Gunter Ulrich (MODUL Universität Wien)
Exploring the predictive ability of LIKES of posts on the Facebook pages of four major city DMOs in Austria
WIFO Research Seminar, 15.9.2020 13:30
Commentary: Oliver Fritz (WIFO)
Please send – up to one day before the event – an informal registration to, if you are interested in attending this lecture.
For online participation: GoToMeeting link will be submitted after registration (
Veranstalter: Österreichisches Institut für Wirtschaftsforschung
Online seit: 17.06.2020 0:00
Using data for the period 2010M06-2017M02, this study investigates the possibility of predicting total tourist arrivals to four Austrian cities (Graz, Innsbruck, Salzburg, and Vienna) from LIKES of posts on the Facebook pages of the destination management organisations of these cities. Google Trends data are also incorporated in investigating whether forecast models with LIKES and/or with Google Trends deliver more accurate forecasts. To capture the dynamics in the data, the autoregressive distributed lag (ADL) model class is employed. Taking into account the daily frequency of the original LIKES data, the mixed data sampling (MIDAS) model class is employed as well. While time-series benchmarks from the naive, error-trend-seasonal, and autoregressive moving average model classes perform best for Graz and Innsbruck across forecast horizons and forecast accuracy measures, ADL models incorporating only LIKES or both LIKES and Google Trends generally outperform their competitors for Salzburg. For Vienna, the MIDAS model including both LIKES and Google Trends produces the smallest forecast accuracy measure values for most forecast horizons.
Forschungsbereich:Strukturwandel und Regionalentwicklung