- From: Owen Ambur <Owen.Ambur@verizon.net>
- Date: Sat, 4 Jan 2020 23:34:49 -0500
- To: Paola Di Maio <paoladimaio10@gmail.com>
- Cc: W3C AIKR CG <public-aikr@w3.org>
- Message-ID: <81a21710-2b07-cc99-1b41-c2fcd4254043@verizon.net>
Paola, machine-readability
<https://en.wikipedia.org/wiki/Machine-readable_document> is not a
U.S.-based standard. It is a concept, i.e., a good practice. Nor are
ISO 15489 or ISO 17469-1 U.S. standards. They are international
standards. So the only issue is why they are not being appropriately
applied -- particularly by organizations like IAC. To the degree there
may be obstacles to doing so, we should explore means of reducing, if
not eliminating them. In the event that lack of awareness may be one of
them, it will be interesting to see if IAC is open to learning about them.
In any event, the U.S. Federal Data Strategy Action Plan includes a
couple dozen references to metadata, including this one
<http://stratml.us/carmel/iso/part2/2020APwStyle.xml#_b19dc80e-2dd4-11ea-a6d4-d5cd0183ea00>
regarding geospatial data:
The Federal Government’s lead agencies for NGDA data assets will
identify, inventory, and publish the status and standards being used
for each of the NGDA data themes and content and services metadata,
consistent with */international standards/* ...
This year's planned actions related to AI
<http://stratml.us/carmel/iso/part2/2020APwStyle.xml#_b19dbd82-2dd4-11ea-a6d4-d5cd0183ea00>
include:
* Investigating barriers to access or quality limitations of Federal
data and models that impede AI R&D and testing. A Request for
Information (RFI) was issued as a Federal Register Notice by OMB
inviting the public to identify needs for additional access to, or
improvements in the quality of, Federal data and models that would
improve the nation’s AI R&D and testing efforts.
* Addressing identified barriers by updating Federal data and source
code inventory guidance for agencies to utilize in enhancing the
discovery and usability of Federal data and models in AI R&D.
Perhaps we should explore prospects for getting someone to brief our
group on those plans. Of course, it would also be nice if the results
were reported in an open, standard, machine-readable format like StratML
so that learning about them were not limited by the constraints of time
and space. See the performance indicators at
http://stratml.us/carmel/iso/part2/2020APwStyle.xml#_66413a26-2dde-11ea-89b4-de7271babdf6
The RFI is targeted for completion next month.
Owen
On 1/4/2020 9:59 PM, Paola Di Maio wrote:
> Thank you Owen
>
> I do like standards as well, they are neat and give a sense of order
> and certainty
>
> at the same time, being a citizen of the world, I seek a global
> perspective. :-) The question is often: is a US standard good also
> for the rest of the world? Does it fit universal requirements?
>
> OK to start from where we are, and from what have got (say the ISO you
> mention) But we should keep in mind that what we have is a starting
> point that needs to be validated, or evolved, to fit a broader spec.
>
> I think here the point for us is avoiding to make country based
> assumptions, and avoiding wanting to impose a single view of the
> world, however pretty :-)
>
> I am shocked at what I see, despite the www making us one world, we
> are still culturally segregated and gliding over too many important issues
>
> For example, just emailed Norvig ccd Vinay Chaudry because he is a
> member of this list, as well as a board member for AAAI JOURNAL where
> this great paper is published:
> https://www.aaai.org/ocs/index.php/AAAI/AAAI16/paper/download/12444/12195
> asking whether this was a US centric paper, but he says he cannot remember
>
> there is no dataset to verify these findings , and not even a mention
> of whether the findings are based on a survey sample population which
> I assume is english speaking and probably US based. In the rest of
> the world, from Latam to Middle and far east, afaik, these findings
> may not true, its hard to tell given the lack of mention
>
> There is an assumption (in the USA) that the US is the center of the
> universe of discourse, and probably true also in other regions.
>
> As much as we all can identify to some extent with US standards, and
> we like them, we need to make sure the scope and limitation are
> clearly stated and hopefully address that
>
> a plurality of cultural and geographic perspectives, or the intention
> to pursue such plurality, should be manifest in this CG work, whatever
> way you want to reflect that
>
> :-)
>
> PDM
>
>
> On Sun, Jan 5, 2020 at 6:25 AM Owen Ambur <Owen.Ambur@verizon.net
> <mailto:Owen.Ambur@verizon.net>> wrote:
>
> Paola, yes, indeed, our focus should be on global goals. However,
> standards and good practices need not be reinvented by
> international bureaucracies if they have already been specified by
> someone else -- not just nationally recognized SDOs but by anyone,
> anywhere on earth.
>
> It seems to me that publishing public information in open,
> standard, machine-readable formats having the attributes specified
> in ISO 15489-1 is such a good practice. It would be nice to think
> IAC might be willing and able to foster adoption of that good
> practice by its stakeholders. That is the prospect that prompts
> my interest in participating in a presentation at their conference.
>
> BTW, here in the U.S. official policy since at least 1998 has
> directed agencies to consider using internationally adopted
> voluntary consensus standards. Here are the applicable sections
> of OMB Circular A-119
> <https://www.whitehouse.gov/wp-content/uploads/2017/11/Circular-119-1.pdf>:
>
> h. Does this policy establish a preference between domestic
> and international voluntary consensus standards?
> This policy does not establish a preference between domestic
> and international voluntary consensus standards. However, in
> the interests of promoting trade and implementing the
> provisions of international treaty agreements, your agency
> should consider international standards in procurement and
> regulatory applications.
>
> i. Should my agency give preference to performance standards?
> In using voluntary consensus standards, your agency should
> give preference to performance standards when such standards
> may reasonably be used in lieu of prescriptive standards.
>
> 7. What Is The Policy For Federal Participation In Voluntary
> Consensus Standards Bodies?
> Agencies must consult with voluntary consensus standards
> bodies, both domestic and international, and must participate
> with such bodies in the development of voluntary consensus
> standards when consultation and participation is in the public
> interest and is compatible with their missions, authorities,
> priorities, and budget resources.
>
> In short, the problem is not the policy but, rather, the
> performance, i.e., the lack thereof in many instances. What's
> needed is not more policy or new "strategies" but more
> accountability and better performance. Hopefully, the U.S.
> Federal Data Strategy Action Plan will make a meaningful
> contribution toward that end, at least with respect to grant
> funding
> <http://stratml.us/carmel/iso/part2/2020APwStyle.xml#_ea289a44-2e58-11ea-bd1a-70248cbabdf6>.
>
> Owen
>
>
> On 1/3/2020 8:02 PM, Paola Di Maio wrote:
>> Thank you Owen
>> it will be great if we could align our work to some of these
>> objectives, please keep an eye on that (my mind being very
>> expanded at the moment)
>> also, can we find alignment of our own work with these US based
>> objectives, also with more global, less US centric strategies and
>> goals.
>> I am thinking UK EU, China and rest of the world as well
>> pdm
>>
>> On Sat, Jan 4, 2020 at 3:35 AM Owen Ambur <Owen.Ambur@verizon.net
>> <mailto:Owen.Ambur@verizon.net>> wrote:
>>
>> The U.S. Federal Data Strategy Action Plan for this year is
>> now available in StratML format at
>> http://stratml.us/drybridge/index.htm#2020AP
>>
>> Action
>> 8:<http://stratml.us/carmel/iso/part2/2020APwStyle.xml#_b19dbd82-2dd4-11ea-a6d4-d5cd0183ea00>AI
>> - Improve Data and Model Resources for AI Research and
>> Development includes direction to provide an updated
>> inventory of technical schema formats.
>>
>> It will be interesting to see if this group may have value to
>> add in support of that objective. If so, the IAC conference
>> <https://docs.google.com/document/d/1iZARdPoWeEZzld1iugl5hlaMh7aaYrJkD7SiEycXvdQ/edit>
>> in September might be a good venue in which to share it.
>>
>> Owen
>>
Received on Sunday, 5 January 2020 04:34:56 UTC