Online review analytics for hotel quality at macro and micro levels

This study was to develop a methodology of online review analytics for hotel quality management at macro and micro levels. The macro level was for understanding the service properties critical to quality and future development. The micro level was for personalized quality diagnosis for individual hotels.

Textual reviews of superior hotels were studied using latent semantic analysis and Kano model to understand what service properties customers concern and expect. Taguchi's quality engineering was applied to establish a quality reference base using superior hotels for evaluating other hotels in the same semantic space. A decision tree algorithm was then used to identify the properties critical to quality discrimination, and the decision rules were used to diagnose individual hotels.

The service properties concerned by customers for superior hotels were identified. The market positioning and value of each property to customers were clarified. For individual hotels, the deficiencies of service properties were diagnosed. With reference to market positioning, deficient properties of priority in improvement and the strategies for enhancing competitiveness were suggested.

The proposed methodology demonstrated the potential value that review analysis can achieve a new and deeper understanding of customer voices and transform it into more specific business operation requirements. The research and application gap that most previous studies only stayed at the macro-level analytics was filled. Moreover, this study effectively applied the diagnostic techniques derived from quality engineering to online review analytics.