A New Instrument for Measuring Customers’ Perceptions of Service Warmth: A Big Data and Machine Learning Approach

Abstract

The hotel industry is placing increasing emphasis on customers’ perception of service warmth. However, the current methods for measuring customers’ perception of service warmth are limited to sample surveys. Thus, a two-stage quantitative approach was adopted in this study to develop a novel measuring instrument. By leveraging big data analysis and machine learning techniques, a weighted lexicon of 95 words was identified, which can be utilized to assess customers’ perception of enthusiasm toward P2P accommodation services. The effectiveness of this new measurement tool was tested through two methods: a five-fold cross-validation approach and multiple regression with controlled variables. The instrument developed in this study enables comparable measurement results of perceived service enthusiasm across different accommodation units. Furthermore, this research contributes to the knowledge of measurement instrument development in the era of big data. The practical implications of using the weighted word list are also discussed.