Multiple Object Types KNN Search Using Network Voronoi Diagram
Abstract. Existing work on k nearest neighbor (kNN) in spatial/mobile
query processing focuses on single object types. Furthermore, they do not
consider optimum path in KNN. In this paper, we focus on multiple type
kNN whereby the interest points are of multiple types. Additionally, we
also consider an optimum path to reach the interest points. We propose
three different query types involving multiple object types. Our algorithms
adopt the network Voronoi Diagram (NVD). We describe two
ways to solve multiple types of KNN queries: one is to create NVD for
each object type, and two is to create one NVD for all objects. The
comparison between these two approaches is presented in performance