CD Skripsi

Analisis sentimen pengguna terhadap transisi aplikasi BNI mobile banking ke wondr by BNI menggunakan indobert

---

Digital transformation has driven Bank Negara Indonesia (BNI) to introduce the wondr by BNI application as a replacement for BNI Mobile Banking. This study aimed to analyze user sentiment regarding this transition through user reviews collected from the Google Play Store. A total of 35,644 review data were classified into three sentiment categories: positive, neutral, and negative. To address class imbalance, an oversampling process was conducted using the random oversampling technique, increasing the total number of data to 40,776. The analysis was conducted using the IndoBERTbase-p1 model with training hyperparameters consisting of 5 epochs, a learning rate of 1 × 10⁻⁵, and a batch size of 32. The model was trained using three different data splitting ratios for training, validation, and testing: 60:20:20, which resulted in a final accuracy of 85%; 70:15:15, with an accuracy of 88%; and 80:10:10, which achieved the highest accuracy of 90%. The 80:10:10 ratio was selected for evaluation due to its superior accuracy. Evaluation results show the best performance in the negative class with an F1-score of 93%, followed by the neutral class at 88%, and the positive class at 87%. This study demonstrated that IndoBERT is effective for sentiment analysis of application reviews.
Keywords: Google Play Store, IndoBERT, Oversampling, Sentiment Analysis, wondr by BNI

---

Ketersediaan

Informasi Detail

Judul Seri

-

No. Panggil

2103112032

Penerbit

Pekanbaru : Universitas Riau FMIPA Sistem Informasi., 2025

Deskripsi Fisik

-

Bahasa

Indonesia

ISBN/ISSN

-

Klasifikasi

2103112032

Tipe Isi

-

Tipe Media

-

Tipe Pembawa

-

Edisi

-

Subjek

SISTEM INFORMASI

Info Detail Spesifik

-

Pernyataan Tanggungjawab

Mutia

Versi lain/terkait

Lampiran Berkas

Komentar

---

Anda harus masuk sebelum memberikan komentar