Pemodelan Topik Pada Ulasan Pengguna Accsess By KAI Menggunakan Metode Latent Dirichlet Allocation (LDA) Dan Latent Semantic Analysis (LSA)

Ajie, Amira Adilla Paramadina (2025) Pemodelan Topik Pada Ulasan Pengguna Accsess By KAI Menggunakan Metode Latent Dirichlet Allocation (LDA) Dan Latent Semantic Analysis (LSA). Other thesis, Institut Teknologi Sepuluh Nopember.

	Text 2043211110-Undergraduate_Thesis.pdf Restricted to Repository staff only Download (5MB) | Request a copy
	Spreadsheet Jadwal Magang Siswa SMKN 5.xlsx Download (9kB)
	Spreadsheet Jadwal Magang Siswa SMKN 5.xlsx Download (9kB)
	Spreadsheet Jadwal Magang Siswa SMKN 5.xlsx Download (9kB)
	Image Brosure.jpeg Download (156kB)
	Image Brosure.jpeg Download (156kB)

Abstract

Kemajuan teknologi digital mendorong transformasi layanan transportasi, termasuk peluncuran aplikasi Access by KAI oleh PT. Kereta Api Indonesia, yang dirancang untuk mempermudah transaksi tiket secara daring. Meskipun memiliki potensi besar, aplikasi ini sering menghadapi masalah teknis, seperti sulitnya login dan lambatnya proses pemesanan, yang tercermin dalam ulasan negatif di Google Play Store dengan peringkat rendah 2,3 dari 5. Penelitian ini menganalisis kualitas layanan aplikasi dengan menggabungkan analisis sentimen menggunakan kamus lexicon-based InSet dan pemodelan topik melalui metode Latent Dirichlet Allocation (LDA) serta Latent Semantic Analysis (LSA). Dari total 8.393 ulasan, sebanyak 67,9% bersentimen positif, didominasi apresiasi terhadap fitur pemesanan tiket dan top-up saldo, sementara 32,1% bersentimen negatif dengan fokus pada masalah teknis. Pemodelan topik mengidentifikasi tema utama dari masing-masing sentimen, di mana LSA memberikan hasil lebih koheren dibandingkan LDA, dengan coherence score tertinggi mencapai 0,626 pada sentimen positif dan 0,486 pada sentimen negatif. Analisis ini memberikan wawasan penting tentang area yang perlu perbaikan, seperti peningkatan stabilitas aplikasi dan optimasi performa sistem. Penelitian ini juga merekomendasikan respons lebih aktif terhadap ulasan pengguna untuk memperbaiki pengalaman pelanggan dan meningkatkan citra PT. KAI sebagai penyedia layanan transportasi digital. Hasil penelitian ini tidak hanya relevan untuk meningkatkan kualitas layanan Access by KAI, tetapi juga memberikan kontribusi strategis untuk pengembangan sistem layanan transportasi berbasis digital di Indonesia.
==================================================================================================================================
Advances in digital technology are driving the transformation of transportation services, including PT Kereta Api Indonesia's launch of the Access by KAI app, designed to simplify online ticket transactions. Despite its great potential, the app often faces technical issues, such as difficult login and slow booking process, which are reflected in negative reviews on Google Play Store with a low rating of 2.3 out of 5. This research analyzes the app's service quality by combining sentiment analysis using InSet's lexicon-based dictionary and topic modeling through Latent Dirichlet Allocation (LDA) and Latent Semantic Analysis (LSA) methods. From a total of 8,393 reviews, 67.9% were positive, dominated by appreciation for the ticket booking and balance top-up features, while 32.1% were negative with a focus on technical issues. Topic modeling identified the main themes of each sentiment, where LSA provided more coherent results than LDA, with the highest coherence score reaching 0.626 for positive sentiment and 0.486 for negative sentiment. This analysis provides important insights into areas for improvement, such as improving app stability and optimizing system performance. The research also recommends more active response to user reviews to improve customer experience and enhance PT KAI's image as a digital transportation service provider. The results of this study are not only relevant for improving the quality of Access by KAI services, but also make a strategic contribution to the development of digital-based transportation service systems in Indonesia.

Item Type:	Thesis (Other)
Uncontrolled Keywords:	Sentiment Analysis, Topic Modelling, LDA, LSA, Access by KAI, Analisis Sentimen, Pemodelan Topik
Subjects:	H Social Sciences > HA Statistics H Social Sciences > HD Industries. Land use. Labor > HD108 Classification (Theory. Method. Relation to other subjects ) Q Science > Q Science (General) > Q325.5 Machine learning. Support vector machines.
Divisions:	Faculty of Vocational > 49501-Business Statistics
Depositing User:	Amira Adilla Paramadina Ajie
Date Deposited:	12 Jun 2025 05:20
Last Modified:	12 Jun 2025 05:20
URI:	http://repository.its.ac.id/id/eprint/118726

Actions (login required)

View Item