Xnavi

Abstract

Though an increasing number of people is now involved in travel planning owing to the spread of the internet, it is still difficult for travelers to plan trips on their own. It is especially difficult for tourists using automobiles because they have several choices of accessible places. To make itineraries easily, travelers require a travel planning system that suggests two types of experiences: experiences characterizing the travel area and experiences stemming from a flow between the former experiences. Existing systems do not list specific spontaneous experiences of interest to travelers. In response, Xnavi, a travel planning system for drivers based on experience flows, is proposed, which provides these types of experiences. To recommend experience flows, Xnavi extracts experience keywords related to the travel area using natural language processing based on the TF-IDF method and also extracts flows of tourist attractions' attributes based on association analysis of driving histories. Trials of the proposed method and a user study were conducted. The results show that Xnavi is effective at suggesting experiences and satisfying tourists with their plans.