Tourismo: A User-Preference Tourist Trip Search Engine

In this demonstration we re-visit the problem of finding an optimal route from location A to B. Currently, navigation systems compute shortest, fastest, most economic routes or any combination thereof. More often than not users want to consider “soft” qualitative metrics such as popularity, scenic value, and general appeal of a route. Routing algorithms have not (yet) been able to appreciate, measure, and evaluate such qualitative measures. Given the emergence of user-generated content, data exists that records user preference. This work exploits user-generated data, including image data, text data and trajectory data, to estimate the attractiveness of parts of the spatial network in relation to a particular user. We enrich the spatial network dataset by quantitative scores reflecting qualitative attractiveness. These scores are derived from a user-specific selfassessment (“On vacation I am interested in: family entertainment, cultural activities, exotic food”) and the selection of a respective subset of existing POIs. Using the enriched network, our demonstrator allows to perform a bicriterion optimal path search, which optimizes both travel time as well as the attractiveness of the route. Users will be able to choose from a whole skyline of alternative routes based on their preference. A chosen route will also be illustrated using user-generated data, such as images, textual narrative, and trajectories, i.e., data that showcase attractiveness and hopefully lead to a perfect trip.