Universitas Syiah Kuala | ELECTRONIC THESES AND DISSERTATION

Electronic Theses and Dissertation

Universitas Syiah Kuala

    THESES
Reni Wahyuni, ANALISIS KUALITAS LAYANAN PERBANKAN DIGITAL MENGGUNAKAN ALGORITMA SUPPORT VECTOR MACHINE DAN FREQUENT PATTERN GROWTH PADA APLIKASI BSI MOBILE. Banda Aceh Fakultas Teknik (S2),2025

Transaksi perbankan yang semakin meningkat telah mendorong adopsi aplikasi mobile banking, termasuk bsi mobile oleh bank syariah indonesia. pada kuartal pertama tahun 2024, bsi mobile mencatat 118,51 juta transaksi, tumbuh sebesar 37,17% dibandingkan tahun sebelumnya, serta peningkatan jumlah pengguna sebesar 29,35%. seiring pertumbuhan ini, ulasan pengguna juga meningkat pesat sehingga diperlukan analisis sentimen untuk mengevaluasi kualitas layanan. penelitian ini bertujuan menganalisis sentimen pengguna bsi mobile dan mengidentifikasi faktor-faktor penyebab keluhan guna memberikan rekomendasi perbaikan layanan. sebanyak 13.793 ulasan pengguna dari google play store yang dikumpulkan antara 4 januari 2021 hingga 30 agustus 2024 dianalisis. klasifikasi sentimen dilakukan menggunakan algoritma support vector machine, dengan kategori ulasan positif, negatif, dan netral. pola asosiasi kata diekstraksi menggunakan algoritma frequent pattern growth untuk mengidentifikasi istilah yang sering muncul dalam sentimen negatif. hasil menunjukkan 60,3% ulasan bersentimen positif, 34,1% negatif, dan 5,7% netral. kata-kata seperti “error,” “update,” “failed,” dan “login” mendominasi ulasan negatif. analisis akar penyebab melalui diagram fishbone menunjukkan faktor manusia (layanan pelanggan tidak responsif), proses (gagal login dan aktivasi), dan produk (crash dan gangguan teknis). rekomendasi perbaikan mencakup peningkatan uji coba aplikasi, akurasi face recognition, pemberitahuan pemeliharaan, layanan pelanggan yang responsif, serta panduan interaktif dalam aplikasi untuk meningkatkan kualitas layanan digital banking secara menyeluruh. kata kunci : bsi mobile, kualitas layanan, analisis sentimen, support vector, frequent pattern-growth



Abstract

The increasing volume of banking transactions has driven the adoption of mobile banking applications, including BSI Mobile by Bank Syariah Indonesia. In the first quarter of 2024, BSI Mobile recorded 118.51 million transactions, representing a 37.17% year-on-year growth, along with a 29.35% increase in users. Along with this growth, user reviews have also risen significantly, highlighting the need for sentiment analysis to evaluate service quality. This study aims to analyze user sentiment toward BSI Mobile and identify the underlying causes of complaints in order to provide service improvement recommendations. A total of 13,793 user reviews from the Google Play Store, collected between January 4, 2021, and August 30, 2024, were analyzed. Sentiment classification was performed using the Support Vector Machine algorithm, categorizing reviews into positive, negative, and neutral. Word association patterns were extracted using the Frequent Pattern Growth algorithm to identify frequently appearing terms in negative reviews. The results showed that 60.3% of the reviews were positive, 34.1% negative, and 5.7% neutral. Words such as “error,” “update,” “failed,” and “login” were dominant in negative reviews. Root cause analysis using a fishbone diagram revealed issues related to people (unresponsive customer service), processes (login and activation failures), and products (app crashes and technical disruptions). The proposed service improvement recommendations include enhancing app testing procedures, improving face recognition accuracy, providing scheduled maintenance notifications, strengthening customer support, and offering interactive in-app guidance to comprehensively improve the quality of digital banking services. Keywords: BSI Mobile, Service Quality, Sentiment Analysis, Support Vector Machine, Frequent Pattern-Growth



    SERVICES DESK