UML Model for Financial Reporting
I am trying to sort some things out in my mind.
UML (the Unified Modeling Language) is communication tool which is used in designing software systems. I have long wondered why there was no UML model created for XBRL. What I think I should have been thinking is why there has not been a UML model created for financial reporting. While a UML model for the XBRL syntax would be helpful in understanding the XBRL syntax, a UML model for financial for financial reporting can help communicate the "moving parts" of financial reporting. Once those moving parts are accounted for and documented in UML, then one can apply the XBRL syntax or any other syntax really to that model.
(Most business users don't really understand what UML is. If you want to understand UML, here is a nice little tutorial. I can recommend a great book that is helpful in understanding UML, UML Explained by Kendall Scott. Understanding UML will take some effort, but it is worth the investment if you are involved in architecting systems which use XBRL, architecting XBRL taxonomies, etc.)
It is not that nothing has been done with UML relating to XBRL.
The US Corporate Actions Taxonomy has UML diagrams. See this presentation by Taku Hasegawa (NTT DATA Corporation) in July 2009.
In what looks like about 2002 a paper Augmenting XBRL Using UML: Improving FInancial Analysis was written by Joseph Callaghan, Arline Savage, and Vijayan Sugumaran of Oakland University. The paper makes the following statement in its abstract:
UML provides a conceptual framework to capture more meaningful semantics, particularly among related, collaborating objects. [CSH: by collaborating objects I think they mean relationships.]
UML capturing semantics? I thought that is what XBRL and OWL were for.
If you do a search in Google on "UML +XBRL" you get 895,000 hits. Many of these documents seem to relate to XBRL GL. From Wikipedia you can go to a UML Profile for XBRL Global Ledger created by OMG.
One thing I don't seem to understand is how OWL relates to UML and XBRL. Is OWL another way to model the domain of financial reporting, or is OWL another syntax just like XBRL is a syntax? Here are some OWL ontologies for XBRL and the US GAAP Taxonomy which seem to model more the XBRL syntax than financial reporting.
Currently all this is somewhat confusing to me. But, it does seem that there really needs to be a UML model for financial reporting. Some people refer to this as a "logical model" I think.
One thing which I do think I am realizing is that the value of XBRL is not the XBRL syntax, the value seems to be the semantics of business reporting and financial reporting and the XBRL taxonomies which are being created such as the IFRS and US GAAP taxonomies for financial reporting.
XBRL, OWL, RDF, UML all seem to be ways to instantiate semantic information relating to financial reporting. PDF, HTML, Word and Excel are other ways but those are for presentation of the information.
Reader Comments (1)
UML can be a Missing Link between business people and technical people, it's just like you said.
As UML is not a tool for specific purpose but a tool for general purpose, both XBRL and OWL syntax can be drawn as UML diagrams.
So that means UML can capture semantics.
But, that is not the essential significance of UML.
UML per se is a communication tool for human, not for system.
(Code generation capabilities is only one aspect of UML.)
With an appropriate granularity of the model, you can facilitate communication between business people and technical people.
In my opinion, UML Class diagrams for financial reporting should be drawn as composite of financial statements and/or networks.
And, in some cases, UML Package is used to indicate DTS.
(Law or regulation level model is too coarse for taxonomy developers; on the other hand, XBRL syntax level model is too fine for regulators.)
Regards,
Taku