ANALISIS SENTIMEN OBJEK WISATA DI GOOGLE MAPS MENGGUNAKAN METODE DECISION TREE

Abstract

Tourism is a popular choice for travelers exploring a city. Typically, people rely on Google Reviews to assess the quality of tourist spots by checking ratings and reading reviews. However, discrepancies between reviews and ratings can occur. In this case study, we focus on Taman Wisata Taman Laut Batam in Batam city. Using the Decision tree method, we conducted sentiment analysis. The results revealed more positive sentiments (200) than negative sentiments (141). The F1-score accuracy was 79.59%, with a margin of error of ±8.78%. Precision for positive predictions was 71.00%, while negative predictions had 77.49% precision. Recall values were 75.53% for positive predictions and 62.09% for negative predictions. Overall, positive sentiment predictions achieved an F1-score of 87.07%, while negative predictions scored 74.73%.