February 18, 2008 at 9:34 pm
Comments posted to this topic are about the item 5 Tips for Becoming a Better Data Modeler
February 19, 2008 at 9:43 am
This article is so "motherhood and apple pie" it says nothing. I do not think the work normalize or denormalize is even used. The author also did not even mention conceptual, logical and physical modelling. These are all extremely important since most data modellers jump right to physical modelling and never understand the data of the application problem that is to be solved.
If you want some meat and not just a bunch of words, I would highly recommend David Hay as an author and his free papers. David has the best book on data modelling called Conventions of Thought. This book will provide a solid foundation in data modelling. David also has a very good white paper on comparison of data modeling notation (Barker, Chen, UML, etc.). It is very important to understand the difference between notation and technique. Notation is just the (drawing) semantic. Technique is how you actually get the answer and what you do with it. David's website is: http://www.essentialstrategies.com/
If you are a beginner, I would highly recommend Richard Barker's book on Entity Modelling. Yes, it is from Oracle Press but it has nothing to do with the Oracle product other than solid data modelling notation AND technique. You can find used versions of this book on Amazon for ~$2. Read it. Study it. Live it.
Here is a link to Amazon for the book: http://www.amazon.com/Case-Method-Entity-Relationship-Modelling/dp/0201416964/ref=sr_1_1?ie=UTF8&s=books&qid=1203439277&sr=8-1
If SQL ServerCentral wants a real article written, feel free to drop me a line.
Bill
February 19, 2008 at 10:24 am
Thank you for the comments, Bill.
Correct - if you're looking for specific examples or "how to", this article isn't it. See my earlier series "Toward Integrity" for that: http://www.sqlservercentral.com/articles/Data+Modeling/61526/.
The purpose of this article is simply to give the aspiring modeler/designer some direction that's sorely lacking if one approaches the discipline from learning the tools (which, to be fair, is how I started out).
I second your recommendation of Hay's books. I have two on my bookshelf and will soon be ordering a 3rd by him.
TroyK
February 19, 2008 at 10:35 am
Troy,
Sorry to be so critical. I have been burned by people who use the title "Data Modeller" and they really are not. Or the Object crowd that don't care about the database design and really just want a file system. So, personally, I am very sensitive to this topic and react quite strongly.
There are few good resources anymore for the beginning data model student. That is why I point to Richard Barker's book. He walks through the initial stages and builds up a very strong business model that is represented in a data model. If you do not have this, spend the couple of $ to get a copy.
Bill
February 19, 2008 at 10:39 am
No worries, Bill. I intentionally kept the content a bit light because I specifically wanted to elicit others' opinions and resources here in the discussion. To that end, thanks once again for sharing those resources that you've found helpful.
TroyK
February 19, 2008 at 6:10 pm
Bill Wimsatt (2/19/2008)
This article is so "motherhood and apple pie" it says nothing. I do not think the work normalize or denormalize is even used. The author also did not even mention conceptual, logical and physical modelling. These are all extremely important since most data modellers jump right to physical modelling and never understand the data of the application problem that is to be solved.Bill
I partially agree that the content was a little light, but at the same time, what is there is all well written and correct. The fact it includes a list of recommended texts is also a good point. When there are an incredible number of books published even on a fairly specialized topic like data modeling and data base administration, having a well thought out and vetting reading list is itself of great value.
---
Timothy A Wiseman
SQL Blog: http://timothyawiseman.wordpress.com/
February 19, 2008 at 9:15 pm
Thanks for reading, Timothy.
Any books, classes, or online resources you've found helpful that haven't yet been mentioned?
TroyK
February 25, 2008 at 10:36 am
I would treat this effort as a class and offer up the following syllabus:
Materials
David Hay’s Conventions of Thought, Data Model Patterns
Richard Barker's Entity Modelling
Some book on 3rd order predicate calculus—need to find my as a reference
Installed Local Database (pick one) with SQL only tool (no TOAD et al)
Goals
Understand databases
-Relational vs. Hierarchical
-ACID test
Understand conceptual data modeling
Understand logical data modeling
-Relationships
-Attributes
-Logical PK
Convert conceptual data model into logical data models
Understand the rules for first, second and third-normal forms
Understand the reasons for de-normalizing a physical data model
Understand components of a database
-Table
-Keys
-Indexes
-Referential Integrity
-Views
-Functions
-Procedures
-Triggers
Understand issues related to data warehousing
-Summary Tables
-Star Schemas
-Point In Time Reporting
-Changes to Logical PK
Understand differences between OLAP and OLTP
Develop a dimensional data model to satisfy Barker’s model
Perform data manipulation using SQL
-Simple
-Complex
-Nested everything
-Models
-Analytic functions
Class
Read Barker
Build Barker as scripts and execute
Load Barker Structures from your scripts
Query Barker
Read Data Model Patterns
Rinse and repeat for the books
When you can write a paragraph on everything listed in the Goals, then you have completed data modeling 101-202. Congrats
February 25, 2008 at 2:23 pm
alittledog (2/25/2008)
I would treat this effort as a class and offer up the following syllabus:Materials
David Hay’s Conventions of Thought, Data Model Patterns
Richard Barker's Entity Modelling
Some book on 3rd order predicate calculus—need to find my as a reference
Installed Local Database (pick one) with SQL only tool (no TOAD et al)
Greetings, alittledog;
Definitely agree on the recommendation for Hay's book.
I agree also on the "(pick one)" recommendation for installing a local db. In fact, one thing I need to do is branch out some and play with MySQL and Postgres. Currently, my only real exposure is MS SQL and Oracle, so it would be good to get some hands on with the other players, too.
Thanks for the comments and recommendations!
TroyK
Viewing 9 posts - 1 through 8 (of 8 total)
You must be logged in to reply to this topic. Login to reply