Oracle FAQ Your Portal to the Oracle Knowledge Grid
HOME | ASK QUESTION | ADD INFO | SEARCH | E-MAIL US
 

Home -> Community -> Usenet -> comp.databases.theory -> Re: Help with our Search Engine

Re: Help with our Search Engine

From: Trevor Jenkins <Trevor.Jenkins_at_suneidesis.com>
Date: 8 Feb 2003 18:54:11 -0600
Message-ID: <slrnb4b1hi.6st.Trevor.Jenkins@suneidesis.com>


On 8 Feb 2003 09:40:13 -0600, Javiermontebruno <javiermontebruno_at_yahoo.com> wrote:

> I don't know if someone know a system that would work better than SQL
> server full-text search and that would enable our engine to ouput
> results faster.

I suspect that SQL Server (as most relational systems seem to do) uses some form of B-tree for its indices. As we all known from Data Structures 101 hash tables are faster than B-trees O(1) compared to O(log n). So replacing SQL Server with a system that uses hash tables for its indices will probably give you faster response.

Alternatively changing from a relational database system to a text retrieval system would probably give you better response too.

If you can find the one text retrieval system that uses hash tables for its indices then you'll be onto a winner. Finding it is an exercise for the reader this being cross-posted to comp.theory.info-retrieval after all. :-) And yes this system does run under Windows NT/2000.

Of course, some of the poor response could be caused by selection of the wrong algorithm or poor programming. Whilst the one system mentioned above will help you it can't make up for human error.

Regards, Trevor

British Sign Language is not inarticulate handwaving; it's a living language. Support the campaign for formal recognition by the British government now! Details at http://www.fdp.org.uk/

-- 

<>< Re: deemed!
Received on Sat Feb 08 2003 - 18:54:11 CST

Original text of this message

HOME | ASK QUESTION | ADD INFO | SEARCH | E-MAIL US