Barbara
What a fascinating analysis - thank you.
I should perhaps have mentioned in my earlier posting that Australian
metadata models were very influential in the PRO 2002 work, albeit that
eventually a simpler approach was chosen.
One point on which I disagree somewhat: the PRO 2002 specification
explicitly groups folders into a hierarchy of "classes" (similar to
"series"), and explicitly allows disposal schedules to be allocated both to
folders and to classes (at any level in the hierarchy), as well as requiring
that disposal schedules and other metadata can be inherited down the
hierarchy.
Marc Fresko
> -----Original Message-----
> From: Barbara Reed [mailto:[log in to unmask]]
> Sent: 15 March 2003 05:25
> To: [log in to unmask]
> Subject: Differences between RM specs, UK and DoD
>
>
> Dear all
>
> Forgive yet another Australian intrusion onto your list, but I've been
> examining various specifications (in addition to writing
> some) for a while.
> There is quite a bit of difference between the US and UK specs. Our
> Australian specs are nearer the UK model than the US. There
> are a few areas
> of major difference.
>
> The first and most significant difference is in the level of
> aggregation at
> which the records management controls are invoked. In
> Australian and UK
> practice (ie the PRO spec) the controls are applied at folder
> level. In the
> US specification the controls are attributed to record
> categories (perhaps
> also known as records series if such a thing exists in the electronic
> world). What this means is that the level of process
> controls (access,
> disposal, classification etc) in US is at this higher level
> and not at the
> folder/item level. I think that applying controls at the
> lower level of
> aggregation (folder/item) makes the UK/Aus approach better
> for actively
> managing the processes of creation and management of items,
> whereas the US
> approach is more about managing groups of folders/items. This is a
> substantial difference. In the ideal future, we'd have both levels of
> aggregations being managed, with the UK/Aus approach being nested to a
> category/series. The category/series would then need some
> different types
> of controls to be defined to manage things like inheritance
> from different
> systems and perhaps migration, and even some grosser level of
> disposal (ie
> all the files from the order processing series)
>
> Use of the fileplan (dreadful name and I'd have some
> suspicions about its
> proprietary origins!). In US and UK there is a set of
> requirements that
> link the allocation of file/folder numbers to the levels of
> the hierarchy
> defined in the classification scheme (aka file plan). Thus meaningful
> numbers where the number equates to a concept or term are
> allowed. Most
> Australian stuff discourages this approach as, from
> experience, it ends up
> being difficult to update and can fossilise. The fileplan in
> the Aus/UK
> approaches is about applying controls on what titling can be
> applied to
> records, an approach which the higher level of aggregationin the US
> precludes. Thirdly the fileplan in all specs then enables
> the linking of
> security/access and disposal controls. Personally I'd argue
> (as I do with
> no avail in Australia!) that what we need is different things
> - one is a
> functional analysis tool which manages the definition of the
> functional
> terms, the mandates etc. From that authoritative analysis,
> various other
> control tools need to hang - the language control (file titling), the
> security/access classification schemes, links to organisational
> structures/people/roles, the disposal schedule. They all
> link, but making
> one tool (called the fileplan-yuk) do all these things is too
> much and makes
> for muddiness in thinking and difficulty in getting particularly good
> implementation. Naturally this latter bit is my personal
> opinion to be
> disregarded in the analysis of the differences.
>
> Process orientation: The specifications all have significant
> numbers of
> processes associated with records and there is a basic
> uniformity on what
> they are (bearing in mind earlier comments on level of
> application). When
> we hit metadata requirements, however, we are at another point of
> difference. The US model doesn't really have a metadata
> standard as such
> associated with it, but does list metadata items required.
> Only in the
> definition of how attribution/review of vital records status and in
> applying/upgrading/changing the security levels are process
> metadata items
> defined. Similarly in the UK metadata specification, the
> process elements
> are best articulated in the internal/external access
> processes. Access
> processes are just one of the recordkeeping processes that
> should maintain
> metadata about themselves (records of recordkeeping
> transactions should be
> managed as records). Australian metadata sets are
> increasingly taking a
> more complex view of recordkeeping metadata to enable them to
> define all the
> records processes as well as the initial creation metadata
> (profile metadata
> if you like).
>
> I think that these are the major areas of difference and the most
> significant of these is the different layers of aggregation
> to which the
> controls are attributed.
>
> regards
>
> Barbara Reed
> Recordkeeping Systems Pty Ltd
> Ph: 61 2 9369 2343
>
|