A deep learning approach for daily tourist flow forecasting with consumer search data

Binru Zhang, Nao Li, Feng Shi, Rob Law

Research output: Journal article publicationJournal articleAcademic researchpeer-review

13 Citations (Scopus)


This study introduces the concept of long short-term memory (LSTM) network to handle complex time series forecasting problems in the tourism industry. To validate the efficiency of the developed method, we used the daily tourist flow and consumer search data of Jiuzhaigou, a popular tourist spot in China, from 8 October 2013 to 7 August 2017 as the experimental dataset for empirical analysis. According to the 150-day forecasting results, LSTM shows the best statistical performance in the training and test sets compared with its counterparts.

Original languageEnglish
Pages (from-to)323-339
Number of pages17
JournalAsia Pacific Journal of Tourism Research
Issue number3
Publication statusPublished - 3 Mar 2020


  • artificial intelligence
  • big data
  • consumer search data
  • daily tourist flows
  • Deep learning
  • forecasting precision
  • long-term dependence
  • LSTM network
  • relatively large sample
  • tourism demand forecasting

ASJC Scopus subject areas

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

Cite this