• Want to receive the latest contracting news and advice straight to your inbox? Sign up to the ContractorUK newsletter here. Every sign up will also be entered into a draw to WIN £100 Amazon vouchers!

fuzzy matching

5.00 star(s)

Olly

Fingers like lightning
Joined
Jan 31, 2007
Messages
758
Likes received
0
Thanks received
0
Hi,
I've got list of people's names that I need to match against another list of peeps

Anyone know of any good routines for this kind of fuzzy matching?
Maybe you know of a table of common mistakes. Mac = Mc etc Mike = Michael ?

thanks very much :)
 
Hi,
I've got list of people's names that I need to match against another list of peeps

Anyone know of any good routines for this kind of fuzzy matching?
Maybe you know of a table of common mistakes. Mac = Mc etc Mike = Michael ?

thanks very much :)

I've used oracle's implementation of this before.

Soundex - Wikipedia, the free encyclopedia
 
oops forgot to add....this is running under MS Access
 
I've looked into this very heavily in the past.

Arguably, SMITH and MSITH are the same name, due to a typo. Or SMITH and WMITH

In my case, I used several algorithms to judge similarity between names and addresses, and weighted them to give a final score.

Here's a link you may find useful:
Approximate string matching - Wikipedia, the free encyclopedia

Also Google for N-Gram, Levenstein and Hamming.
 
right then...I'm getting there with this
I've built a small table from existing matched data of names and abbreviations
eg. William = Bill

I'd really like a full multicultural data set...any pointers please?
 

Advertisers

Back
Top