|
Track: Real Estate Services
Mark Harley
Geographic Data Technology, Inc.
11 Lafayette Street
Lebanon, NH 03766-1445
Telephone: 603-643-0330
Fax: 603-643-6808
E-mail: markh@gdt1.com
Building a Massive Real Estate Database Using SDE
Defining Issue: There are many decisions to make when purchasing a home in a new area. Organizing real estate data spatially allows a broker to provide detailed and comparative information about potential homes and neighborhoods. Constructing a seamless, nation-wide, geographic database with hundreds of millions of records presents a unique challenge and requires powerful GIS tools. GIS Solution: An SDE database was designed jointly by Esri, GDT, and the Realtor's Information Network (RIN) partners. GDT integrated data from numerous sources and built the SDE data set. Esri developed an extensive client/server software system to retrieve data, display maps, generate reports, and log transactions. Methodology: RIN needed a spatial database covering the entire United States including a basemap and census/postal geography with attached demographics and regional information. Also required were data on homes including each address, current telephone listing, property features, assessment, and any home
sales, foreclosures, nearby schools and hazardous materials sites. SDE was selected to be the repository for the information. The data was integrated from multiple vendors and much of it required a "spatial enabling" step before being displayed on a map. This was accomplished through geocoding and by attaching fields as attributes to existing polygonal regions. The system was designed to be updated often, due to frequent changes in the information. With an overall size in the hundreds of gigabytes, a partitioning scheme was developed to keep individual pieces manageable and yet appear seamless to the end user. Software: The central repository was built in SDE with the server applications running in a Sun/UNIX environment. The client software was prototyped in ArcView with a final implementation using MapObjects. A wide area network was established between the data integrator and central server using a T12 line to allow for continuous, high band width updates. This paper discusses the challenges and
corresponding solutions found for building and maintaining such a large SDE data set.
|
|
