Analisis Sentimen Aplikasi Identitas Kependudukan Digital pada Ulasan Google Play Store menggunakan IndoBERT dan Bi-LSTM

Ananto, Amadeus Terra (2026) Analisis Sentimen Aplikasi Identitas Kependudukan Digital pada Ulasan Google Play Store menggunakan IndoBERT dan Bi-LSTM. Other thesis, Institut Teknologi Sepuluh Nopember.

Text 5026211075-Undergraduate_Thesis.pdf Restricted to Repository staff only Download (7MB) | Request a copy

Abstract

Transformasi digital telah mengubah berbagai proses aktivitas sehari-hari menjadi jauh lebih cepat dan efisien. Pemerintah Indonesia dalam upaya meningkatkan efektivitas dan efisiensi proses birokrasinya, turut aktif melakukan transformasi digital dalam pelayanan publik, salah satunya dalam bidang administratif. Aplikasi Identitas Kependudukan Digital (IKD) adalah salah satu inisiatif pemerintah dalam simplifikasi administrasi Indonesia yang diawali dengan identitas digital. Aplikasi IKD disediakan pada Google Play Store untuk diunduh masyarakat Indonesia. Tujuan Tugas Akhir ini adalah untuk menguji performa IKD melalui analisis opini masyarakat dengan metode analisis sentimen pada ulasan pengguna pada Google Play Store. Sebanyak 3000 ulasan diambil secara acak dan dilabeli mejadi tiga kategori : positif, netral, dan negatif. Data yang sudah dilabeli kemudian di bersihkan melalui pra-pemrosesan agar siap dianalisis. Alat yang digunakan untuk analisis sentimen berupa IndoBERT dengan lapisan Bi-LSTM yang akan dibandingkan dengan lapisan alternatif RNN, GRU, dan LSTM. Performa model akan dievaluasi melalui f1-score dan akurasi dan didapatkan paling optimal 91% dan 90% pada model IndoBERT Bi-LSTM.
===================================================================================================================================
Digital transformation has changed various daily activity processes to be much faster and more efficient. The Indonesian government in an effort to improve the effectiveness and efficiency of public services, is actively carrying out digital transformation in public services, one of which is in the administrative field. The Identitas Kependudukan Digital (IKD) application is one of the government's initiatives in simplifying Indonesian administration, commencing with digital identity. The IKD application is provided on the Google Play Store for the Indonesian people to download. The purpose of this study is to test the performance of IKD through public opinion analysis with the sentiment analysis method on user reviews on the Google Play Store. A total of 3000 reviews were taken randomly and labeled into three categories: positive, neutral, and negative. The labeled data is then cleaned through pre-processing so that it is ready to be analyzed. The tool used for sentiment analysis is IndoBERT with a Bi-LSTM layer which will be compared with alternative layers of RNN, GRU, and LSTM. Model performance will be evaluated through F1-score and accuracy. To get the most relevant sentiment, a special sentiment analysis of reviews for each version of the application is also carried out. The sentiment of the latest version is then compared with the previous version to get complete sentiment towards IKD

Item Type:	Thesis (Other)
Uncontrolled Keywords:	Identitas Kependudukan Digital, IKD, sentiment, IndoBERT, BI-LSTM, Identitas Kependudukan Digital, IKD, sentimen, IndoBERT, Bi-LSTM.
Subjects:	P Language and Literature > P Philology. Linguistics > P325 Semantics. 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) T Technology > T Technology (General) > T385 Visualization--Technique T Technology > T Technology (General) > T57.5 Data Processing
Divisions:	Faculty of Intelligent Electrical and Informatics Technology (ELECTICS) > Information System > 57201-(S1) Undergraduate Thesis
Depositing User:	Amadeus Terra Ananto
Date Deposited:	02 Feb 2026 04:09
Last Modified:	02 Feb 2026 04:09
URI:	http://repository.its.ac.id/id/eprint/131496

Actions (login required)

View Item