Efficient R-Tree Exploration for Big Spatial Data

With the era of Big data, People, companies and devices are all becoming factories of data which generate an incredible amount of information. In several real-life applications, exploiting data with spatial characteristics is of great interest. For this purpose, different approaches have been proposed to deal efficiently with this kind of data (e.g., Spatial-Hadoop, Location-Spark). Spatial index based on R-tree is one of the modern solutions proposed to accelerate access to the desired information. Exploring such index could seriously impact performance. In this paper, we propose a novel approach to explore the R-tree spatial index, making it possible to minimise the number of disk accesses and consequently the execution time. Our experiments show that our approach outperforms the standard approach used by existing systems.