Junk Dimension Naming

What do you call your junk dimensions? Do you use "junk" in the dimension name? Junk implies garbage so I don't like using it. Most dimensions are easily named since they represent a single person, place, thing, concept, or event. Junk dimension contents are like a pot luck dinner.

Well, since its pot-luck, how about 'casserole' or '7 layer bean dip'?

I just try to use something that fits. I agree 'junk' isn't one you should use... users have no sense of humor. Heck, even Ralph uses 'mini-dimension' in recent books... I hate that term... its just too PC.

I like the concept of junk dimension but not the name. I am working on a BI project managing student information for all public schools. Therefore my junk dimension is collection of low cardinality demographic attributes about students so it's called DimStudentDemographic. Profile could be another choice if I need to include something more general. However anything is better than DimStudentJunk.

In my experience, not many people like the concept of junk dimension, maybe because of the name, or its unusual handling uncomfortable to traditional data modellers. Name is a powerful factor when you try to defend something that others try to get rid of.

I'd love to use some other name instead of junk dimension, even just to refer to the concept, but am not sure if the term mini-dimension can be used interchangeably in all cases. Mini-dimension seems to be a dimension with a group of low cardinality attributes extracted out of a large or rapidly changing main dimension, and both dimension keys appear as FK in fact table to reflect their correlation. Whereas junk dimension can be a bucket of related or unrelated low cardinality attributes that can be de-normalised into a single dimension instead of many discreet small dimensions.

I tend to use "Attribute" in the name of Junk dimensions, as in d_claim_attribute, d_member_attribute, etc.

How about "Miscellaneous Dimension" as an alternative term without the negative connotation of "junk"?