Forecasting tourist arrivals with the help of web sentiment: a mixed-frequency modeling approach for big data

• Irem Önder ^[1] ; Ulrich Gunter ^[2] ; Arna Scharl ^[2]
  1. [1] University of Massachusetts System

     University of Massachusetts System

     City of Boston, Estados Unidos

     
  2. [2] MODUL University Vienna

     MODUL University Vienna

     Innere Stadt, Austria

     
• Localización: Tourism analysis, ISSN 1083-5423, Vol. 24, Nº. 4, 2019, págs. 437-452
• Idioma: inglés
• Texto completo no disponible (Saber más ...)
• Resumen

  • Online news media coverage regarding a destination, a form of big data, can affect destination image and influence the number of tourist arrivals. Sentiment analysis extracts the valence of an author's perception about a topic by rating a segment of text as either positive or negative. The sentiment of online news media can be viewed as a leading indicator for actual tourism demand. The aim of this study is to examine if web sentiment of online news media coverage of four European cities (Berlin, Brussels, Paris, and Vienna) possesses information to predict actual tourist arrivals. This study is the first to use web sentiment for forecasting tourism demand. Automated semantic routines were conducted to analyze the sentiment of online news media coverage. Due to the differing data frequencies of tourist arrivals (monthly) and web sentiment indicators (daily), the MIxed-DAta Sampling (MIDAS) modeling approach was applied. Results indicate that MIDAS models including various web sentiment indicators outperform time-series and naive benchmarks in terms of typical accuracy measures. This study shows that utilizing online news media coverage as an indication of destination image can improve tourism demand forecasting. Because destination image is dynamic, the results can vary depending on time period of the analysis and the destination. A managerial implication of the forecast evaluation exercise is that destination management organizations (DMOs) should add models incorporating web sentiment data to their forecast modeling toolkit to further improve the accuracy of their tourism demand forecasts.