User-generated content across social media platforms is playing an increasingly important role in tourism. Understanding tourists' experiences and opinions about tourism destinations has led to numerous opportunities to provide tourism providers with greater insights. Identifying sentiments, detecting topics of interest, and exploring loyalty behaviors from user-generated content can provide valuable direction for managerial decisions. This paper presents a novel and inclusive approach that uses different analytical techniques such as sentiment analysis and topic modeling to extract sentiments and topics of interest from tourists’ conversational data on TripAdvisor from 2002 to 2019. Sentiment analysis revealed that some touristic locations in Jasper National Park are outperforming others in terms of sentiment scores, despite the fact that tourists less frequently reflect on their experiences at these places on social media. Such higher rankings suggest that average sentiment score can be a more informative measure than simple TripAdvisor rankings. This paper also explores destination loyalty statements using a keyword clustering approach. Previous destination loyalty literature was used to develop a keyword list that was applied to search for expressions of loyalty in online reviews. The robustness of loyalty clusters and optimal number of clusters was also assessed prior to final analysis. Four leading loyalty-focused categories of destination offerings were observed: glaciers, waterfalls, lakes and islands, and hiking and trails. Prioritization of visitor experience enhancements relating to these loyalty-inducing destination components are discussed.

跨社交媒体平台的用户生成的内容在旅游业中发挥着越来越重要的作用。理解游客对旅游目的地的经验和看法，已经带来了许多机会，可以为旅游提供商提供更深刻的见解。识别情绪，检测感兴趣的主题以及从用户生成的内容中探索忠诚度行为可以为管理决策提供有价值的指导。本文提出了一种新颖且包容的方法，该方法使用了不同的分析技术，例如情感分析和主题建模，以从2002年至2019年TripAdvisor的游客对话数据中提取感兴趣的情感和主题。情感分析显示，贾斯珀国家公园的一些旅游地点在情感分数方面胜过其他人，尽管事实上，游客在社交媒体上较少反映他们在这些地方的经历的事实。如此高的排名表明，与简单的TripAdvisor排名相比，平均情绪评分可以提供更多信息。本文还探索了使用关键字聚类方法的目的地忠诚度声明。以前的目的地忠诚度文献用于开发关键字列表，该列表用于搜索在线评论中的忠诚度表达。在最终分析之前，还评估了忠诚度群集的健壮性和群集的最佳数量。观察到有四种主要的以忠诚度为重点的目的地产品类别：冰川，瀑布，湖泊和岛屿以及远足和徒步旅行。讨论了与这些引起忠诚度的目的地组成部分相关的访客体验增强的优先级。