Analisis Sentimen Ulasan Pengunjung Terhadap Tempat Wisata Religi Walisongo Menggunakan Metode Supervised Learning

Abstract

The rapid development of information technology has made visitors of Walisongo Tourisms able to provide reviews through the map application that often used by public, Google Maps. With the reviews given by other users, it can help others who want to visit the place and become a source of data material for the local government in optimizing the services, facilities, and infrastructure of Walisongo Tourisms. To help simplify the process of processing reviews, a sentiment analysis process is needed which is assisted by an algorithm that will be in charge of classifying these reviews. This study aims to carry out a sentiment analysis of visitor ratings of the Walisongo Tourisms using supervised learning, such as Decision Trees (CART), K-Nearest Neighbor, Multinomial Naïve Bayes, and Support Vector Machine kernel RBF. The result of this study indicate that the model who has the highest accuracy value is SVM-RBF model classification with accuracy value 87,12% and f1-score value of negative class 89%, 68% for neutral class, and 91% for positive class.