Analisis Sentimen Berbasis Aspek Pada Ulasan Pengguna Aplikasi POLRI Presisi

Bramhatchi, Farhan Bramhatchi (2026) Analisis Sentimen Berbasis Aspek Pada Ulasan Pengguna Aplikasi POLRI Presisi. Other thesis, Institut Teknologi Sepuluh Nopember.

Text 2043221064_Undergraduated_Thesis.pdf - Accepted Version Restricted to Repository staff only Download (7MB) | Request a copy

Abstract

Aplikasi POLRI Presisi merupakan inovasi layanan publik digital yang dikembangkan oleh Kepolisian Republik Indonesia untuk mempermudah masyarakat dalam mengakses layanan kepolisian secara daring. Aplikasi POLRI Presisi masih memperoleh rating dengan nilai 3,3 dari nilai 5 dengan 34,8 ribu ulasan secara keseluruhan dan masih ditemukan berbagai keluhan pengguna terkait kinerja dan kestabilan sistem aplikasi. Penelitian ini menggunakan 18.874 ulasan pengguna aplikasi POLRI Presisi versi 2.0.27 yang dikumpulkan melalui proses scrapping pada hari Jum'at, 8 Agustus 2025 dengan tujuan mengidentifikasi aspek layanan yang menjadi sumber keluhan serta kecenderungan sentimen pengguna menggunakan pendekatan Aspect-Based Sentiment Analysis (ABSA). Identifikasi aspek dilakukan menggunakan metode Latent Dirichlet Allocation (LDA), sedangkan analisis sentimen dilakukan dengan model IndoBERT. Setelah melalui tahapan text preprocessing, jumlah data yang digunakan dalam analisis berkurang menjadi 10.395 ulasan. Hasil identifikasi dan pelabelan topik menghasilkan dua topik utama, yaitu “Kinerja Aplikasi dan Pengalaman Pengguna” serta “Proses Registrasi dan Upload Dokumen”. Analisis sentimen menunjukkan bahwa topik “Kinerja Aplikasi dan Pengalaman Pengguna” didominasi sentimen positif, sedangkan topik “Proses Registrasi dan Upload Dokumen” didominasi sentimen negatif yang menunjukkan tingginya tingkat keluhan pengguna. Evaluasi menggunakan Net Reputation Score (NRS) berdasarkan sentimen maupun rating pengguna menunjukkan bahwa topik “Proses Registrasi dan Upload Dokumen” memiliki tingkat ketidakpuasan tertinggi dan menjadi prioritas utama perbaikan, sedangkan pada topik “Kinerja Aplikasi dan Pengalaman Pengguna” secara umum memperoleh penilaian yang relatif baik. Hasil penelitian ini diharapkan dapat menjadi dasar pengambilan keputusan strategis dalam meningkatkan kualitas aplikasi dan layanan publik digital kepolisian.
=================================================================================================================================
The POLRI Presisi application is a digital public service innovation developed by the Indonesian National Police to facilitate public access to police services through online platforms. Despite obtaining an overall rating of 3.3 out of 5 with 34.8 thousand user reviews, various user complaints related to application performance and system stability are still evident. This study utilizes 18,874 user reviews of the POLRI Presisi application version 2.0.27, collected through a scraping process on Friday, August 8, 2025, with the aim of identifying service aspects that serve as the main sources of user complaints as well as analyzing user sentiment tendencies using an Aspect-Based Sentiment Analysis (ABSA) approach. Aspect identification was conducted using the Latent Dirichlet Allocation (LDA) method, while sentiment analysis was performed using the IndoBERT model. After undergoing text preprocessing, the number of reviews used in the analysis was reduced to 10,395. The topic identification and labeling process resulted in two main topics, namely “Application Performance and User Experience” and “Registration Process and Document Upload.” Sentiment analysis results indicate that the “Application Performance and User Experience” topic is dominated by positive sentiment, whereas the “Registration Process and Document Upload” topic is dominated by negative sentiment, reflecting a high level of user complaints. Further evaluation using the Net Reputation Score (NRS) based on both sentiment classification and user ratings shows that the “Registration Process and Document Upload” topic has the highest level of dissatisfaction and should therefore be considered the top priority for improvement, while the “Application Performance and User Experience” topic generally receives relatively favorable evaluations. The findings of this study are expected to serve as a basis for strategic decision-making in improving the quality of the application and digital public services provided by the police.

Item Type:	Thesis (Other)
Uncontrolled Keywords:	ABSA, LDA, IndoBERT, Net Reputation Score, ABSA, LDA, IndoBERT, Net Reputation Score.
Subjects:	Q Science Q Science > Q Science (General) Q Science > Q Science (General) > Q325.5 Machine learning. Support vector machines.
Divisions:	Faculty of Vocational > 49501-Business Statistics
Depositing User:	Farhan Bramhatchi
Date Deposited:	28 Jul 2026 08:35
Last Modified:	28 Jul 2026 08:35
URI:	http://repository.its.ac.id/id/eprint/136529

Actions (login required)

View Item