Progressive top- k subarray query processing in array databases

Abstract

Unprecedented amounts of multidimensional array data are currently being generated in many fields. These multidimensional array data naturally and efficiently fit into the array data model, and many array management systems based on the array data model have appeared. Accordingly, the requirement for data exploration methods for large multidimensional array data has also increased. In this paper, we propose a method for efficient top- k subarray query processing in array databases, which is one of the most important query types for exploring multidimensional data. First, we define novel top- k query models for array databases: overlap-allowing and disjoint top- k subarray queries. Second, we propose a suite of top- k subarray query processing methods, called PPTS and extend them to distributed processing. Finally, we present the results of extensive experiments using real datasets from an array database, which show that our proposed methods outperform existing naïve methods.