Xenia: A context aware tour recommendation system based on social network metadata information

Tour planning and point-of-interest (POI) recommendation are two challenging and time-consuming tasks for tourists, predominately due to the large number of POIs a travel destination may contain and the complex constraints and parameters associated with the trip itself (e.g., time, budget, etc.). In this paper we present Xenia, a context-aware platform aiming to construct travel routes (i.e., ordered visits to various POIs that maximize the user's travel experience) that adhere to the aforementioned limitations by modeling and solving the tour planning dilemma through the Orienteering Problem (OP). To achieve this, we use geo-tagged photos, collected from Flickr and exploit their metadata (e.g., time-stamps, geolocation and user-generated tags). By utilizing these spatio-temporal data, we are able to identify the trajectory patterns of tourists during their vacations and determine the most popular POIs in any given city, along with the tourists sequential POIs visits and their corresponding durations. Finally, we evaluate the effectiveness of the proposed system against a set of typical baseline approaches.