Mining for Data
How do you sift through massive amounts of online information to find the real ore?

By James L. Sipes, ASLA

Images.com/Keith Skeen

As the saying goes, “garbage in, garbage out.” That is certainly true in landscape architecture, where decisions based on inaccurate, outdated, or biased data usually wind up being bad decisions. The key to making good design and planning decisions is to base these decisions on information that is accurate, valid, relevant, and correct. Unfortunately, that is easier said than done.

There is no shortage of digital data available via the Internet and organizational databases. At the same time, the amount of data that can be stored on disk drives increases dramatically every year, and the cost of this storage continues to drop—but there’s still no easy way to sift through all of this data to find information that is valid and meaningful.

Data Mining

One way to find data is to use a technique called data mining. Data mining involves sorting through massive amounts of data to find useful information. Data clearinghouses are becoming popular places to search for data. USGS’s Geospatial One Stop (www.geodata.gov) is a web-based portal for accessing maps, data, and other geospatial services. I used Geospatial One Stop extensively when I was mapping environmental impacts of hurricanes along the Gulf Coast.

Data mining software such as SAS Data Mining Software (www.sas.com), Statistica (www.statsoft.com), and SPSS (www.spss.com) allows you to sort through data, analyze it, organize and categorize it, and then identify patterns and trends. One of the really cool things about data mining is that it is possible to extract information from a database that you did not know even existed. For example, you can sift through information from a real estate service that provides information about the number and location of building permits issued in an area and then use that to determine potential growth patterns and purchasing trends.

…To read the entire article, subscribe to LAM!