Generating tourism path from trajectories and geo-photos

The pervasiveness of GPS devices enables tourists recording their trajectories and uploading geo-tagged photos. Geo-related data has emerged as new source for travelers to refer to when making tourism decisions. As the increasing availability of these user-generated experiences on the social networks, there is a need to automatically discovering useful patterns for potential travelers. In this paper, we propose a tourism path by incorporating the trajectories and geo-photos. Specifically, we provide an algorithm for precisely matching user-uploaded photos to tourism sites and a density based clustering approach to identify the place of interests inside tourism sites. We then build a model that adapts the well-known HITS algorithm to detect interesting points and trajectories with high utility scores and design an algorithm for efficiently computing rational routes for visiting tourism sites. Finally, experimental results illustrate the advantage of the proposed density-based algorithm and confirm the effectiveness applicability of our tourism path discovering approach.