Geospatial Semantic Web Blog

In geospatial information systems (GIS), data integration is often a problem. Different systems may use different vocabularies to represent the same abstract concept, and different systems may express data values in different unit of measure (UOM). This problem may be of interest to the Semantic Web community because it’s a different kind of semantic interoperability problem.

I believe GIS data integration is not much about building knowledge representation models for things in the world or developing logical inference for reasoning about properties of things. It’s about how to detect misalignment in data representation and align functional computation to produce accurate and consistent results. An instance of the GIS data integration problem is the management of coordinate reference system (CRS).

CRS defines how georeferenced spatial data relates to real locations on the Earth’s surface. There are many different CRS datasets — in the order of thousands. A common CRS that most people are familiar with is WGS84, which is a CRS used in GPS devices to describe latitude and longitude coordinates. However, WGS84 is not a universal CRS. For example, in the oil drilling industry, WGS84 is only one of many CRS datasets that companies use. For this reason, it’s often challenge for independently developed GIS applications to reuse and integrate data.

To make the data integration problem worse, different countries and regions often use different CRS. In additional, through out the GIS history, various versions of CRS standards have been developed. In 1994, the European Petroleum Survey Group (EPSG) publishes a set of CRS shared by major oil companies. In 2005, EPSG was absorbed by OGP (International Association of Oil & Gas Producers). The current release of the EPSG CRS dataset is published in Microsoft Access. Soon this will be replaced by an online Web Service registry.

So, what has been done to address this data integration problem? Markup language GML is an emerging standard that encourages the integration of GIS data. Companies are developing web service registry (e.g., INdicio Geodetic Registry) for developing and sharing CRS definitions over the Web.

Some people may be surprised that the foundation of the GIS world is not built on a single coordinate reference system (e.g., WGS84 latitude and longitude). While this creates obstacles in enabling system interoperability, but it also creates new opportunities for semantic web research, finding applications of semantic technology in real-world problems.

Resource: Managing the Earth’s Precious Natural Resources