Design of Distributed Heterogeneous Data Retrieval Engine Using Service Oriented Architectures and Rich Internet Applications

Abstract

As the demand for geospatial data rapidly increases, the lack of efficient way to find and retrieve appropriate data embedded in the mountain of available data on the Internet becomes critical. To meet this challenge, this paper presents a novel design of the data retrieval engine (DRE) targeted at facilitating one stop retrieval of geospatial data from heterogeneous data sources distributed on the Internet. The engine enables users to search, access and visualize the geospatial data of different types by setting query criteria in a map-based web portal. The engine is based on the service oriented architecture. It contains a built-in middleware composed of a harvest server, a central service database and a service search engine. Meta information of Open Geographic Consortium (OGC) web service are harvested by the harvest server and stored in the service database. When searching for particular geospatial data, those cached OGC services providing access to the data can be identified by the service search engine. The core functionalities of the middleware are exposed through a set of web services which can be consumed at client side. Two popular Rich Internet Applications (RIAs) technologies including Microsoft Silverlight and ArcGIS API for Silverlight are fully utilized to develop the web portal. The feasibility and effectiveness of the engine are carefully tested and demonstrated through investigations.