Saturday, February 21, 2015

Naming conventions in Dimensional Warehousing models

I am looking at a good and robust set of naming conventions to adopt for a Dimensional Data Warehousing Model that I am developing. As I looked around, there were different conventions with opposing perspectives. I still do not know which makes more sense, but I will first enumerate the list of questions I have.

1. Should the table name suggest whether the table is a fact or a dimension?
2. Should Entity name prefixed in each table attribute?
3. Should the attribute name follow some convention?
4. Should the fact table reveal something about the grain?
5. Should a fact measure provide some clue of the units the measure is recorded in?

I will update the post, as I get closer to my answer, but already I have seen some links/ blogs from Kimball group and another set of standards from Oracle BI Applications. Here are the links to these resources as I will revisit this post.

1. Oracle naming conventions

2. Kimball Group Blog on naming convention

I will explore this further and keep updating my preferences as we go along...