Big Data: Does it really improve Forecasting techniques for Tourism Demand in Spain?

In this study, innovative forecasting techniques and data source from Big Data are used for the study of Hotel Overnight Stays for Spain, from January 2012 to December 2018. The unstoppable development of the tourism sector, together with the application of Big Data technologies, allow to make efficient decisions by economic agents. In this paper, univariate forecasting methodologies such as SARIMA and SSA are used. The use of the data obtained from the Google Data Mining tools allows to obtain knowledge. The ARDL models with seasonality explain easily when economic agents will make their decisions. ECM allows make forecasting for short-term and long-term. This fact means that tourist offers and demands can be perfectly adjusted at every moment of the year. As a criterion for the selection of models, the innovative Matrix U1 Theil is proposed, this allows to quantify how much a model is better than another in terms of forecasting.