st: matching strings on words

[Jeph Herrin <junk@spandrel.net>]

I'm not sure what to call this - if I did, I might have
better luck with my searches for a utility. Basically,
I want to do something similar to the utility -nmatch-
which matches first and last names, but I have more than
two words per record.

The problem: I have two files with lists of hospital names.
Hospital names tend to consist of multiple words, that get
used to different extent; the same hospital might be listed
as:

 st joseph's
 st joseph's memorial
 st joseph's memorial hospital
 st joseph's memorial hospital of danbury

etc. (There is also a lot variation on eg "Saint vs "St." and
"Memorial" vs "memorial", but I have trapped most of those
already.)

What I'd like to do is match these on "words", and generate
a _merge variable which indicates how many words match vs
how many words there are. Then I (or some unlucky grad student)
can trawl through the matches and decide which ones are the
same hospital.

I can see how to write a program to do such a thing, but am hoping
there is already a solution out there that I overlooked?

thanks,
Jeph

*
*   For searches and help try:
*   http://www.stata.com/help.cgi?search
*   http://www.stata.com/support/statalist/faq
*   http://www.ats.ucla.edu/stat/stata/