Analisis Sentimen Terhadap Ulasan Aplikasi Gojek Menggunakan Naive Bayes Classifier dengan BoW

Abstract

In this digital era, technology pervades every aspect of daily life, revolutionizing industries and interactions. Within this landscape, online-based transportation services have emerged as transformative solutions, notably exemplified by Gojek in Indonesia. This study delves into sentiment analysis of Gojek reviews using Multinomial Naive Bayes and Bag-of-Words extraction, aiming to gauge user perceptions and responses. Leveraging a dataset of 9.996 App reviews, the research undertakes comprehensive preprocessing, including case folding, filtering, tokenization, stopword removal, and stemming, followed by sentiment labeling. By employing Bag-of-Words feature extraction, textual data is converted into numerical vectors, enabling the application of the Multinomial Naive Bayes classification model. Evaluation metrics, derived from a confusion matrix, reveal an accuracy rate of 86.29%, with precision, recall, and F1-Score values of 86.94%, 86.41%, and 86.26% respectively. This study underscores the efficacy of the adapted Multinomial Naive Bayes model with Bag-of-Words feature extraction in discerning user sentiments towards Gojek, offering valuable insights for enhancing service applications in the digital realm.