LAUSR

Abstract It is important to mine moving object gathering pattern, but the traditional algorithms cannot effectively analyze the trajectory data with mass scale. A parallel mining algorithm based on Resilient Distributed Datasets (RDD-Gathering) and R-tree index is proposed to discover the gathering pattern in the massive trajectory data, and the mining functions are provided as cloud services on Spark. MongoDB cluster is constructed to store massive trajectory data, R-tree index structure is used to store the trajectory data, and the RDD-Gathering algorithm interface is encapsulated with RESTful cloud service. The RDD-Gathering algorithm is implemented based on Apache Spark and performances are better than normal methods in real trajectory dataset testing.