Forecasting Tourism Demand with Decomposed Search Cycles

Xin Li, Rob Law

Research output: Journal article publicationJournal articleAcademic researchpeer-review

67 Citations (Scopus)


This study aims to examine whether decomposed search engine data can be used to improve the forecasting accuracy of tourism demand. The methodology was applied to predict monthly tourist arrivals from nine countries to Hong Kong. Search engine data from Google Trends were first decomposed into different components using an ensemble empirical mode decomposition method and then the cyclical components were examined through statistical analysis. Forecasting models with rolling window estimation were implemented to predict the tourist arrivals to Hong Kong. Results indicate the proposed methodology can outperform the benchmark model in the out-of-sample forecasting evaluation of Choi and Varian (2012). The findings also demonstrate that our proposed methodology is superior in forecasting turning points. This study proposes a unique decomposition-based perspective on tourism forecasting using online search engine data.

Original languageEnglish
Pages (from-to)52-68
Number of pages17
JournalJournal of Travel Research
Issue number1
Publication statusPublished - 1 Feb 2020


  • ensemble empirical mode decomposition
  • Google search data
  • tourism demand
  • tourism forecasting

ASJC Scopus subject areas

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


Dive into the research topics of 'Forecasting Tourism Demand with Decomposed Search Cycles'. Together they form a unique fingerprint.

Cite this