<![CDATA[Purpose - This study aims to analyze customer sentiments in hotel reviews to gain insights into guest experiences, service quality, and overall satisfaction. The goal is to demonstrate how sentiment analysis using natural language processing (NLP) can support decision-making, service enhancement, and strategic management in the tourism and hospitality industry. Methodology/Design/Approach - The research employs two main approaches for sentiment classification—lexicon-based (VADER) and machine learning-based (BERT). Textual hotel reviews are processed and categorized into positive, negative, and neutral sentiments. The performance of both methods is compared to evaluate accuracy and effectiveness in interpreting customer feedback. Findings - The results reveal that key sentiment drivers include service quality, cleanliness, and value for money. The BERT model outperforms the lexicon-based VADER method in classification accuracy, demonstrating its superior ability to understand contextual nuances in customer reviews. The study confirms that advanced NLP models can provide deeper and more reliable insights for reputation management and marketing strategies. Originality/Value - This paper contributes to the growing field of artificial intelligence applications in tourism by showcasing how NLP-based sentiment analysis can transform qualitative feedback into actionable intelligence. It highlights the potential of AI in improving customer experience analytics and suggests future research directions in multilingual sentiment analysis and real-time monitoring for dynamic decision support, benefiting hotels, travel agencies, and policymakers.]]>
