Selectivity estimation in spatial networks

Modern applications requiring spatial network processing pose many interesting query optimization challenges. In many cases, query processing depends on the corresponding graph size (number of nodes and edges) and other graph parameters. This dependency may be local or global. In this paper, we present novel methods to estimate the number of nodes in regions of interest in spatial networks, towards predicting the space and time requirements of range queries. We examine all methods by using real-life and synthetic spatial networks. Experimental results show that the number of nodes can be estimated efficiently and accurately with small space requirements, thus providing useful information to the query optimizer. Categories and Subject Descriptors H.2.8 [Database management]: Database applications-Spatial databases and GIS General Terms Algorithms Performance Keywords estimation, optimization, spatial networks

• Type: Journal
• Year: 2008
• Where: SAC
• Authors: Eleftherios Tiakas, Apostolos N. Papadopoulos, Alexandros Nanopoulos, Yannis Manolopoulos
• Keywords: Applied Computing, Interesting Query Optimization, SAC 2008, Spatial Networks, Synthetic Spatial Networks