Re: help with EER diagram

Wookie wrote:

> These are the entities that I got:
>
> book, fiction, non-fiction, paperback, large paperback ,
> hardback,other merchandise, magazines, stationery items, greetings
> cards, seasonal items, receipt, special offer voucher, sales person
> code.
>
> Are there more?

Those aren't entities, but attributes.

If you think about it, each of those "entities" could be expressed as a numerical value in a main table called eg. "products". That table could have columns like "number in stock", "number sold" etc. The "entities" that you have described above, are "item types", and could be enumerated in another table where eg. "books" could be assigned a number 1, fiction as number 2, etc. In the next step, you may find that "books" is worthy of yet another table, with attributes like "fiction", "non-fiction" etc.

The way to go about a solution of an assignment like this, is to write everything up in a huge table with every attribute in a separate column. Following the rules of normalization, you will find that each "text field" that tend to repeat itself, should be broken out into another table.

It's not rocket science. Start with a crude model, and then refine it in an iterative process.

A quite good example of normalization of just the kind of data model that you describe here, with some very useful commentaries, can be found as Appendix A of this document:

<url:http://xml.coverpages.org/GENTECH-DataModelV11.pdf>

-- 
Leif Biberg Kristensen
http://solumslekt.org/