Are customer reviews just reviews? Hotel forecasting using sentiment analysis

Doris Chenguang Wu, Shiteng Zhong, Richard T.R. Qiu, Ji Wu

Research output: Journal article publicationJournal articleAcademic researchpeer-review

16 Citations (Scopus)


Despite the important role of the sentiment of customer reviews in the tourist decision-making process, it has received limited attention from the field of tourism demand forecasting. This study aims to examine the sentiment information of customer reviews and explore its potential in enhancing hotel demand forecast. Empirically, four Macau luxury hotels are selected and their customer reviews are crawled from two popular online platforms. A deep learning method, the Long Short-Term Memory model, is used to extract sentiment information from consumer reviews. Three sentiment indices, namely, bullish index, average index and variance index, are constructed and examined. The effectiveness of these sentiment indices is further evaluated by the autoregressive integrated moving average with exogenous variables model. Empirical results indicate that the inclusion of the sentiment indices helps with improving forecast accuracy. The findings of this study further emphasise the importance of textual content to hotel practitioners in terms of strategy formulation, revenue management and competitiveness enhancement.

Original languageEnglish
Pages (from-to)795-816
Number of pages22
JournalTourism Economics
Issue number3
Publication statusPublished - May 2022
Externally publishedYes


  • big data
  • hotel demand forecasting
  • long short-term memory model
  • online review
  • sentiment indices

ASJC Scopus subject areas

  • Geography, Planning and Development
  • Tourism, Leisure and Hospitality Management


Dive into the research topics of 'Are customer reviews just reviews? Hotel forecasting using sentiment analysis'. Together they form a unique fingerprint.

Cite this