so i've been off working in the search group here at microsoft. it's been a few months and i'm definitely having a lot of fun working on some very cool stuff.
looks like folks have been having their fun at techEd and sql 2k5 is moving along.
just saying hi to the good folks over here...
An article I wrote on Yukon Full-Text Search has (finally!) been published on MSDN. Check it out:
http://msdn.microsoft.com/library/en-us/dnsql90/html/yukonftsearch.asp
I actually wrote the article a long time ago, and you would not believe how long it takes to run the whole thing through the wringer to get published.
I believe you will find what I wrote to be most satisfying. It covers:
- New FT features in Yukon
- A lay-person's overview of how search really works (this is probably helpful for non-db folks interested in search as well)
- Unadulterated, full-on 'disclosure' of our ranking algorithms
- Much more fun stuff...
Please read it and let me know what you think. As a side note, I should mention that I have taken on an exciting new role in Search in Microsoft, so I have moved on from the Yukon project. I can pass on your feedback to my successor, or you can just post it directly in the microsoft.public.sqlserver.fulltext newsgroup (I don't really get through there that often anymore).
Enjoy!
I've posted the third and final article in my series on testing Full-Text components using C# and COM interop. I hope you have enjoyed reading the articles as much as I have writing them. Happy new year!
http://sqljunkies.com/weblog/acencini/posts/716.aspx
I've posted part II of the series of articles on testing Full-Text components via COM interop and C#. This time, we look at how to test out a stemmer component. Check it out:
http://sqljunkies.com/weblog/acencini/posts/616.aspx
As I promised in a recent blog entry, I will be posting some code up here that shows you how to test out the various components that are used by SQL Server Full-Text Search -- in particular, wordbreakers, stemmers and filters.
Rather than cover all the ground at once, I've split the story into a three-part series (mostly because a one-part series would be enormous) -- the first part, looking at how wordbreakers work, will provide you with a tutorial on how to wrap the right stuff in C# so you can test wordbreaker output. More can be found here (I think this is the right link): http://sqljunkies.com/weblog/acencini/posts/595.aspx
I'll try and get part II posted later this week or early next week. Stay tuned. As always, questions and comments are quite welcome.
Well it has been a while since I have updated this blog, and I just got a few seconds free so I thought I would post an update.
It is good to see Michael Rys posting up here as well -- if you think he is prolific on the blog, imagine working with him ;-)
So I got back from a nice restful vacation to Mexico where I did very little at all. A nice change of pace. A short while ago I finished writing a white paper on Yukon Full-Text Search (per my last post, “Yukon Full-Text Secrets Revealed”) -- it should be up on the web hopefully before the new year but you wouldn't believe how much goes into the process.
Today it is very windy in the Seattle area -- I wouldn't be surprised if the power goes out here on the Microsoft campus... it seems like it has been a while since that has happened. I like these windy days, though... quite relaxing -- though I hear most of the surrounding areas are currently without power.
Anyway, if I can figure out a good way to do it, I've got some nice C# code that can be used to test stemmers and wordbreakers for Full-Text Search. There are many folks out there who really would like a good way to understand what is really going on during Full-Text index/query when their data is being tokenized / stemmed. Unfortunately when I posted the stemming bit to the microsoft.public.sqlserver.fulltext newsgroup, the formatting got eaten for lunch. If I get a few good seconds I'll do two installments which are also kind of a nice tutorial on COM interop with C# as well as implementing a callback between unmanaged and managed code. Stay tuned.
Just kidding... I can't tell about Yukon Full-Text Search just yet ... ;-)
If you're interested in learning more about Full-Text Search in SQL Server 2000, though, I thought I would share a couple of my favorite links (well, favorite because I wrote one of them):
There are a lot of great uses out there for Full-Text Search -- I'm always interested in hearing how you are working with Full-Text in your applications, and hope the articles above are of use to you.
Enjoy,
--andrew
Thanks to Donny for setting me up with this Blog account...
To get started, a bit about me:
I'm from the SQL Server development team -- I'm a Program Manager in the SQL Engine and am responsible for Full-Text Search among other things. I've been at Microsoft for a little more than 2 years now ... grew up in Boston, MA and went to school in NY. Aside from spending a lot of time with SQL Server, my other interests include goofing around a lot with C#, travel, and hockey. I actually used to have a Blog of my own a few years ago (when I had more time on my hands)... in those days I actually wrote back end code myself -- quite a lot of fun.
Depending on when it is ok to share Yukon stuff, I'll definitely be talking about Full-Text, and also other random tidbits you may or may not find interesting. I'm looking forward to hearing more from folks out there and you can reach me at acencini@ microsoft.com or you can hit me up on my personal email at andrew@ cencini.com (remove the space between the @ and domain)... See you later...
--andrew