Posted by Reilly Atkinson on August 15, 1999 at 22:08:42:

In Reply to: Some thoughts about Data Warehousing posted by Martyn R Jones on August 15, 1999 at 17:22:59:

Martyn -- Applause, kudos, and congrats; your extracts on the role of the user and the efficacy of an iterative approach are right on. I say this for two reasons; first, I have been operating this way since my graduate student days, and have had consistently positive results both as a physicist and as a business analyst/consultant. The same is true for many colleagues.

The reason, of course, is that you describe major rules of scientific research -- in most general terms, the users are, say, the physics community, or, perhaps an experimental team for whom you are doing computations. As one clever scientist put it, most researchers spend most of their time being wrong, so you have to keep trying, and learn from your mistakes.

As I see it, you describe some of the "best practices" of the scientific community - developed over centuries of effort. And, this is exactly how Einstein worked (although he was not always in touch with his users as much as he might have been). Clearly, then, no way do these best practices inhibit creativity. Rather, once in play your methodological approaches greatly enhance the posibility of the emergance of creativity -- I've seen it happen many times

What is new is the successful selling of this approach to the business community -- 25 years ago few were buying. Although I have my own notions, I'm curious about the causes of the explosion of interest in data warehouses, data mining, knowledge management and the like. It's not like any of these things are new; they have all been around for centuries.

Also, what I don't see is any mention of how the users are going to get their desired information from the nicely organized data. Who is going to tell you whether OLAP results are reliable, or whether some automated neural network data mining package is giving you sense or nonsense? In my humble opinion, an experienced, sophisticated data analyst ought to be part of the data warehouse development team from the start. Is this often the case?

Thanks for your clear thinking and writing.
Reilly

---

Follow Ups:

• DW & KM Denham 12:53:40 8/16/99 (1)
  • Re: What's Next? Kees de Vos 15:24:04 8/16/99 (0)

---

Click Here to Post Follow Up in New Forums

    Knowledge Management Think Tank (New)

  

 Subject: 

Message: