PopTour: Discovering Journey Group T-Patterns from Instagram Trajectories to Recommend Hot Travel Routes
Shuangyu Yu, Yaxin Yu, Yulong Li, Xin Liu · 2014
Instagram is an popular photo-sharing smart phone application of social network and is widely used among tourists to record their journey information such as location, time and content. Consequently, huge volume of spatio-temporal data are generated in the form of trajectories. Discovering useful patterns from these trajectories can reveal valuable knowledge to a variety of critical applications. In this light, we propose a novel concept, called Journey Group (JG), which is a group trajectory pattern reflecting a large number of users who walk through a common trajectory and depart from the trajectory for several times allowed. In this paper, we focus on data generated by Instagram to discover the JG Trajectory Patterns i.e., JG T-Patterns, from travel trajectories. Previous researches on T-Patterns mining concentrate on GPS-based data, which is different from Instagram data, a kind of UGC-based (User Generated Content based) data. GPS-based data is dense because it is often generated automatically by moving devices in a certain pace, while UGC-based data is sparse because data is generated randomly by the uploading of users. Aiming at this, a novel JG T-Patterns mining strategy from UGC-based data is also proposed. Finally, a demo for discovering hot travel routes based on JG T-Patterns from Instagram trajectories, named Pop Tour, is implemented and experimental results show that Pop Tour is an effective system.