MRFM: An efficient approach to spatial join aggregate

Abstract

Spatial join aggregate(SJA) is a commonly used but time-consuming operation in spatial database. Since it involves both the spatial join and the aggregate operation, performing SJA is a challenging task especially facing the deluge of spatial data. A popular model nowadays for massive data processing is the shared-nothing cluster using MapReduce. Thus, to explore SJA in MapReduce, a Map-Reduce-Filter-Merge(MRFM) algorithm is proposed.Map step divides the total SJA task into disjoint sets, then Reduce step aggregate each set individually, a Filter operation will filter those aggregate results of single assignment spatial objects.Finally, Merge step further aggregate the partial results of multiple assignment spatial objects using an efficient merge algorithm. Extensive experiments in large real spatial data have demonstrated the efficiency, effectiveness and scalability of the proposed methods. © 2012 Springer-Verlag.