Data Modelling Principles

william burkett

Data Modelling Principles
There are a ton of data modelling books out there that provide wide ranging rules, principles, and guidance for creating data models. Do you have a favorite data modelling book? Or do you have a set of rules/principles/guidelines that you use in developing data models? For me, the best work on "data modelling" isn't even about data modelling. It's an ISO standard on terminology: ISO 704. https://en.wikipedia.org/wiki/ISO_704

Eric Schiller

RE: Data Modelling Principles
(in response to william burkett)

I'm still partial after all this time to the old school Kimball book, even if I'm not working directly out of the dimensional modeling world as often now.

That said, I'm VERY interested in Fact-Oriented Modeling after seeing Marco Wobben's presentation at EDW.  There's a book  and also this website to cover it -- http://www.factbasedmodeling.org/.

Edited By:
Eric Schiller[All DATAVERSITY Members] @ Mar 26, 2019 - 07:01 AM (America/Pacific)

Matthew David

RE: Data Modelling Principles
(in response to william burkett)

I agree Kimball is pretty good but it the book is quite thick.
There's an EdX course on data modeling which is ok but is a bit slow
I started the new Data Engineering course from Udacity and think it is pretty good. A lot of good hands on examples.

Eric Schiller

RE: Data Modelling Principles
(in response to Matthew David)

Yeah, I found that the first five or so chapters of Kimball are the really essential stuff, then most of the rest of the book is templates for specific situations until you get to the ETL stuff and a few other things at the end.

That saves a LOT of reading up front so that you can go back and reference later as needed.

Definitely worth looking into other stuff like canonical modeling, and some other more general principles as well to keep up with the agile world.

Charles Harbour

RE: Data Modelling Principles
(in response to Eric Schiller)

It's been my experience that form almost always follows function.  What I mean by that is that there are commonalities across all healthcare databases (operational and analytic), across supply chain, across service model.  So, your basic structures would be the same across all implementations within the same industry (with your own customizations, of course).  

 

For example, there is progress being made to get all of the Blue Cross Blue Shield network participants to use a common storage format, as well as a common communication format, to promote automatic electronic data handling.  You might think that dragon would have already been slayed, but it has not - there are a large number of translation/adapter fittings in place to keep the data flowing, with unsophisticated validation and a lot of manual processes.  But the more we can work together to share the details, to define that industry standard, the easier it will be to put automated processing rules in place and to increase the overall data quality.

 

So, depending on your needs, you should review (and maybe experiment) for what works best with your business model - and then decide the best fit for your company.

Mark Matten

RE: Data Modelling Principles
(in response to william burkett)

My favourite books are Len Silverston's Data Model Resource Book series due to the insight into business domains and the patterns provided to rapidly getting a modelling project started. I'm currently working my way through The Nimble Elephant too (John Giles is another data modelling hero).

Graeme Simsion's book also must-reads.

Ralph Kimball's Data Warehouse Tookit is a classic (I couldn't say the same for Building the Data Warehouse though).

Aaron Fuller

RE: Data Modelling Principles
(in response to william burkett)

The data modeling book I use the most is Chris Adamson's Star Schema: The Complete Reference. I design dimensional databases a lot and it gives me a great tool to quickly apply the best design practices to my particular situations. I also like Len Silverston's books.

Aaron Fuller, CBIP

Founder & Principal Consultant

Superior Data Strategies 

517-803-0714

superiordatastrategies.com

Lawrence Witner

RE: Data Modelling Principles
(in response to william burkett)

Hi William,

For beginners I recommend "Data Modeling: A Beginner's Guide" by Andy Oppel.   For MongoDB I recommend the publicly accessible Mongo Manual online.

William McKnight

RE: Data Modelling Principles
(in response to william burkett)

I like Data Modeling Made Simple by Steve Hoberman and Data Modeling for the Business by Hoberman, Burbank and Bradley.

Phil Forestall

RE: Data Modelling Principles
(in response to Mark Matten)

I like your choices, Mark. For full-time modelers Simsion's 'Data Modeling Essentials' (get the third edition) is a great job manual for data modeling. Silverston's three-volume compendium has been a very useful lookup reference to me too. Kimball's Toolkit is the warehousing breakthrough classic.

I spent a lot of time on Chris Date's 'Introduction to Database Systems', rigorous but a tough read. William Kent's little book 'Data and Reality' earns every rave it gets. Have a look at these too.

Cheers

Vladimir Sosiurko, CDMP

RE: Data Modelling Principles
(in response to William McKnight)

Definitely Steve Hoberman is what you can start right from the beginning of your career and get it to the next few levels

Gary Jordan

Data Modelling Principles
(in response to William McKnight)
I agree. Hobermans books provide a good solid foundation for Data Modeling.

I just finished David Hay's new book, *Achieving Buzzword Compliance. *It
gives some good background on Data Architecture and Data Modeling terms and
concepts. It has really helped me fill in some of the gaps in my
understanding of these areas.

Gary Jordan

Data Architecture

Red Hat NC https://www.redhat.com

[login to unmask email]
M: 9192103078
https://red.ht/sig





On Fri, Jul 19, 2019 at 11:05 AM William McKnight <
[login to unmask email]> wrote:

> I like Data Modeling Made Simple by Steve Hoberman and Data Modeling for
> the Business by Hoberman, Burbank and Bradley.
>
> -----End Original Message-----
>

Jose Mari Taleno

RE: Data Modelling Principles
(in response to Vladimir Sosiurko, CDMP)

Kimball Data Model is still applicable for me. I'm keen on learning Data Vault Modelling however, the investment to start right with it is so high that makes it hard to get support from the business.