Re: How to do a "fuzzy" or approximate matching of strings in a SQL where clause

"Jeremy" <jeremy0505_at_gmail.com> wrote in message news:MPG.1dffabfd5556a3e298a058_at_news.individual.net...
> In article <1123532934.979790.255840_at_g49g2000cwa.googlegroups.com>,
> Barbara Boehmer says...
>> There are many options, most of which have already been mentioned by
>> others, which can be used individually, or in combination. You can use
>> built-in functions like LIKE and INSTR or you can use Oracle Text,
>> which requires creation of an index and periodic synchronization of the
>> index, but is an excellent combination of efficiency and flexibility,
>> with various search options and scoring that can be used for ordering
>> or further limiting. You can also use the Levenshtein distance (edit
>> distance) formula, which can also be used for ordering and will find
>> wrong spellings and things that the others do not, but tends to be slow
>> on large datasets. The Levenshtein distance formula is best used for
>> ordering and further limiting result sets after other methods have been
>> applied. What combination you choose depends on your exact needs. I
>> have provided a small sampling of some of these methods below for your
>> perusal.
>>
>
>
> <snip>
>
>
> What an excellent set of examples - filed for later use!
>
> --
>
> jeremy
> _______________________________________
> jeremy0505 at gmail.com
>

Oracle 10g also makes it possible to use regular expression in SQL with the built-in REGEXP function ...

Matthias Received on Tue Dec 06 2005 - 15:27:05 CST