Machine learning for classifying international tourists using big data/social media photos

Keywords: Big data, Machine Learning, tourism activity
Session Type: Virtual Paper Abstract
Day: Monday
Session Start / End Time: 2/28/2022 03:40 PM (Eastern Time (US & Canada)) - 2/28/2022 05:00 PM (Eastern Time (US & Canada))
Room: Virtual 7

Abstract

The recovery of the tourism sector is crucial for global economy – now more than ever. To achieve this, a set of agreed upon and reliable data is required. As such, the first and most basic step in gathering reliable data on tourism is creating an applicable and consistent definition of the term tourist. However, a key issue relates to the range of existing definitions of “tourists” that results in different methods for counting tourists and measuring tourism.
The objective of this study is to offer a practical and adaptable model for identifying international tourists and measuring tourism using machine learning and big data. The innovative non-linear model for defining international tourists relies on photographs posted on the Flickr social media platform and based on the city structure. The developed classification model produced above 90% success rate in identifying international tourists in Vienna, Prague, and Manhattan, offering new travel indicators that differentiate international tourists, like repeat visitations, travelling distances, and short stays.
The classification model developed, and the ideas promoted and implemented in this research provide a new perspective to the definition and classification of tourists. It integrates existing knowledge in the field, while using new methods and non-authoritative datasets, together with new viewpoints of what can be considered as tourism activity – and where. Moreover, the model offers a practical, adaptable, and scalable application for tourism practitioners, as well as researchers, for measuring and analyzing tourism activity and behavior – a crucial competence in today’s struggling tourism industry.