以網際網路為基礎的社群媒體興盛時代，社群上常常可見由使用者生成的評論 來表達各式各樣的體驗過程或感受。旅遊評論更是社群評論中的大宗，旅客常會透過線 上評論平台表達他們對於旅行的整體體驗，線上評論不僅為旅客提供一個與業者溝通的 管道，更是一種藉由大眾的意見共同評估特定飯店品質的方法。因此，許多人在搜尋和 選擇飯店的過程也會受到線上評論和評分的影響。尤其是國際旅客，更是僅能基於可見 的線上評論作為評估與下訂單的基礎。線上評論亦容易受到當下整體經濟環境因素的影 響，而為使用者帶來不一樣的看法。因此，本研究將透過使用文字探勘技術分析台灣飯 店評論以了解不同類型的國際旅客對於顧客滿意度及反饋以及影響滿意度的重要因素。 透過主題模型方法得知所有顧客評論中的句子所隱含的主題，此主題可視為影響顧客滿 意度之關鍵因素;並透過情感分析技術來標記每個評論中的句子之各自的情感面向，最 後將兩個結果結合，採用 KANO 模型其中的三個類別，必要條件品質、一維品質、魅力 品質，將針對不同旅遊類型之旅客影響顧客滿意度之因素更近一步的分類。 ;In the age of social media and the internet, user-generated comments are commonly found on various platforms, expressing experiences and emotions. Travel reviews play a significant role in this community feedback, as travelers often use online review platforms to share their overall travel experiences. These reviews not only allow travelers to communicate with service providers but also serve as a collective assessment of hotel quality based on public opinions. Many people rely on online reviews and ratings when searching for and selecting hotels, especially international travelers who heavily depend on visible online reviews for evaluation and booking decisions. However, online reviews can be influenced by the current economic environment, leading to diverse perspectives among users. Therefore, this study aims to analyze hotel reviews in Taiwan using text mining techniques to understand the satisfaction levels, feedback, and important factors influencing international travelers. By employing topic modeling, we can identify implicit themes conveyed in customer reviews that are considered key factors affecting customer satisfaction. Additionally, sentiment analysis techniques allow for the identification of sentiment aspects within each review. By combining these outcomes and adopting three categories from the KANO model—basic quality, performance quality, and excitement quality—we can further classify the factors influencing customer satisfaction among different types of travelers.
