Application of Text Mining Techniques in Hotel Service Quality Diagnosis

• Main Authors: HSIAO, YU-TING, 蕭伃婷
• Other Authors: HSIAO, YU-HSIANG
• Format: Others
• Language: zh-TW
• Published: 2018
• Online Access: http://ndltd.ncl.edu.tw/handle/rm97hb

碩士 === 國立臺北大學 === 企業管理學系 === 106 === With the popularity of the Internet and mobile devices, the platform for user-generated content (UGC) such as online reviews, community websites, and blogs has rapidly grown. Companies can listen to consumers' opinions on their products and services through online reviews on their official websites or third-party websites. It can be seen that online reviews play an important role in electronic word-of-mouth. This study uses the reviews on the Booking.com site as a sample source and uses the Latent semantic analysis of text mining to summarize and identify the customer satisfaction and dissatisfaction service characteristics from the textual reviews. The Kano model is used to classify two-dimensional quality features to understand customer needs and serve as directions for hotel operations and services. Further, a decision tree is used to establish the classification diagnosis rules. The rules draw out the key service characteristics of good hotels and bad hotels. Last, a quality diagnostic model for service features was developed to diagnose the quality of service for individual hotel, improve the missing service features, and make recommendations for the diagnosis. The research uses Business Intelligence to help the hotel analyze its own merits from the existing reviews and provide hotel manager with appropriate measures or plans for new managements to enhance their competitiveness.