Normalisation

[gary223]

Hi,

I want to do 0NF to 3NF for a few databases but really can't get my head around on how to do it correctly. The examples I've found online just don't make sense to me. I've spent the last couple of hours trying to understand

I attached an ER diagram of one of my databases. If you could show me how to solve that one, I would be able to do the rest as they're similar.

Thanks.

[ajetrumpet]

are you actually concerned about normalization?? IMO, following it explicitly should really only be done in a classroom.

[gary223]

I need to understand it for a project so its pretty important I know how to do it.

[ajetrumpet]

OK, so obviously you've read about it. What exactly are the issues you're having with level one and level 3 of the process??

and what article, if any on the web, did you read?? no two are the same.

[gary223]

I've read info and tried to follow examples on many sites but did not understand well enough to apply it to my own.

http://db.grussell.org/section008.html
http://www.sqa.org.uk/e-learning/MDBS01CD/page_26.htm

...among many more I checked out.

The main issue is I don't know where to start and how to do it basically.

[ajetrumpet]

I don't know where to start and how to do it basically.

Gary, saying "don't know how to do it" is really a question for a teacher. first and foremost.

secondly, I'm referring to this link: http://db.grussell.org/section008.html

1NF is basically explaining the fact that access is different from excel. access is 3-dimensional, while excel is not. It's also telling you that "groups" of data (the "main" data that has "sub-type" data sets related to it) should never have repeating records. As in, if your data "group" is student names, a table in 1NF will never have the same student name listed twice.

The 3NF is just ridiculous, in my eyes. That article goes so overboard with theory it's hard to even understand what they're talking about! First of all, how old are you?? The reason I ask: I don't think that article should be read by college students, or people close to that age. I could hardly understand it myself.

IMO, the 3NF section is nothing more than further explanation about 1NF data "groups". Honestly, 3NF will probably fall into the place naturally if you understand the basic hierarchy of how database information is supposed to be structured.

I personally think one of the issues that you can fix is to not look at sites like: http://www.sqa.org/

Society of Quality Assurance?? You can be sure that the amount of "theory" you'll find in organizations like that will be extremely high. And when theory is that important, you get unnecessary complexities and incredible redundancy (sometimes). The USA governmental operation is a prime example of this.

I would find others sources on the net, like fmsinc.com, or mvps.org. Those sites, I believe, are run by actual business operations and the tutorials are easy to read.

As a general rule, that I follow anyway, business people are 500% more likely to keep things simple than organizations that exist solely for the purpose of "maintenance" and "governance". Anything that has "society" in the name is probably one of those organizations.

Oh, and by the way, if you're interested:

• "non-key attribute" means "a piece of child data"
• "repeating group" means "a record in the parent table where the PK field data is repeated"
• "transitive functional dependencies" probably means "child data that is in the right place"
• "whole key" probably means "the PK, regardless of whether it's one field or a composite of two or more fields"

Hope ya learned something!

[gary223]

Thanks.

...