Skripsi

KNOWLEDGE DISCOVERY MENGGUNAKAN ANALISIS SENTIMEN DAN PEMODELAN TOPIK PADA ULASAN PENGGUNA BCA MOBILE DAN MYBCA

---

The rapid growth of mobile banking in Indonesia underscores the need for secure and innovative digital financial services. However, studies that compare user experiences across multiple applications within the same bank remain limited. To address this research gap, we analyze BCA’s two primary applications, BCA Mobile and myBCA, which serve millions of users in Indonesia. Data collected from Google Play reviews were analyzed with sentiment analysis and topic modeling to deliver actionable recommendations for service refinement and trust enhancement. Reviews were automatically labeled with IndoBERT and evaluated using seven classification models. The GRU classifier achieved the best performance, with 89.70% accuracy, 0.89 precision, 0.90 recall, 0.89 F1-score, and an AUC-ROC of 0.95. For topic modeling, BERTopic achieved the highest coherence score relative to LDA, extracting eight topics from negative reviews of BCA Mobile and myBCA. The resulting topics indicate differences in user complaints between the applications, login failures, balance deductions, and missing features are more dominant in BCA Mobile, while post-update errors, login problems, and face-verification issues are more prevalent in myBCA. The findings provide service-improvement recommendations by combining sentiment classification and topic modeling to drive enhancements aligned with SDG 9

---

Availability

Detail Information

Series Title

-

Call Number

T1892802025

Publisher

Indralaya : Prodi Sistem Informasi, Fakultas Ilmu Komputer Universitas Sriwijaya., 2025

Collation

xvii, 97 hlm.; ilus.; tab, 29 cm

Language

Indonesia

ISBN/ISSN

-

Classification

006.350 7

Content Type

Text

Media Type

-

Carrier Type

-

Edition

-

Subject(s)

Prodi Sistem Informasi
Analisis Sentimen--Proses Data Mobile

Specific Detail Info

-

Statement of Responsibility

MI

Other version/related

File Attachment

Comments

---

You must be logged in to post a comment