The Role of Sentiment and Toxicity in Digital Narratives Surrounding Sulawesi's Wildlife Tourism: A Content Analysis for Enhancing Conservation Strategies

This research explores the intersection of wildlife tourism and digital narratives, focusing on Sulawesi's endemic species. Utilizing the Digital Content Reviews and Analysis framework, the study combines content analysis, sentiment classification, and toxicity assessment to uncover critical insights. The findings highlight digital narratives' significant role in shaping public perceptions and behaviors toward conservation and ecotourism. Through systematic content analysis, themes such as biodiversity, conservation, and local community involvement emerged as effectively communicated, resonating with audiences and promoting sustainable tourism practices. The framework's structured approach enabled a thorough examination of digital content's impact on wildlife tourism narratives, identifying critical patterns and themes. The study also employed advanced machine learning techniques, specifically the SVM algorithm enhanced by SMOTE, which achieved a sentiment classification accuracy of 88.76% ± 3.11% and an AUC of 0.977, demonstrating its effectiveness. However, toxicity assessment revealed that while most interactions were civil, specific posts contained significant levels of toxicity, with a peak score of 0.64912, underscoring the need for better moderation and engagement strategies. The research emphasizes integrating conservation-focused elements into digital narratives to foster positive engagement and support for wildlife preservation. The study provides practical recommendations for enhancing the positive influence of digital narratives on conservation and sustainable tourism, offering a foundation for future initiatives to optimize digital communication strategies in ecotourism