Discovering spatial high utility frequent itemsets in spatiotemporal databases

Abstract

Spatial High Utility Itemset Mining (SHUIM) aims to discover all itemsets in a spatiotemporal database that satisfy the user-specified minimum utility (minUtil) and maximum distance (maxDist) constraints. The popular adoption and successful industrial application of SHUIM suffers from the following two limitations: (i) Since SHUIM determines the interestingness of an itemset without taking into account its support within the data, SHUIM facilitates sporadic itemsets with high utility to be generated as SHUIs. In particular, items in long transactions can combine with each other and be generated as SHUIs. (ii) SHUIM is a computationally expensive process because the generated itemsets do not satisfy the downward closure property. This paper introduces Spatial High Utility Frequent Itemset Mining (SHUFIM) to address these two issues. A SHUI in a spatiotemporal database is said to be a SHUFI if its support is no less than the user-specified minimum support (minSup) constraint. The usage of minSup not only facilitates the proposed model to be tolerant to the long transactions within the data but also facilitates us to employ additional pruning techniques to reduce the computational cost. A single scan fast algorithm has also been proposed to discover all SHUFIs in a spatiotemporal database. Experimental results demonstrate that the proposed algorithm is efficient. We also demonstrate the usefulness of the proposed model with a real-world application.