Eric, The problem with one pass no restrictions, you will get errors as bad as a different state and zipcode.
John D. Haynes Director Geodata Consultants, Inc. 1-800-838-6661 ex.10 [EMAIL PROTECTED] www.geodataconsult.com ----- Original Message ----- From: "Eric Frost" <[EMAIL PROTECTED]> To: <[EMAIL PROTECTED]> Sent: Tuesday, February 18, 2003 10:30 AM Subject: MI-L MapMarker procedure > > Listers, > > I recently inherited a MapMarker project for which the internal documentation says to go through several passes starting at the strictest level and then slowly relaxing each of the Match Restrictions, Multiple Match, and Fallback restrictions respectively. However, I seem to get better results just allowing a single pass with all restrictions relaxed. > > >From manual inspection of 20 that the single pass method seemed to match better, 15 are indeed considerably better, whereas five are "false positives" for which the several pass method did a better job. > > Is there an "expert" procedure perhaps somewhere between the two that more or less consistently does the best job for geocoding? Are other people playing with doing several passes as opposed to a single pass with all restrictions relaxed? > > Thanks, > > Eric > --------------------------------------------------------------------- List hosting provided by Directions Magazine | www.directionsmag.com | To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED] Message number: 5556
