From Chaos, Order: SKOS type Framework for Financial Reporting

XBRL International, its members, regulators, and other stakeholders in XBRL and XBRL-based reporting are defining important "stuff" and they are representing that "stuff" in the XBRL technical syntax and putting that "stuff" in the XBRL International Link Role Registry (LRR), Functions Registry, Data Type Registry, Units Registry, and foundational/standard XBRL taxonomies such as those created by the SEC,  ESMA, and other such regulators.

This sort of "silo mentality" needs to be avoided, each party implementing XBRL does their own thing.  Also, this "stuff" needs to be organized into a framework, not be provided as a bunch of individual pieces that may, or may not, work together effectively.  This silo mentality and lack of coordination makes it harder for software vendors to provided consistent functionality, interoperable functionality, and robust functionality and features that take advantage of a coordinated framework.

One example of a higher level framework is SKOS, Simple Knowledge Organization System which is provided by the W3C. To understand SKOS, have a look at the SKOS Primer. SKOS tends to deal with "is-a" types of associations.

Another big issue is made apparent by the document, A Taxonomy for Part-whole Relations. Mereology is the theory of parthood relations: of the relations of part to whole and the relations of part to part within a whole. This is more about "has-a" and "part-of" or "composition" or "composed of".

Here are some of the pieces provided for XBRL-based financial reporting by different parties (this is not a complete or exhaustive list, it is only some examples):

• General-special associations, provided by XBRL International.
• Wider-narrower associations, provided by ESMA. (which begs the question, when should "general-special" be used, when should "wider-narrower" be used, and what is the difference between the two?  This also begs the question, how does "wider-narrower" relate to "broader-narrower" which is provided by SKOS)
• Class-subclass associations, provided by XBRL International. (Not sure how this relates to "class-subclass" as defined by W3C's OWL)
• Instant-inflow and instant-outflow, provided by XBRL International
• Instant-accrual, provided by XBRL International.
• Instant-contra, provided by XBRL International.
• ArcroleGrossNet, provided by EDINET (FSA Japan)
• ArcroleGrossAccumulatedImpairmentLoss, provided by EDINET (FSA Japan)
• ArcroleGrossAccumulatedDepreciationAndImpairmentLoss, provided by EDINET (FSA Japan)
• Restated label role, provided by XBRL International
• Originally stated label role, provided by Seattle Method (i.e. not provided by XBRL International or anyone else)
• Document and Entity Information, provided by SEC.
• Country and Currency Codes, provided by XBRL International.
• Geographic area codes, not provided by anyone. (for example, Europe, North America, South America, Pacific Islands)
• Parts of an economic entity, not provided by anyone. (for example, Reporting Unit, Asset Group, etc.)

Another example of what probably should be provided is a Dublin Core Metadata Element Set type set of elements that is used to provide information about "resources" (a.k.a. document).

Rather than having what amounts to a hodge-podge; one framework could be provided that pulls all these individual pieces together into one organized framework.  Alternatively, the separate pieces could be pulled together, but to do that...each of the resources needs to be represented using XBRL.