How to apply Data Mining (association rule) to a huge database?

Question

What I want to do is apply the data federation method in my SQL Server 2000 database. An association rule is something like "searching for the most common elements that appear together in the database."

For those who don’t know or want to remember what an association method is, take a look at this presentation about the association rule in Data Mining.

http://www.authorstream.com/Presentation/sushiltry-108428-association-rules-data-mining-science-technology-ppt-powerpoint/

The 17th slide provides a good example of applying an association rule to a database.

So can you help me on how to write SQL codes (if that is enough, of course)

Thank.

Edit: I adjusted the link.

Answer 1

You need to implement the apriori algorithm . Check link-1 as well as link-2 .

Answer 2

You need to use Analysis Services MS SQL Server. Analysis Services has a Microsoft association algorithm, which is an algorithm for developing join rules. You need to provide a data set, support and trust, and the Analysis Service will take care of the rest. If the datset is very large, you might consider setting the size of the template.

http://technet.microsoft.com/en-us/library/ms174916.aspx http://technet.microsoft.com/en-us/library/ms167047.aspx

(Apriori, FP-Growth, Eclat ..)

Answer 3

Java- Apriori, FPGrowth, Eclat .., : http://www.philippe-fournier-viger.com/spmf/

SQL, .

, , SQL- . , , .