based Sentiment Analysis (ABSA) approach and perceptual map visualization based on user reviews from the Google Play Store. The ABSA process is conducted in two stages: aspect classification using LDA results and sentiment classification into positive and negative categories. Both stages are implemented using Support Vector Machine (SVM) and Long Short-Term Memory (LSTM) methods. The results show similar patterns across applications, where the ease of use and service satisfaction aspect is dominated by positive sentiment, while financial and payment, account access and verification, application performance and stability aspects are dominated by negative sentiment, indicating existing issues in these areas. The comparison of the methods shows that SVM outperforms LSTM in both aspect classification and sentiment classification, achieving an accuracy of 87.15% for aspect classification and 95.72% for sentiment classification. Furthermore, the perceptual map shows that BRImo achieves the highest sentiment score of 0.7393 for the ease of use and service satisfaction aspect, -0.6669 for the financial and payment aspect, and -0.8804 for the account access and verification aspect. Meanwhile, Livin' by Mandiri records the highest sentiment score of -0.8864 for the application performance and stability aspect. Higher sentiment scores indicate fewer user-perceived issues than other mobile banking applications.

Item Type:	Thesis (Other)
Uncontrolled Keywords:	Mobile Banking, Analisis Sentimen, LSTM, SVM, Perceptual Map, Mobile Banking, Sentiment Analysis, LSTM, SVM, Perceptual Map
Subjects:	Q Science > Q Science (General) > Q325.5 Machine learning. Support vector machines. Q Science > QA Mathematics > QA336 Artificial Intelligence Q Science > QA Mathematics > QA76.87 Neural networks (Computer Science)
Divisions:	Faculty of Mathematics and Science > Statistics > 49201-(S1) Undergraduate Thesis
Depositing User:	Annissa Hasna Diana
Date Deposited:	04 Aug 2026 04:37
Last Modified:	04 Aug 2026 04:37
URI:	http://repository.its.ac.id/id/eprint/140533

Actions (login required)

View Item