Article Preview
Top1. Introduction
Geospatial data integration can be used to improve data quality, to reduce costs, and to make data more useful to the public(Auer et al.,2009; Bittner et al.,2009; Brodt et al.,2010; Kuhn,2002; Su et al.,2012; De Carvalho et al.,2012; Su & Lochovsky,2010; Ballatore et al.,2014; Buccella et al.,2010; Fonseca, Egenhofer et al.,2002; Malik et al.,2010;Vaccari et al.,2009). However, the large amount of data is produced by a variety of sources, stored in incompatible formats, and accessible through different GIS applications. Thus, geospatial data integration is difficult and becoming an increasingly important subject.
To implement the geospatial data integration, four problems need to be addressed: geospatial data retrieving, modeling, linking and integrating. This paper proposes corresponding approach for each issue. Besides, our work takes advantage of Karma (Szekely et al.,2011; Knoblock et al.,2012; Taheriyan et al.,2012; Tuchinda et al.,2011; Knoblock et al.,2011), which is a general information integration tool. It supports importing data from a variety of sources including relational databases, spreadsheet, KML and semi-structured Web pages, and publishing data in a variety of formats such as RDF. The source modeling work is based on these functions: