Abstract
The influence zone method is used to answer reverse k nearest neighbours (RkNN) queries using a region approach without having to verify each object; this makes the answering of a RkNN query more efficient than does the conventional point-to-point approach. However, the influence zone is unable to answer the dynamic value of k efficiently as it needs to know this value in advance. In this paper, a concept is introduced whereby the influence zone is expanded to enable answering the dynamic value of k as well. Furthermore, a concept is proposed that expands the influence zone of k= 1. Experimental results indicate that the expanded influence zone is able to answer RkNN queries even when the k value is dynamic, without the need to recompute it. Furthermore, the experiments show that the use of pre-computed regions will provide a stable query time for any value of k.