Real Estate Solutions
The ability to mine and leverage the vast amount of information in property appraisals
is a critical business advantage. There are many appraisal software applications
that capture the data entered by appraisers nationwide. Each vendor defines and
stores data differently and with a wide variety of data quality. PDSA’s technology
has created an appraisal import, data transformation, and data persistence approach
to construct an appraisal data repository. The repository can now be used to produce
a vast amount of business intelligence reports.
|
|
What is a Real Estate Appraisal?
Real estate appraisal, property valuation or land valuation is the practice of developing
an opinion of the value of real property, usually its market value. The need for
appraisals arises from the heterogeneous nature of property as an investment class:
no two properties are identical, and all properties differ from each other in their
location - which is the most important determinant of their value.
How are Appraisals Created?
The appraisal practice is regulated and standardized. Several software companies
have created tools to help appraisers develop and create appraisals. Two very popular
software vendors in this field are ACI and ‘a la mode, inc’. Each vendor supports
all the necessary forms for appraisers to fill out. These are all great tools. But
if we think of the downstream process where organizations may later want to report,
analyze, conduct business intelligence against this data, then you must properly
prepare the data for it to have any value.
Standardizing the Appraisal Data Model
|
The data inputting and quality of appraisal data is, at times, weak. For example,
a zip code may be entered as the word ‘same’ implying the ‘same’ as the subject
property. But if this data was entered into a database for reporting purposes it
would be considered an invalid zip code. Each vendor has chosen their own data model
to capture the necessary data. Hence, there will be differences between vendors
but most data may be similar. To solve this problem, mappings must be defined so
as to map a field from one vendor into the same field from the other vendor. Our
data model has 50+ tables and 3,800+ fields!
|
Leveraging XML – But You Need More
|
|
The appraisal software allows the appraisal to be exported in XML. XML, in combination
with other standards, makes it possible to define the content of a document separately
from its formatting, making it easy to reuse that content in other applications
or for other presentation environments. Unfortunately the structure of the XML between
vendors is different. Our challenge therefore was to define a generic data model
and provide the intelligent mapping engine to map all of the XML elements from any
vendor into the appropriate database table and field. XML is a great tool, but since
each vendor defined their own XML data model, we needed to provide the intelligent
mapping engine to ‘normalize’ the data for enhancing its usage.
|
|
Using Geocoding Intelligently
|
An element of mapping and scrubbing the data is also to apply geocoding technology.
Geocoding may sound easy but it has many complexities. Difficulties arise when distinguishing
between ambiguous addresses such as 742 Evergreen Terrace and 742 W Evergreen Terrace
or you are attempting to geocode new addresses for a street that is not yet added
to the geographic information system database. Hence, it is important to use a company
such as PDSA which has a rich experience in producing the highest hit rate possible
based on your particular needs. We’ve even found that using the same address and
calling different geocode services sometimes yield different addresses in completely
different areas, so we need to examine the results to select the more reasonable
one. We’ve also found that the same service may return more than one address, leaving
it up to the consumer to decide which is best for their use.
|
|
The Appraisal Repository – The Goal
The appraisal repository is the database that houses all of the appraisals that
were imported by using our technology. Now that the data has been normalized, scrubbed
and prepared for use, others can now develop report queries against the data. For
example, show me all appraisals that were performed in zip code 92780 between January
1, 2009 and January 30, 2009 with 3 or more bedrooms. Now your team can present
numerous questions to do the necessary ‘mining’ for valuable information to accelerate
your business.