Spatial Query Processing Using Generalized Filter

This is the author's version of the work. It is posted here with the permission of the publisher for your personal use. No further distribution is permitted. If your library has a subscription to these conference proceedings, you may also be able to access the published version via the library catalogue. Spatial data, ranging from various land information data to diierent t ypes of environmental data, are typically collected and used by diierent custodians. Full beneets of using spatial data can be achieved by c o m bining the data from diierent sources covering a common region. Due to organizational, political and technical reasons, it is unreal-istic to physically integrate vast amount of spatial data managed by diierent systems in diierent organizations. A practical approach i s t o p r o vide interoperability t o s u p-port multi-site data queries. In this paper, we study the performance aspect of complex spatial query processing. We propose a framework for processing queries with multiple spatial and aspatial predicates using data from multiple sites. Using a new concept called generalized lter, a query is processed in three steps. First, an aspatial lter that incorporates some conditions derived from spatial predicates is used to nd a set of candidates, which is a superset of the nal query results. Then, the candidates are manipulated and a reenement step is executed following an optimized candidate sequence. Finally, a post-processing step is used to handle spatial expressions in query results. The focus of this paper is to generate enhanced lters in order to minimize the need of transferring and processing complex spatial data.