The application in mobile computing of spatial association rules mining algorithm based on separating support items

Abstract

In mobile computing many data expressing spatial locations are very important to extract spatial association rules, which can provide potential and useful information for mobile clients. In order to fast mine these spatial association rules and improve efficiency of mobile intelligent system, this paper proposes an algorithm of spatial association rules mining based on separating support items, which can extract location association of spatial object in mobile intelligent system according to mobile client demand. The algorithm firstly uses the way of spatial buffer analysis to extract spatial predicate values among these given spatial predicate, target object and non-target object, and then uses every target object to create a transaction and turns these transactions into transaction vector, finally uses the down search method of separating support item to extract spatial association rules of meeting mobile client demand. Neither does it need create candidate frequent item sets nor compute support to extract spatial association rules, and so the algorithm avoid to create redundant candidate and reduces repeated computing to improve its efficiency. The result of simulate experiment indicates that the algorithm is faster and more efficient than present mining algorithms when extracting spatial association rules of meeting mobile client demand in mobile computing. © 2011 Springer-Verlag Berlin Heidelberg.