Proceedings of The 11th International Conference on Modern Research in Management, Economics and Accounting
Google Trends and Tourism: Regression Cluster Analysis
Miguel Ángel Ruiz Reina
The appearance of Big Data technologies and economic development have led to a new generation of data that must be analysed. This new data structure allows identifying consumer behaviour; modelling offers and allowing countries to design Economic Policies efficiently. The service sector increasingly weights in the GDP of the nations; in particular, in Spain, it is around 12 per cent. Search engines from Google collect this data information by areas and time. In particular, the crossing of data from primary sources and secondary data is a comparative advantage over traditional analyses. We analyse datasets in this work coming from official statistical sources in Spain and data from Google Trends. The objective of this work is explaining the hotel demand trough previous searches of consumers from January 2008 to December 2019 for the Spanish case. The application of the method called Regression Cluster Analysis (RCA) represents an improvement in forecasting modelling of the literature. This model allows searching profiles to be made as a previous step to what is known in the literature as Next Best Activity. We compare three models (RCA, ADRL + Seasonality, SARIMA) thought the Ratio Theil U2, for a time horizon of six months. This methodology involves adding one more model to the debate on the use of Big Data. At the end of the work, we propose three ideas for “sustainable” development of this type of analysis with open data.
Keywords: Big Data, Forecasting, Google Trends.