Alternatives to SQL FTS.

Thread view:

Enable EMail Alerts

Start New Thread

Thread rating:

slartyb - 25 Apr 2005 15:19 GMT

I've been using FTS for 5 years now on 1.5million rows of data. Certain
things have alway annoyed me about it however:

1. I never managed to succesfully move/copy ft catalogs.
2. Noise words/symbols/single characters e.g. c#, c#, "I am not".
3. Search speed and ranking/relevance quality.
3. Population performance.

I know by constantly staying on top of the excellent advice in this forum I
can probably overcome these issues, but frankly I'm not sure I want to take
this route. Not at least before I've explored the alternatives. So my
question is what alternative products would anyone recommend taking a look at
? The ft queries would be made from a C#/.Net/ASP/NET app, and I don't mind
if the text records are stored in a database or maybe an XML file.

I haven't looked at Yukon yet - so that may well be a valid answer.

Any takers ?

Reply to this Message

Hilary Cotter - 25 Apr 2005 16:07 GMT

I would have a look at Yukon for the following reasons

1) you can now a) detatch and reattach full text catalogs with the database
files
b) your full text catalogs can be backed up with
your back and restored. This is the default behavior
c) you can replicate full text indexes as long as
your subscriber database is full text enabled
2) by default c and j are indexed if they are preceeded by a ++ or a #. SQL
2005 (Yukon) can be made to be noise insensitive - ie your queries will not
break with a noise word
3) Search preformance is greatly improved. The ranking algorithm yeilds
different results but is much the same.
4) radical improvements in indexing performance.

There are other options out there - like SQL Turbo
(http://www.imceda.com/SQLTurbo_Description.htm) and omnidex
(http://www.disc.com/).

Signature

Hilary Cotter
Looking for a SQL Server replication book?
http://www.nwsu.com/0974973602.html

Looking for a FAQ on Indexing Services/SQL FTS
http://www.indexserverfaq.com

> I've been using FTS for 5 years now on 1.5million rows of data. Certain
> things have alway annoyed me about it however:
[quoted text clipped - 14 lines]
>
> Any takers ?

Reply to this Message