Analisis Sentimen Ulasan Wisata Alun-Alun Brebes pada Google Maps Menggunakan Support Vector Machine

Abstract

The rapid development of information technology has encouraged the use of digital platforms as media for sharing opinions, including tourism reviews on Google Maps. Alun-Alun Brebes, as one of the most frequently visited public spaces, has generated thousands of reviews with diverse textual characteristics, making manual analysis inefficient and impractical. This study aims to analyze visitor sentiment toward Alun-Alun Brebes by applying a text mining approach using the Support Vector Machine algorithm. The dataset consists of 1,000 Google Maps reviews, including 327 reviews manually labeled as positive and negative sentiments and 673 unlabeled reviews. The research stages include data collection, text preprocessing, feature extraction using the Term Frequency–Inverse Document Frequency (TF-IDF) method, Support Vector Machine model training and testing, and automatic labeling of unlabeled data. Model performance was evaluated using a confusion matrix with accuracy, precision, recall, and F1-score metrics based on the manually labeled data. The results show that the Support Vector Machine model with a linear kernel achieved an accuracy of 100% with an F1-score of 1.00, indicating excellent sentiment classification performance. Furthermore, word cloud visualization reveals that positive sentiment is dominated by aspects related to comfort and facilities, while negative sentiment is associated with cleanliness, crowd density, and environmental management. These findings provide data-driven insights into key aspects that should be maintained and improved in managing Alun-Alun Brebes as a public space.