Interesting that you mentioned InterMeida Text. I would not think it
can solve your problem, unless your post did not tell the full story
of your situation. ORACLE text does have the fuzzy search capability -
by using the "fuzzy" operator. But I don't think that's what you need.
You actually want to compare rows and identify patterns, this is not
the same as fuzzy text search.
"Giovanni Azua" <bravegag_at_hotmail.com> wrote in message news:<ank4uu$f0r4s$1_at_ID-114658.news.dfncis.de>...
> Hello all,
>
> I am currently facing a problem that perhaps
> one of you could help me to solve out sooner. The
> problem, apparently simple, is that there are reapeated
> entries in tables in a very big existing RDB and I am required
> to find such repeated records, there are e.g. duplicated addresses
> with typo errors and several differences which would make fail
> an exact LIKE or INSTR operators, those textual fields are in
> english and german (swiss german) is there any approach I
> could use to achieve an inexact scored matching? Would that work
> with German as well? e.g. IR stemming algorithms are obviously
> specifics for the target language.
>
> Any help would be greatly appreciated,
>
> TIA,
> Best Regards,
> Giovanni
Received on Fri Oct 04 2002 - 19:10:40 CDT
