Tourist Flow Simulation in GAMA Using Historical Data Parameters

Abstract

Decision makers in the tourism sector deal with various issues and need high-quality information to support their decisions. We propose a data-centric approach that analyses historical point of interest (POI) check-in data to determine parameters for an Agent Based Model (ABM). ABM simulation is then run multiple times to simulate possible outcomes in terms of the tourist flow. We have tested the proposed approach on the city of Salzburg using check-in data from Salzburg Card users across 29 POIs. These data were used to parameterize the ABM model with the number of people, the number of POIs a person visits per day, and the preference for selecting POIs to visit. The simulation was performed in GAMA ABM platform and the spatial environment was based on buildings and roads from OpenStreetMap (OSM). Simulation for the duration of 1 day has been repeated 50 times to generate POI visiting patterns. The simulation results have been compared to the ground truth data for the same day and they show that the approach can recreate the long-term pattern of POI visits, but has over-estimated several POIs that had lower visitor counts on that specific day.

Bridging Disciplinary Divides through Computational Social Sciences and Transdisciplinarity in Tourism Education in Higher Educational Institutions: An Austrian Case Study

“…The research on the individual level comprises spatiotemporal consumer behavior and attitudes (e.g., [32,33]), whereas the organizational level deals with marketing and financial performance. On an industry level, tourism flow [34,35], demand volume [36][37][38], and sustainable development [39,40] are of particular interest. Naturally, security and privacy issues are also addressed, especially when dealing with location-based social media [41,42].…”