MRSLICE: Efficient RkNN Query Processing in SpatialHadoop
Date
2019Language
en
Keyword
Abstract
Nowadays, with the continuously increasing volume of spatial data, it is difficult to execute spatial queries efficiently in spatial data-intensive applications, because of the limited computational capability and storage resources of centralized environments. Due to that, shared-nothing spatial cloud infrastructures have received increasing attention in the last years. SpatialHadoop is a full-edged MapReduce framework with native support for spatial data. SpatialHadoop also supports spatial indexing on top of Hadoop to perform efficiently spatial queries (e.g., k-Nearest Neighbor search, spatial intersection join, etc.). The Reverse k-Nearest Neighbor (RkNN) problem, i.e., finding all objects in a dataset that have a given query point among their corresponding k-nearest neighbors, has been recently studied very thoroughly. RkNN queries are of particular interest in a wide range of applications, such as decision support systems, resource allocation, profile-based marketing, location-based services, etc. In this paper, we present the design and implementation of an RkNN query MapReduce algorithm, so-called MRSLICE, in SpatialHadoop. We have evaluated the performance of the MRSLICE algorithm on SpatialHadoop with big real-world datasets. The experiments have demonstrated the efficiency and scalability of our proposal in comparison with other RkNNQ MapReduce algorithms in SpatialHadoop. © 2019, Springer Nature Switzerland AG.