03-Analisis Sentimen Penggunaan Aplikasi ChatGPT Pada Lingkungan Akademik-Izmi Panji Gumilang

Ulasan Aplikasi ChatGPT

Abstract

Generative Pre-trained Transformer (ChatGPT) is an artificial intelligence technology developed by OpenAI, designed to understand natural language and interact with humans contextually. Its high popularity has brought diverse experiences and opinions from users, particularly regarding its use in academic environments. Sentiment analysis research is necessary to understand user perceptions of the ChatGPT application. User review data were collected from the Google Play Store through web scraping, totaling 5,000 reviews from July 2023 to July 2025. The Naïve Bayes algorithm was used to classify the review texts using a Fine-Grained Sentiment Analysis approach into three sentiment categories: positive (72.3%), negative (18.5%), and neutral (9.2%). Testing using the Confusion Matrix with an 80% training and 20% testing data split showed that the model achieved an average accuracy of 81%, precision of 82%, recall of 81%, and F1-score of 81%. The results of this study are expected to provide insights into the extent to which ChatGPT is considered helpful or effective, especially in contexts related to academic environments.