Viewing 15 posts - 16 through 30 (of 38 total)
@jobudzin, I love Confio Ignite, but I quickly realized the pain of naming many SQL statements legibly... and then re-naming them, without duplication, when they were recompiled. A...
January 13, 2014 at 2:38 pm
One reason for doing this instead of built-in Full-Text Search, is that once you have isolated your keywords, you can suggest alternatives when a user has misspelled one. See my...
September 25, 2012 at 4:29 pm
@rajesh-2 R-424835: Thanks for saying it's awesome, although you had an issue! The only limitation on size of column being searched, is that only NVARCHAR(253) is passed into...
September 25, 2012 at 4:28 pm
@amy.G I believe you need a database of first names; because the most an algorithm could do for you is to normalize the names. I googled "infer gender...
September 25, 2012 at 4:14 pm
@martin-2 Bastable @fmendes Yes, the algorithms I've heard of most in a SQL context (using extended stored procedures) include Jaro-Winkler[/url] and Metaphone[/url] so I've linked those...
September 23, 2012 at 4:59 pm
@all: Scripts updated 2012-09-21 from v1.0.0 to v1.1.0
Performance of [udfDamerauLevenshteinOpHTML] has been improved, by doing one less INSERT and one less SELECT on the table variable. (The values of...
September 21, 2012 at 5:36 pm
@dmusick: That (replacing a table variable with string variables) is essentially what I've done in my Lim UDF. Since I only need to keep three rows, I only...
September 19, 2012 at 6:11 pm
@ChrisM@Work: Interesting approaches, but I'm wondering why you use DATALENGTH instead of LEN, are you trying to include trailing spaces? (In order to work with NVARCHAR instead of...
September 19, 2012 at 9:32 am
@dmusick: Interesting approach, and certainly useful. Note that DLD shows the same similarity for 'farm' and 'charm' (40% diff = 60% sim). For performance, I would try to...
September 18, 2012 at 11:12 am
@martin-2 Bastable: Thanks for your kind words, and thanks to those who proofread various drafts of my article, including Kellie Sanchez, Dan Hess, Eric Freeman, Ernest Cook, and...
September 18, 2012 at 3:42 am
@Davin21: Indeed, I had originally linked the very informative SimMetrics Web Site, which was linked from your sourceforge link, but it seems to be permanently down now:(
September 18, 2012 at 3:31 am
@Davin21 I have some non-T-SQL implementations (including a Jaro-Winkler) listed on my Reference[/url] page - please let me know of others I should consider linking to.
September 18, 2012 at 1:26 am
@ZA_Crafty: Perhaps CLR could be faster, but a Transact-SQL-only solution was important to me and to my clients.
September 18, 2012 at 1:14 am
Nice summary! Of course we all know that the database WILL be expected to parse pieces of data that are allowed to be combined in one column. (Just witness the...
June 30, 2011 at 12:43 am
Rather than changing the server default collation just to avoid the issue on the next database creation, how about changing the collation in the model database?
April 11, 2011 at 6:54 pm
Viewing 15 posts - 16 through 30 (of 38 total)