How to detect two similar words

Algorithm to match 'new york' and 'newyork'

[guarriman]

Hi.

I'm creating a program with Perl to know if word typed by user is within a DB table.

If I do "LIKE 'newyork'" in SQL it doesn't match "new york", so I'm interested in finding an algorithm which detects if two words are very similar.

Do you know it? Thank you very much.

[guarriman]

Found:
"Levenshtein Distance Algorithm"
[google.com...]

[rocknbil]

The soundex [dev.mysql.com] or sounds like mysql function may be of use to you.