Yesterday, I had the pleasure of giving one of the plenary presentations at the 2007 Florida Geospatial Conference at Cape Canaveral. The theme of the conference and the title of my presentation is Integrating Geospatial Data.
Following are the slides and some of the commentary. I ran the presentation as more of a workshop as opposed to a speech so some of the ideas expressed below are attributable to the group as a whole.
What does integrating geospatial data mean?
I believe it means different things to different people. Its meaning is dependent upon context—the context of the data being integrated, the question to be answered by the melding of data and the people doing the integration.
Google says there are about 905,000 web pages that talk about integrating geospatial data. Therefore it must be important!
It is important if it leads us to greater understanding of our business, organization and the world. Geographic visualization often has a transformational effect on our understanding of information. It allows us to perceive patterns that could not be seen in the same data if only viewed as records in data table.
In the 1980’s and 1990’s the focus of our discussions on integrating geospatial data revolved around the technical problems of integrating different map layers into a single map.
We talked about map accuracy, scale and source lineage. While these issues still exist, I believe they are well understood and no longer represent a significant barrier.
During this time the related problem of data interoperability was a high priority in the geospatial community. Data interoperability is the idea that information in one software package can be read by another (often competing) software package. Data interoperability is no longer a significant issue either.
Today our integrations should be focused on bringing business data into geospatial systems for analysis and visualization to support decision making. After all, this is the unfulfilled promise of the geographic information system.
Are we well integrated now? Yes and no.
We’re much better than we were a decade ago, but we are at the beginning of a new type of integration with non-geographic information. We’re at a tipping point and I think we’re going to see highly creative integrations of data from many sources. The information will be pulled together by its geography and will be visualized on a map.
In doing so we will understand patterns that we did not understand and find solutions that we could not find before. Where we are going will be transformational.
Who’s in charge? You are! Why? Because integrations are context driven.
Gone are the days of standards committees that met for months and years to try to get 10 or 20 or 50 organizations to agree to use the same data model to represent real world objects inside the geographic information system. I was involved in some of these efforts and, looking back, they were mis-guided.
What we really need is data semantics. Semantics are definitions of sorts, where I can create my own model for, say, a road. In my model the road has attributes like “name” and “lanes.” You also have a model for roads in your GIS, but your road’s attributes are called “rdname” and “numlanes.” Semantics help our systems talk to each other by acting as a translator. In this case telling the two systems that “name” and “rdname” both store the name of the road.
The beauty of this is that now neither one of us needs to agree on a data model. I can use the model that makes sense for me and you can use the model that makes sense for you, but we can still share information because we have an easy way to translate between models.
Unfortunately, we are usually the obstacle to data integration. People and politics are the most commonly cited barriers to successful integrations. We’ve tackled most of the big technical problems—they were easy to solve. Now we’re left with the people problem. It stems from the false premise that information is power. Information hoarded is not power, information shared is power.
There are a number of new technologies that are making the integration of geospatial data much simpler. The most important new technology for data integration is web services.
Google Earth is fueling a geographic awakening in society. People that would never know that GIS existed have used Google Earth or seen it on CNN. It’s given geographic visualization to the masses. It has more effectively communicated events such as the crisis in Darfur, Sudan than textual reports could.
Mashups are great examples of the integration of geospatial and non-geospatial data. They are simple to implement, accessible by non-technical people and don’t require bulky client applications to view them.
Mashups are driven by a technology called web services. Web services are used to publish both data and computing functions over the web for others to use in their own applications. For GIS this is driving us away from redundant mapping where each organization often mapped base layers like roads, land parcels, topography and aerial imagery. It is driving us towards the concept of federated or distributed GIS. In this new concept an organization will focus on delivering the data and or analysis functions for areas where they are the experts. The organization will pull in the other basic data they need from web services created by other organizations.
Web services are also used in back office business systems. This means that we can use one technology to integrate geospatial and non-geospatial data in a mashup.
Some web services will be offered for free, while others will be for-pay services. That’s one of the beauties of the web service concept—it supports different business models. While we’d all like to get things for free, the idea of you get what you pay for still holds true on the web.
Some great examples of mashups can be found at:
Geo-social networking builds on the social networking concept used on sites like myspace, facebook and flikr by adding a location to the information posted on those sites. Some of the tools for geo-social networking include:
These tools are changing the way we find things and I think they will make small businesses as easy to find as large national brands.
There are big societal trends that are beginning here. Small is overtaking big. Grassroots is becoming more powerful than the institution.
GeoRSS is an emerging standard for adding location tags to RSS feeds. This has the potential to make is vastly easier to find information about what’s around you. It also has the potential to be the pipe that connects different GIS systems to one another and keep them updated as data changes. It’s just being formulated now, but has the potential to huge!
Read more at www.georss.org.
So is all of this stuff just for college kids with a myspace account? Not by a long shot! The idea of geo-social networking is already being talked about for disaster response as evidenced by the LoSoNet Conference last year. When you think about it, it’s the perfect way to gather information from many different sources and visualize it on a map. It should make the task of gaining situational awareness much easier.
I can see the local fire station or police precinct or mobile response unit reporting on the conditions at their location. All of it will be done with simple web based tools (no expensive desktop mapping software). Then the crisis managers will see it all in one unified, map centric view. They’ll get the whole picture instead of just bits and pieces of information.
What a great way to use these “social” tools!
Integrating geospatial data is personal—it depends on your context. So how will you do it?
