I don't see any difference between knowledge discovery and Data Mining.
The diference between knowledge discovery and knowledge extraction (my opinion) is that the first is not driven by the user, meanwhile the second yes.
Other reason for using this surrogates keys.
If the operational changes its ids, the facts tables in the DWH is not affected. May be you only have to change in the dimension tables the id's traduction for the operational to the multdimensional. But you don't have to update the fact tables...
This can help you anyway when you are using large dimensions with more than 64k members.
You can make a hierarchy over the same 'id'.
Refer to the Grouping property in Analysis Services, that should be automatic or refer to the book OLAP Train Reed Jacobson page 110.
The order is different of the hierarchize.
Define the hirarchize for instance Year-Quarter-Month-Week-Day, and forget the number id of your primary key
regards
This site uses cookies to help personalise content, tailor your experience and to keep you logged in if you register.
By continuing to use this site, you are consenting to our use of cookies.