<![CDATA[This study addresses the research problem of understanding public sentiment towards tourism-themed content on YouTube, with a specific focus on "A Visual Journey of Wonderful Indonesia." The primary aim is to explore how viewers perceive and depict Indonesia as a tourism destination through their comments on YouTube videos. Employing the Cross-Industry Standard Process for Data Mining (CRISP-DM) methodology, sentence analysis is conducted using the Support Vector Machine (SVM) algorithm with the Synthetic Minority Over-sampling Technique (SMOTE) to classify sentiments within a dataset of YouTube comments as positive, negative, or neutral. The analysis of frequently used words in the comments provides valuable insights into Indonesia's perception, revealing positive sentiments reflected in terms such as "beautiful," "wonderful," and "amazing," emphasizing the country's aesthetic appeal. Notably, terms like "orang" and "Indonesian" indicate appreciation for Indonesia's rich cultural heritage and its people. These findings highlight the pivotal role of destination branding efforts in shaping positive perceptions and emotions toward Indonesia. The results indicate the efficacy of the SVM-SMOTE model, achieving high accuracy (84.26%), precision (100.00%), recall (68.51%), f-measure (81.25%), and AUC (0.996) in accurately classifying sentiment patterns within analyzed YouTube content. This offers practical implications for destination managers and marketers. Conversely, the SVM algorithm without SMOTE demonstrates impressive accuracy, precision, and recall scores of 97.08%, but its AUC value of 0.607 suggests potential challenges in discriminating between positive and negative sentiment instances. These findings provide valuable insights into the role of digital media platforms in shaping destination perceptions and offer practical implications for destination marketers and managers]]>
Copyrights © 2024
