Book Review: Data Modeling Made Simple – 2nd Ed

In this book review, I address each section in the book and provide what I found most valuable as a data modeler. I compare, as I go, how the book’s structure eases the new data modeler into the subject much like an instructor might ease a beginning swimmer into the pool.

Data Modeling Made Simple – Second Edition
Reviewed by Johnny R. Gay
Published: Fall of 2009
Author: Steve Hoberman
Publisher: Technics Publications, Second Edition 2009
ISBN 978-1-9355042-7-6

This book begins like a Dan Brown novel. It even starts out with the protagonist, our favorite data modeler, lost on a dark road somewhere in France. In this case, what saves him isn’t a cipher, but of all things, something that’s very much like a data model in the form of a map! The author deems they are both way-finding tools.

The chapters in the book are divided into 5 sections. The chapters in each section end with an exercise and a list of the key points covered to reinforce what you’ve learned. I find myself comparing the teaching structure of the book to the way most of us learn to swim.

SECTION I: Data Modeling IntroductionThe first section is like the shallow end of the pool, where as a beginning swimmer, you can dip your toes in to test the water. These easy chapters are short and concise. Here the author uses very common objects to describe what a data model is, and why it is so valuable. His first examples made excellent use of what’s truly a universal data model to millions of computer users – in school and business – the spreadsheet. 

SECTION II: Data Model ComponentsIn the second section, Steve Hoberman introduces you to the simplest components that make up a data model, and explains the important terms that we apply when we discuss them. By the end of section 2, you now have both feet comfortably in the water. You’re ready and eager to plunge deeper into the depths of this pool of data model knowledge.

SECTION III: Subject Area, Logical, and Physical Data ModelsYou’ve made it to the deep end of the pool where you get a real workout as you lap through the 3 levels of data models: subject area (or conceptual), logical, and physical. Just as there are different strokes for different folks, there are different models for different audiences. By the end of section 3, you’ll be able to swim through the intricacies of a data model like a barracuda, and you’ll know more than many working data modelers do today. Calling one’s self a data modeler and being a data modeler are two very different things indeed.

SECTION IV: Data Modeling QualityJust as swimmers can kick-start their movement through the water with the use of swimming aids (maybe a flotation device or fins will help), you can utilize Steve’s 4 favorite templates to collect and organize the requirements that will define your data model. They allow you easily wade through the river of requirements you’ll collect. You may recall the scorecard the Olympic judges use to rate a dive. Steve introduces his Data Model Scorecard, which applies a quality rating to a data model.  It’s an objective look at the quality of the model built. I too have seen how a number of quick and dirty data models eventually led to the untimely death of the databases for which they were created. I’m convinced that applying the scorecard helps us deliver a higher quality data model resulting in a higher quality database that will live a much longer and healthier life. We are actually adopting this tool where I work, after applying our own weightings to his 10 criteria.

One of the things I like about this author is that he’ll be the first to tell you that he doesn’t know everything. He isn’t shy about asking for help. And so he does engaging three of the most renowned go-to-guys in the data world, all experts in their own right.  In this section, Graeme Simsion wrote a chapter on working with others. In the next section, Bill Inmon wrote a chapter on unstructured data, and Michael Blaha wrote a chapter on UML. I was very pleasantly surprised to find that all three contributions together helped complete the pieces missing in the first edition. 

Graeme Simsion shares his advice on working with others from the viewpoint of the data modeler as a consultant. Here he takes a look at the other side of the quality question, and cautions you not to get too focused on quality to the point that it becomes a detriment to the success of the project. I saw myself as he talks about getting caught in the perfection trap. I’ll heed his advice that the data model is ultimately judged by how valuable it is to a project’s success and not by how close it is to perfection. He also provides his own list of seven habits for highly effective data modelers to help us stay on track. My biggest take-away: the client should own the deliverable, not the data modeler.

SECTION V – Beyond Data ModelingBelieve it or not, you’re ready to leave the pool and jump head first into a small part of the ocean of outside influences that affect a data modelers’ work. Very few (if any) of us work in a vacuum, except maybe an astronaut.  Bill Inmon tackles unstructured data with taxonomies. Here he simply provides the best explanation about taxonomies and ontologies that I’ve found. And I’ve seen far more than I care to admit! Thanks to Bill and his clear explanation, they are both far simpler to understand than they sound.

Michael Blaha, who literally wrote the book on the subject of the Unified Modeling Language (UML), follows with an introduction about UML. You are getting it straight from the best. You can’t help but notice the similarities between a data model diagram and a class diagram as he describes the later. They both describe a data structure using similar components, and you can generate a database schema by applying a set of similar rules to either. Steve ends by answering the 5 most frequently asked modeling questions that he has encountered.

I dare say, by the end of the book you could know more than most data modelers working today. This revision took the first edition up several notches from what some deemed a data modeling for dummies book, to what is now a full-fledged textbook. It’s easy to see how it could quickly and easily light the way for many future data modelers in any classroom.  I have it on good authority that the author wrote this book to be the most easy-to-read and comprehensive data modeling text on the planet. I agree. This is in itself a wonderful way-finding tool for data modelers that’s very easy on the eyes and complete in its coverage. It will save you countless hours of confusion, and possibly even years of data modeling in the dark. The only really big flaw I found with this book is that it didn’t exist 15 years ago when I started as a beginner lost in a dark classroom somewhere in Texas, oh so many years ago.

Share

submit to reddit
Top