Skripsi

PENERAPAN DATA MINING UNTUK KNOWLEDGE DISCOVERY ULASAN APLIKASI BLU BY BCA MENGGUNAKAN SVM, CNN, DAN INDOBERT

---

The ever-evolving digital banking services require a comprehensive understanding of user perceptions. This study applies a knowledge discovery approach to analyze the sentiment of user reviews of the Blu by BCA application by comparing three classification algorithms, namely Support Vector Machine (SVM), Convolutional Neural Network (CNN), and IndoBERT. Data was obtained through the process of scraping user reviews on the Google Play Store platform, and after filtering and preprocessing, 22,923 valid reviews were obtained. Model performance was evaluated using accuracy, precision, recall, F1-score, and AUC-ROC metrics. The results showed that IndoBERT had the best and most stable performance with accuracy, precision, recall, and F1-score values of around 0.87, and the highest AUC-ROC value of 0.89, followed by CNN with accuracy in the range of 0.83–0.86 and SVM in the range of 0.83–0.85. Additionally, WordCloud analysis was used as part of the knowledge discovery process to identify dominant words in each sentiment category, revealing aspects of the service that users frequently appreciate as well as issues that often arise in reviews. Overall, this study produced knowledge discovery in the form of sentiment patterns and characteristics of user perceptions of the Blu by BCA application based on public review data. Keywords: Sentiment Analysis, Knowledge Discovery, Text Mining, IndoBERT, Blu by BCA

---

Availability

Detail Information

Series Title

-

Call Number

T1947422026

Publisher

Indralaya : Prodi Sistem Informasi, Fakultas Ilmu Komputer Universitas Sriwijaya., 2026

Collation

xvi, 50 hlm.; ilus.; 29 cm

Language

Indonesia

ISBN/ISSN

-

Classification

006.312 07

Content Type

Text

Media Type

unmediated

Carrier Type

-

Edition

-

Subject(s)

Data mining
Prodi Sistem Informasi

Specific Detail Info

-

Statement of Responsibility

KA

Other version/related

File Attachment

Comments

---

You must be logged in to post a comment