Updates to Recent Articles

Just a couple of updates. If you found the article last week on Oracle OLAP memory management useful, Chris Chiappa sent an email through to me this evening with some clarifications and corrections. I've annotated the article with Chris' comments and you can read the updated version here. Thanks very much Chris, much appreciated.

If you've been following the postings on mine and Abhinav Agarwal's blog about Siebel Analytics and the future of Oracle's BI products you might be interested to note that there's a couple of very interesting presentations being given by Oracle at the ODTUG conference next June.

"Siebel Business Analytics Platform: Next-Generation Business Intelligence Architecture" by Aydin Gencler (one of Oracle's business intelligence product managers) has the abstract "The Siebel Business Analytics platform is a next-generation business intelligence solution that helps you integrate and analyze data from multiple enterprise sources and turn it into actionable insight for thousands of users." whilst Christina Kolotouros, another Oracle product manager, is presenting "Roadmap to Oracle's Next Generation Business Analytics with Oracle Fusion" which has the abstract "The next generation of Oracle business intelligence from ad hoc query reporting, dashboarding, alerting, and analytic workflow capabilities is a truly integrated suite on the BI platform. This presentation will explain the roadmap of the new Oracle Business Intelligence Suite" I've had a couple of papers accepted for the OTDUG conference and so therefore I'll be able to see and hear firsthand what's coming up for Oracle's BI & reporting products. Should be very interesting. George Lumpkin is doing the "Oracle Data Warehousing Architectures" presentation I missed last Open World but blogged about subsequently, I'll also be very interested to go along to that one.

Marcos Campos, a former Thinking Machines Senior Scientist who now works as a Development Manager for Oracle Data Mining, has started a new blog. There's a particularly interesting article on comparing how time-series forecasting is done using the OLAP Option (and Express Server before that) and with the new Support Vector Machine feature in the Data Mining Option. I won't pretend to understand half of it (data mining isn't my forte) but it's an excellent article and so far an excellent blog.

Finally, I've just got back from a customer engagement where we helped to implement Oracle Business Rules as a customer front-end for defining an account adjustment process - see my previous posting from last weekend where I first looked at the product. I'm pleased to say that between us we managed to get the product up and running, and integrated in with their ETL process, with a Java program calling the Oracle Rules Engine, which contained some "Rule Sets" defined earlier that updated or amended financial figures according to some rules set down by the business. It's not an entirely straightforward process - you have to build an Java API over the tables you wish to adjust, together with another Java program that calls the rules engine and runs the rules (this white paper explains the process better), but it works and it meets the customer's requirements that the business can define what is in effect the ETL routine. I won't mention who the customer is as we've got an NDA but if they're reading this, you've got a pretty nifty Java programmer and DBA there, well done. Hopefully work with you again soon.