Re: Editing for fake names; Data cleansing

[CRPence <CRPbottle@xxxxxxxxx>]

> On Wednesday, June 10, 2009 10:43 AM Jim Franz wrote:
>
> Looking for a list of fake names to edit for exclusion. 
> Have already excluded
> FRED FLINSTONE
> MICKEY MOUSE
> UNKNOWN (and several variations)
> N/A
> Googled for a list but mostly found name generators...
> Anyone have a more complete list to share.

  I would suggest generating a list of real names against which to 
test for a match, to test for /inclusion/ instead of /exclusion/. 
And to optionally assign a weight to the likelihood that any name is 
presumed legitimate, as part of a[n a]synchronous service if a[n 
obvious or exact] match is not found in the database of known names; 
a null weight could indicate the name has not been evaluated.  Some 
services such as a phone directory lookup may serve as a database of 
presumed-to-be-legitimate names, instead of building a private 
version of such a database.  FWiW I would probably process in 
batches, versus process upon input, for many applications; the 
application was not described in the original post [which I did not 
reply to directly since that message does not appear on the group].

  From what I have read, there is an entire industry which does 
/data cleansing/ for names & addresses, and they may have [web] 
services available which use their proprietary databases and 
algorithms to make such decisions.  Probably address, GIS, or some 
other correlative information is required to accompany any names. 
Most utilization of such services is presumably for what is often 
called /deduplication/, but perhaps there is some manner in which 
they diagnose fake\frivolous information as data that should be 
purged.  With a legitimacy rating\weighting from an internal lookup, 
and possibly an internal software evaluation, better choices can be 
made [limit the amount of rows] for which data needs further human 
or paid cleansing action; noting of course the obvious loss of the 
ability to consistency check or de-duplication against any other 
presumed-valid rows in the database.

http://www.google.com/#hl=en&num=100&q="data+cleansing"+services
http://www.ibm.com/developerworks/webservices/library/ws-soa-infoserv3
http://www.google.com/#hl=en&num=100&q="data+cleansing"+"web+services";

Regards, Chuck