Analisis Sentimen Ulasan Aplikasi ChatGPT Pada Play Store Menggunakan Pendekatan Data Mining Dengan Evaluasi Kinerja Model

Salamah, Rabbiatul Rafi'ah (2025) Analisis Sentimen Ulasan Aplikasi ChatGPT Pada Play Store Menggunakan Pendekatan Data Mining Dengan Evaluasi Kinerja Model. Skripsi thesis, UIN Imam Bonjol Padang.

	Text (Cover - Daftar Isi) Cover - Daftar Isi.pdf - Published Version Download (1MB)
	Text (BAB 1) Rabbiatul Rafi'ah Salamah. NIM 2117020035. BAB 1.pdf - Published Version Download (170kB)
	Text (BAB 3) Rabbiatul Rafi'ah Salamah. NIM 2117020035. BAB 3.pdf - Published Version Download (253kB)
	Text (BAB 5) Rabbiatul Rafi'ah Salamah. NIM 2117020035. BAB 5 - Daftar Pustaka.pdf - Published Version Download (285kB)
	Text (Full Text) Rabbiatul Rafi'ah Salamah. NIM 2117020035. Full Text.pdf - Published Version Restricted to Repository staff only Download (3MB)

Abstract

Artificial intelligence (AI) has become a significant innovation in the digital age, capable of processing data and providing automated solutions across various fields. One widely used AI implementation is ChatGPT, a text-based application developed by OpenAI. ChatGPT has received millions of reviews on the Google Play Store, making it crucial to conduct a comprehensive analysis of user perceptions. This study aims to analyze user sentiment toward the ChatGPT application and compare the performance of four classification algorithms: Naive Bayes, Support Vector Machine (SVM), Decision Tree, and Random Forest. A total of 2,289 reviews were collected via web scraping from the Google Play Store between January and March 2025. The research followed the CRISP-DM stages, from data understanding to model evaluation. The data was analyzed through preprocessing and then represented using the TF-IDF method. Sentiment was automatically labeled based on rating scores. The results showed that the majority of reviews were positive (82.45%), followed by negative (11.90%) and neutral (5.65%). The data was divided with an 80:20 ratio for model training and testing. Evaluation was performed using a confusion matrix with accuracy, precision, recall, and F1-score metrics. The SVM algorithm showed the best performance with an accuracy of 98.32%, followed by Random Forest (96.55%), Naive Bayes (90.71%), and Decision Tree (88.05%). Overall, ChatGPT received positive reviews from users, particularly in terms of learning, communication, and personal interaction. However, some reviews highlighted technical glitches and unstable features. Therefore, developers need to make continuous improvements to ensure an optimal user experience.

Actions (login required)

View Item