Parallel gathering discovery over big trajectory data

The advances in location-acquisition technologies have generated massive spatio-temporal trajectory data, which represent the mobility of a diversity of moving objects over time, such as people, vehicles, and animals. Discovery of traveling companions on trajectory data has many real-world applications. Most of existing discovery approaches are limited to centralized computing, while these techniques for handling large-scale trajectory data require considerable performance improvement. Parallel computing essentially provides an alternative method for handling this problem. In this work, we first present the design and implementation of both batch and streaming gathering patterns discovery algorithm in a distributed parallel computing fashion. Afterwards, we further propose several optimization techniques for efficient computation. Finally we conduct extensive experiments based on a public dataset to evaluate the efficiency of our approaches and effectiveness of optimizations using Amazon EC2 clusters.