<![CDATA[This research aims to improve sentiment analysis of reviews related to Garden by the Bay, a prominent tourist destination in Singapore, by leveraging the CRISP-DM methodology and Synthetic Minority Over-sampling Technique (SMOTE). The study employs a comprehensive approach, integrating CRISP-DM phases to systematically collect, clean, and analyze data from online reviews. The dataset comprises a substantial number of reviews, reflecting diverse visitor experiences. Using SMOTE, class imbalance issues within the dataset are addressed, leading to enhanced performance of sentiment analysis algorithms. The evaluation of Decision Tree (DT) and Support Vector Machine (SVM) algorithms, both with and without SMOTE, reveals significant improvements in accuracy, precision, recall, and F-measure metrics when SMOTE is applied. These findings underscore the efficacy of SMOTE in optimizing sentiment analysis algorithms for the Garden by the Bay dataset, thereby facilitating a deeper understanding of visitor sentiments and experiences, which inform strategies for enhancing the tourism experience at Garden by the Bay.]]>
Copyrights © 2024
