An Association Rule Mining-Based Exploration of Travel Patterns in Wide Tourism Areas using A Wi-Fi Package Sensing Survey

Abstract

The number of inbound tourists in Hokkaido is increasing annually due to tourism promotions and policies, including unique terrain, natural eco-tourism, and outdoor activities. The attractive destinations and journeys information can assemble to predict traveler mobility. Distance and travel times affect travel behaviors in Hokkaido; also, most of the attraction locations are in the countryside and away from the city center. Hence, travel patterns are considered to maintain and improve the tourism atmosphere. In this research, travel patterns analyzed employing Wi-Fi probe data, which was collected in Asahikawa and Furano tourism areas. First, fundamental analysis was conducted at each spot. Next, the three indicators support, confidence, and lift parameters were calculated using the Apriori algorithm, which generally is used in marketing purposes. The results, the usefulness of association rule mining calculation provided rules of the transaction of migration travel characteristics — sequential travel patterns illustrated to identifying significant location toward sustainable tourism development.