FedNet: A US Government Web of Things

All,

With a migration to RESTful identifiers, government data bases often suffer from an administrative incoherence which has little or nothing to do with data content.  In particular, the Gold Standard for longitudinal research remains Cohort Analysis.  And in particular, small sub-lists of small code spaces break the Component-Cohort Model by fragmenting Components into commercially important Cohorts and "others" (fly-over).  This creates havoc with the fundamental assumption of Cohort Analysis - that the subjects of study distill "herd" policy if given (and only if given)  a high number of choices (degrees of freedom). When the Web of Things is distilled down to a few highly populated Things, the "Open World Assumption" becomes an imperative.

A pre-filtering of undifferentiated Cohorts into "good" and "bad" or "important" and "marginal" is a bureaucratic (administrative) conceit and can be left right there under almost all circumstances. But ...   the Federal Information Processing Standards (FIPS) and the ANSI Feature Codes replacement are both flawed frameworks with respect to Cohort Analysis.

>From a practical standpoint, the Net of an undifferentiated Cohort can be replicated with virtual serial numbers and generational differences visualized.  For example, the opportunistic advantage a Community has to progressively educate the young in winter months while reserving the summer for "making hay" is inversely proportional to abs(Latitude) alone (see issue 4/4).  The local economic decision to *not* educate the young in order to save money has dire consequences, but so is the failure to realize that Latitude (and transportation issues) are not a basis for general Educational Policy.

There are (at least) 4 issues around semantic friendly bookkeeping for Organizations.

http://www.rustprivacy.org/2013/education/fednet.html

Have fun, and Happy Holidays

--Gannon

Received on Friday, 20 December 2013 16:49:03 UTC