Written by William Kent, updated by Steve Hoberman
“Data and Reality” is described on the cover as “timeless”. The book was written in 1978, and I started in IT a few years later. At my first job, I think they understood the issues described in this book, and realised how to deal with them, or at least attempted to deal with them. It was only when I moved on that I realised that organisation was exceptional by the standards of the time. Fast forward to 2012, and how much has changed? Surely organisations now understand about `data and reality’, don’t they? Nope – it’s as if time has stood still in some ways, so this book is still amazingly relevant.
William Kent takes a philosophical view of data, and manages to describe data modelling without you realising that’s what he’s doing! He asks whether data modelling is adequate as a formal modelling system. Well, the focus of data modelling is traditionally well structured data, which lends itself well to a structured description, but it is filtered and incomplete. That’s because structured data and information systems that contain it are themselves a possibly ambiguous representation of reality. As Steve Hoberman says in his commentary (quoting Douglas Adams) “I think the problem, to be honest with you, is that you’ve never actually known what the question is.” Remember that, the next time you’re asked why you need a data model, when all you’re doing is implementing an expensive ERP package.
If the creators of today’s legacy systems had read this book, perhaps they would have understood the importance of data modelling, and I wouldn’t be spending my time trying to understand the data in their applications.
<openness statement> I co-authored a book with Steve Hoberman in 2011 </openness statement>
June 12th, 2012
Ronald G. Ross and Gladys S.W. Lam
This book is targeted at Business Analysts, but I would definitely include it in the must-read book list for any data modeller, especially if they need to produce a “conceptual data model” or “business data model”. A “conceptual” or “business” data model is about real-world concepts, using terms that business people would naturally use – it is very close in concept to the ‘Fact Model’ described in chapter 9. The Fact Model employs terms (nouns or noun phrases) and noun-and-verb constructions (such as “a customer must place at least one order”) – these constructions transform into business rules, and can often be reproduced directly in a data model. You may be able to configure your data modelling tools to use the notation shown in this book; if you can, that’s a real help.
Data modellers should also check out chapter 10, where the authors discuss business milestones, and how they represent changes in state (e.g. “Order shipped”); such state changes are very important to data modellers.
This book is a great illustration of the crossover between the roles of Business Analyst and Data Modeller – they are both concerned with placing the business rules and vocabulary at the fingertips of business people, business analysts, and anyone else interested; we must avoid burying the rules and vocabulary in software requirements.
Like another reviewer, I quibble with the notion that “business rules spring to life only after the solution has been deployed” – the business rules are necessary prerequisites for an effective solution. Notwithstanding that single point, this book should be required reading for Business Analysts, Programme and Project Managers, and anyone with responsibility for improving the way the business operates.
January 3rd, 2012
I know a data modelling trainer who has given hundreds of copies of the first edition of this book to those attending his courses. He did this because it was a great supplement to the material he covered on his training course. Is this second edition a worthy successor?
The author has expanded and restructured the book for this edition; it has grown considerably in size, from 134 to 360 pages. Additional topics have been added, partly based upon the presentations the author makes at seminars and conferences. Some of this additional material has been provided by experts in the respective fields – Bill Inmon, Michael Blaha and Graeme Simsion. This extra material doesn’t come for free – the list price has increased two-fold.
This book is a well-scoped and well-written introduction to data modelling and related topics. The author’s friendly presentation style really comes across in the text, avoiding the temptation to use geek-speak to impress the reader.
There is no CD included, but there is interaction of a kind, provided by 15 exercises to test your understanding.
I think all the material is great, but I have one minor gripe. In the first edition, there’s a separate chapter on the importance of good definitions, describing the characteristics of good definitions (clarity, completeness and accuracy). While it is covered in the new book (page 102), it doesn’t have the same punch as the original version. Getting people to put the effort into creating good definitions is one of the key challenges with data modelling, so I’d prefer more emphasis.
Should you buy this book?
* If you already have the 1st edition, then you buy the second edition for the extra material; don’t throw away the first edition, use it as part of your internal marketing process, by giving it to someone who wants to know what data modelling is about. They’ll thank you for it.
* If you’re a business or IT person and need to understand more about ‘doing’ data modelling, buy this book. The first 100 or so pages should be compulsory reading for anyone who has to deal with data models or data modellers.
* If you want to understand more about data modelling without getting into details of ‘doing it’, consider Data Modeling for the Business: A Handbook for Aligning the Business with IT Using High-Level Data Models (Take It With You), which Steve Hoberman co-authored.
Please note – I wrote this review several months before I was asked to co-author “Data Modeling Made Simple with PowerDesigner”.
December 15th, 2010