Please start any new threads on our new
site at https://forums.sqlteam.com. We've got lots of great SQL Server
experts to answer whatever question you can come up with.
| Author |
Topic |
|
AskSQLTeam
Ask SQLTeam Question
0 Posts |
 Posted - 2003-04-28 : 07:37:20
|
| Jason writes "I have an address field that, for the purpose of standardizing, has been populated consistently with the following:[Street Number]['-'][Apartment number/letter][' '][Street Name]i.e '123-A Main St'If no apartment number, the '-' is dropped as follows:i.e. '123 Main St'There is also a 24 character limit on the size of the field, which I have no control over as it is part of a 3rd party application. Therefore, some street names can get cut off.i.e. '123-321 Theodore Roosevelt Dr' becomes '123-321 Theodore Rooseve'I need a way to best-match our address to an OCR-read maybe-CASS-certified/maybe-not-CASS-certified utility invoice address. Looking at the previous example, I would be looking at matching up any or more of the following variations:123 Main St. Apt. A123 Main St Ste A123 Main St. #AI have already spoken with Rob Volk, and he said that '123' and 'A' would probably be dropped as "noise words" from the full-text index. This would rule out my initial thought of using the CONTAINS function.Do you have any ideas on how I can at least match up some of these addresses? Any we can match will save that much more data entry time.Thanks for your help!" |
|
|
|
|
|
Searching a field for the closest match