Post reply

New to full text search and need some pointers

  • June 17, 2008 at 4:43 pm

    #183581

    I have a project where I need to try to match and address file to a master address file and create a list of possible matches. I can create a list using the weighting feature of the ContainsTable function to match on different pieces of a company name, but how would I check for possible versions of personal names like Bill and William? Also where are there good examples of different "real life" situations that have use this?

    Thanks in advance

    Kelly

  • June 18, 2008 at 5:52 am

    #830323

    You may want to reconsider using FullText for this. A fuzzy lookup in SSIS would probably be more appropriate. It is designed to find "close" matches based on criteria you can configure.

    Leaving both of those options behind, lots of address verification web services have popped up over the past couple of years like this one:

    http://www.intelligentsearch.com/address_verification/verify_address.html#xml

    I have never used this one, but most address verification web services will return the corrected USPS version of an address after verification. If you verified two addresses that were the same but had different attention lines in your database, the corrected versions would match exactly.

    Address matching can be a bit complicated, for a couple hundred dollars you can sign up for a service created by people that do nothing else. It will probably be cheaper, easier, and more accurate.

  • June 19, 2008 at 7:22 am

    #831084

    I agree, there's plenty of companies that will do this for you.

    But if this is something you're going to have to do more of, then I'd advise you to make a case for spending more, and invest in some matching software. They've got built in lists of all the abbreviations and variations that can occur in business names (Limited = Ltd), addresses (Street = St) and personal names (William = Bill) and use all those fancy fuzzy and phonetic searching techniques.

    I use one called DedupeExpress:

    http://www.dqglobal.com/deduplication_software_dedupeexpress.html

    and it's great. You get a free training session with it too, and their customer service is top notch. And no, I don't work for them! But seriously, there are plenty around, and well worth an hour or two of internet research.

Viewing 3 posts - 1 through 2 (of 2 total)

You must be logged in to reply to this topic. Login to reply