Re: rank algoritme - best bet search

Hi Jack,

yes it does seem that I need to look at data mining. I didn't quite know what the difference between data mining and OLAP was... still don't ;)

But those articles you pointed me towards seem to be what I am looking for. From the brief introduction it looks like that is what I am trying to do. Hmmm... he is talking about 100 elements and that would need 10 to the power of 20 GB worth of storage if done in the way I thought of at first. I have 100.000 elements... not really feasable ;)

Thanks for pointing me in the right direction. I need to do a lot of reading I can tell. Do you know of any good books on data mining?

Pete!

"jack_dull" <jack_dull_at_163.com> wrote in message news:<bbtt4f$1mkl$1_at_mail.cn99.com>...
> Hi pete,
>
> I'm not sure if your question is really about "rank" that can be handled
> efficiently by OLAP. SQL SERVER 2000 has no SQL/OLAP extenstion. But
> Microsoft Analysis service is a both OLAP and dataming product. There are
> plenty of document and tutorials about rank functions that you can search in
> the google with the key words "rank SQL OLAP".
>
> It seems that your problem is related to association rule that is part of
> dataming. There is a brief introduction at
> http://dsl.serc.iisc.ernet.in/~vikram/mining.html
> Another General Survey about "Algorithm for Association rule Mining" is
> http://www.acm.org/sigs/sigkdd/explorations/issue2-1/hipp.pdf.
> microsoft.public.sqlserver.datamining is a news group about datamining.
>
> It seems that your question can be divided by two different parts: Firstly,
> how to find associate rules. Secondly, how to use those associate rules.
>
> I hope those information can be any help for you.
>
> Best regards,
>
> Jack
Received on Sun Jun 08 2003 - 17:54:27 CEST