Indexing a Local Newspaper Index on the Web (variation of topic)

[Kenneth R. Irwin]

Hi folks,

I came rather late to this thread and have a related question. It's a bit
beyond the scope of this list, but since we have 
several folks working on similar projects, I'd be interested in hearing
some off-list ideas about my quandry.

 I've been running working on converting a cardfile index of variable
quality (kept by a variety of librarians and non-librarians 
over several decades) into something resembling an index with a controlled
vocabulary and an authority file. This is for a 
college newspaper 1880-1990. About 10 years ago, a librarian started
creating a controlled vocab. and using a database
to index the paper, but so many hands have touched the index (even since
then) that what we have is an "official" controlled 
vocab, except it controls almost nothing because not everyone uses it. Does
anyone have ideas about pushing an 
uncontrolled system of index terms into the confines of a controlled vocab? 

(Also, has anyone found a way to speed the process of inputing old card
data? It seems like it might be faster to type than
to scan, for such a small amount of data.)

Thanks,
Ken, (whose index, for the record, is kept MS Access (pththh... not my
choice of DBs) and searched with a bad home-grown perl script. I'm hoping
to move it all to mySQL this fall)
Kenneth R. Irwin                                    kirwin at wittenberg.edu
Reference/Electronic Resources Librarian            (937) 327-7594
Thomas Library, Wittenberg University