Question - Nulls as Dimension Attributes

I'm wondering if this is a common practice to add the following 2 'standard' attributes to dimension tables. For example:

I have a TRANSACTION_DIM table, I have these 2 'standard' attributes (plus the rest of the business attributes)
transaction_type, and transaction_type_desc

These two 2 'standard' attributes allow me to :
1) provide a description of whether a dimension row exists or not (or not known) when joined to the Fact.
Example: transaction_type = NA, and transaction_type_desc = Not Applicable
2) classify the transaction information.
Examples: transaction_type = Claim, and transaction_type_desc = Claim Transation
transaction_type = ENRL, and transaction_type_desc = Enrollment Transaction
or
3) default the 'standard' attributes value to 'DATA' for instance, when unable to classify a transaction record due to some null attributes in the dimension table.

I read the Kimball Design Tip #43 but didn't see if having these 2 columns are ok or not.

Apologies for the delay in reply. Those attributes look fine. Generally speaking, it's difficult to put in too much metadata about data quality and so on. Lots of times we don't share this kind of attribute with the business users, depending on requirements.

I do trust that you're ensuring referential integrity between facts and dims... not putting in a fact row unless the dimension row exists.