In light of the growing prevalence of location-based applications, a substantial amount of location data has been accumulated, facilitating various services and research endeavors such as traffic flow analysis, identification of similar movement patterns, and detection of social circles. This surge in data usage has, however, prompted heightened concerns about location privacy among users. Our work aims to fortify users' location privacy without compromising the advantages brought by location-based services. Specifically, we propose an innovative deep learning model called GeoGRCNN which is capable of generating synthetic trajectories indistinguishable from real human trajectories. Such synthetic trajectories can then be mixed with real human trajectories to serve as decoys that prevent users from being profiled by the service provider. Our model seamlessly combines Graph Convolutional Networks(GCN) and Recurrent Neural Networks(RNN) to faithfully replicate authentic movement patterns within urban environments. Our evaluation focuses on measuring the spatial accuracy of these trajectories and their conformity to actual road networks. The experimental results demonstrate that our model effectively replicates actual urban layouts with minimal deviations from road networks. We achieve trajectory generation accuracy of 95.06 % for Beijing and 97.37% for Atlanta.

The widespread use of smart mobile devices has resulted in a massive accumulation of trajectory data by service providers. The analysis of human trajectories, particularly semantic location information, has opened up avenues for discovering common social behavior and enhancing social connections, leading to a range of applications such as friend recommendations and product suggestions. However, the exponential growth of trajectory information generated every day presents significant challenges for existing trajectory analysis algorithms, which are no longer capable of delivering timely analysis results. To address this issue, we propose a highly efficient algorithm that can recommend social communities for new users in real time by leveraging knowledge gained from large-scale semantic trajectories. Specifically, we develop a novel two-branch deep neural network model that extracts semantic meanings at different levels of granularity from human trajectories and uncovers the hidden relationship between trajectories and social communities. We then utilize this model to perform instant social community recommendations. Our experimental results have demonstrated that our approach is not only significantly faster than traditional trajectory analysis algorithms in terms of social community recommendation, but also preserves high prediction accuracy with F1-score above 97%.

In today's society, location-based services are widely used which collect a huge amount of human trajectories. Analyzing semantic meanings of these trajectories can benefit numerous real-world applications, such as product advertisement, friend recommendation, and social behavior analysis. However, existing works on semantic trajectories are mostly centralized approaches that are not able to keep up with the rapidly growing trajectory collections. In this paper, we propose a novel large-scale semantic trajectory analysis algorithm in Apache Spark. We design a new hash function along with efficient distributed algorithms that can quickly compute semantic trajectory similarities and identify communities of people with similar behavior across the world. The experimental results show that our approach is more than 30 times faster than centralized approaches without sacrificing any accuracy like other parallel approaches.