Trip Pattern Mining Using Large Scale Geo-tagged Photos

Photo-sharing websites allows people to display their experiences on the Web through rich media data such as photos and videos. These photos contain spatial context in terms of latitude and longitude where the photo was taken. The geotagged photos disclose much information about people travel behavior and tourist density. As web-based and mobile-based technologies advance, geo-tagged photos are increasingly collected beyond the capability of human analysis; so, have provided new research opportunities and challenges. In this work, we study on analyzing the travel behaviors based on the geo-tagged photos. This research proposes a framework that is used to organize a large collection of geotagged photos collected from Flickr to extract trip patterns from geo-tagged photos in Europe. This framework includes two data mining techniques based on a MapReduce framework: clustering, and frequent pattern mining. We explain using these techniques to analysis a large photo collection, while also revealing various interesting relations about popular cities and landmarks at Europe scale. We report interesting experimental results.