My Photo

Your email address:

Powered by FeedBlitz

April 2018

Sun Mon Tue Wed Thu Fri Sat
1 2 3 4 5 6 7
8 9 10 11 12 13 14
15 16 17 18 19 20 21
22 23 24 25 26 27 28
29 30          
Blog powered by Typepad

Become a Fan

« Data Mining, Predicate Triage and NSA Domestic Surveillance | Main | What’s In A Name? »

March 14, 2006


Feed You can follow this conversation by subscribing to the comment feed for this post.

Tanton Gibbs

I used to work for a data recognition company. I called this concept "Quantum Recognition". Basically, a person existed in a number of different states until you provided enough information to determine what "version" of the consumer you wanted.
Good to see someone else thinking along the same lines.

Alex Simonelis

It would be more accurate to say that these particular entities have n attributes - like an n-dimensional vector - and that truth is not in the eye of the beholder (the values for each dimension are fixed at any one time), but rather each user defines what is important to him (his "truth") as one (or multiple), but not all, of the dimensions.


Why is the industry always pushing one version of the truth.

I agree "truth is in the eye of the beholder".

We could not agree an a recent MDM effort and scrapped it. But a seperate app to app integration project I was on adopted an open-spec approach we found on sourceforge called Jumper metamodel. This gave us the best of MDM with much needed flexibility.

Bruce Wallace

great minds...I had a series of brainstorms in early 2006 that paralleled this discussion...

It has led me to undertake a study of the Philosophy of Identity, and to attempt to put into a book for general software developers the 2500 year old conversation that Philosophers have had on this topic. [Thesis: Because their bread-and-butter activity involves modeling the world, ALL software developers need to know about Philosophy/Metaphysics rather than only exotic post-graduate researchers.]

Euan Cochrane

It seems to me that this issue comes down to the use of terms to refer to concepts.
The terms "name" and "address" can refer to many different concepts. If these concepts were properly identified and differentiated from each other and instances of these concepts identified and differentiated from each other then the conflicts described in the post would not happen.

We need to build systems that differentiate between concepts, not just terms, and have systems for documenting concepts that are as unambiguous as possible.
I find the guidance given by the Australian Institute of Health and Welfare's data modelling team to be quite useful for providing some rigor and rules for this: pages 19-23

The comments to this entry are closed.