Analisis sentimen ulasan aplikasi traveloka pada app marketplace menggunakan DistilBERT

Kevin, Kevin (2024) Analisis sentimen ulasan aplikasi traveloka pada app marketplace menggunakan DistilBERT. Bachelor thesis, Universitas Pelita Harapan.

Preview	Text (Title) Title.pdf Available under License Creative Commons Attribution Non-commercial Share Alike. Download (33kB) | Preview
Preview	Text (Title) Title.pdf Available under License Creative Commons Attribution Non-commercial Share Alike. Download (57kB) | Preview
Preview	Text (Abstract) Abstract.pdf Available under License Creative Commons Attribution Non-commercial Share Alike. Download (404kB) | Preview
Preview	Text (ToC) ToC.pdf Available under License Creative Commons Attribution Non-commercial Share Alike. Download (308kB) | Preview
Preview	Text (Chapter1) Chapter1.pdf Available under License Creative Commons Attribution Non-commercial Share Alike. Download (286kB) | Preview
	Text (Chapter2) Chapter2.pdf Restricted to Registered users only Available under License Creative Commons Attribution Non-commercial Share Alike. Download (679kB)
	Text (Chapter3) Chapter3.pdf Restricted to Registered users only Available under License Creative Commons Attribution Non-commercial Share Alike. Download (738kB)
	Text (Chapter4) Chapter4.pdf Restricted to Registered users only Available under License Creative Commons Attribution Non-commercial Share Alike. Download (618kB)
	Text (Chapter5) Chapter5.pdf Restricted to Registered users only Available under License Creative Commons Attribution Non-commercial Share Alike. Download (264kB)
Preview	Text (Bibliography) Blbliography.pdf Available under License Creative Commons Attribution Non-commercial Share Alike. Download (273kB) | Preview
	Text (Appendices) Appendices.pdf Restricted to Repository staff only Available under License Creative Commons Attribution Non-commercial Share Alike. Download (2MB)

Abstract

Traveloka merupakan sebuah platform terkenal di Indonesia yang menyediakan layanan pemesanan tiket pesawat, kereta api, hotel, dan lain-lain. Di era digital saat ini, aplikasi perjalanan seperti Traveloka telah menjadi aplikasi bagi banyak individu untuk merencanakan dan memesan perjalanan. Sebagai platform perjalanan terkemuka di Indonesia, Traveloka mendapatkan banyak umpan balik pengguna dalam bentuk ulasan, yang mencakup aspek seperti kualitas layanan dan pengalaman pengguna. Penelitian ini menganalisa sentimen ulasan pelanggan pada aplikasi Traveloka menggunakan model DistilBERT dan berfokus pada pengembangan model analisis sentimen ulasan aplikasi Traveloka di app marketplace untuk memahami kepuasan pelanggan dalam menggunakan aplikasi Traveloka dan mengidentifikasi potensi peningkatan fitur layanan Traveloka. Metodologi yang dilakukan termasuk pengumpulan data dari App Store dan Google Play, pra-pemrosesan data, pemisahan data, pemodelan pembelajaran mesin dengan BERT, evaluasi performa model, dan analisis performa model. Jumlah dataset setelah proses tersebut adalah sebanyak 7.149 data. Evaluasi performa model yang dilakukan menunjukkan model analisis yang dikembangkan memiliki akurasi sebesar 70,21% menggunakan confusion matrix. Hal ini menunjukkan model mampu mengklasifikasikan sentiment ulasan dengan cukup baik. / Traveloka is a well-known platform in Indonesia that provides services for booking flights, trains, hotels, and more. In the current digital era, travel apps like Traveloka have become essential tools for individuals to plan and book their trips. As a leading travel platform in Indonesia, Traveloka receives extensive user feedback in the form of reviews, covering aspects such as service quality and user experience. This research analyzes customer review sentiments on the Traveloka app using the DistilBERT model. It focuses on developing a sentiment analysis model for Traveloka app reviews in app marketplaces to understand customer satisfaction in using the Traveloka app and identify potential enhancements to Traveloka's service features. The methodology includes data collection from the App Store and Google Play, data preprocessing, data splitting, machine learning modeling with BERT, model performance evaluation, and performance analysis. After these processes, the dataset consists of 7,149 entries. The model performance evaluation carried out shows that the analytical model developed has an accuracy of 70.21% using the confusion matrix. This shows that the model is able to classify review sentiment quite well.

Actions (login required)

View Item