RE: Requirement for 'Valid Time' Issue-16

Frans, Kerry

I now have Issue-16 in both the body and subject of the email for the tracker to spot it. Thanks Kerry.

So is the debate between Kerry and Frans: “entirely in-scope for us to take a position” and "Ensure alignment with existing methods for expressing the time in which data are valid (e.g. http://purl.org/dc/terms/valid)."
The problem I have with Dublin Core as it stands is that it specifies “Date valid” and it is not clear whether this is inclusive of “Time” with hours minutes seconds etc.
Chris
--------------------------------------------------
From: Kerry Taylor [mailto:Kerry.Taylor@acm.org]
Sent: Tuesday, November 10, 2015 9:26 PM
To: Little, Chris
Cc: public-sdw-wg@w3.org; Frans Knibbe; Simon Cox; Tandy, Jeremy; Kerry Taylor (Kerry.Taylor@acm.org)
Subject: Re: Requirement for 'Valid Time' issue-16

I agree. I also heard a requirement of this form expressed by the wot interest  group in sapporo. We should either refer to the met ocean wg reference time, or build our own if that is not good enough.

This is not strictly owl-time, but it is very near by and entirely in-scope for us to take a position.

( chris, mentioning ISSUE-16 in the email gets the message against  the issue in the tracker).

Kerry

From: Frans Knibbe [mailto:frans.knibbe@geodan.nl]
Sent: Wednesday, November 11, 2015 1:42 PM
To: public-sdw-wg@w3.org
Cc: Simon Cox; Tandy, Jeremy; Kerry Taylor (Kerry.Taylor@acm.org); Little, Chris
Subject: Re: Requirement for 'Valid Time'

Thanks to Chris for continuing the discussion. My comments are inline:

2015-11-10 17:47 GMT+01:00 Little, Chris <chris.little@metoffice.gov.uk<mailto:chris.little@metoffice.gov.uk>>:
Dear colleagues,

Unfortunately, tomorrow I will be travelling by train, with rubbish phone connections, so it will be highly unlikely that I can join in, so I will add my views on UCR-16 here:

1. I agree that 'Valid Time' is outside the scope of the Time OWL ontology, as are lots of other time semantics (verification time, time of last update, creation time, publication time, etc).

2. There are widely used ontologies that recognise different kinds of time e.g. Dublin Core.

3. It seems to be out of scope for spatial data as opposed to just 'data on the Web'. However, there are lots of examples in the spatial domain, such as environmental observational data and weather forecasts. Increasingly, to know a location, one needs to know when.

True, but that by itself does not warrant us adopting this requirement. Examples of many aspects of handling data can be found in the spatial domain, but that does not mean we should consider working on them. Why make an exception for this particular example?

4. The Met Ocean Domain WG of OGC produced a Best Practice profile of the OGC WMS1.3 standard to capture some of these time concepts and improve interoperability arising from real world experience of interoperability problems. The BP gave a standard name for these non-standard time semantics ("Reference Time").

As I see it, OGC semantics form a more or less closed system. Other than very basic data types they can not rely on semantics that are defined outside of the system. Hence the need to standardise aspects of reality that are not strictly spatial in the OGC. But since the SDWWG is acting in a web environment, it should be OK to have to rely on semantics that are defined somewhere out of our sphere of control.

5. I do not know if there is a generic 'framework' for capturing different time application level semantics in Data on the Web.

6. I think this is important for enough SDWWG Use Cases that it should not be ruled out of scope unless some other 'owner' is identified to take it forward.

Yes, we could take it upon ourselves to make some other party the problem owner. But that would be something very different from taking it upon ourselves to work on semantics for valid time.

This remark has led me to see the resemblance between ISSUE-16<https://www.w3.org/2015/spatial/track/issues/16> and ISSUE-11<https://www.w3.org/2015/spatial/track/issues/11>. There was a risk of having a requirement that would make us have to deal with the semantics of provenance. Now the provenance requirement<http://w3c.github.io/sdw/UseCases/SDWUseCasesAndRequirements.html#Provenance> states that we should seek alignment of our work with existing standards for provenance. I believe that issue was resolved to everyone's satisfaction.

In this case we could go the same way: have a requirement for OWL Time that says that the work should be aligned with existing methods of expressing valid time. That way we do acknowledge that valid time is important for spatial data, but we do not make it a primary responsibility for us to have semantics for valid time.

This could lead to a requirement like:

"Ensure alignment with existing methods for expressing the time in which data are valid (e.g. http://purl.org/dc/terms/valid)."

In this case the appropriate related deliverable would be only OWL Time.

Greetings,
Frans




HTH, Chris


Chris Little
Co-Chair, OGC Meteorology & Oceanography Domain Working Group

IT Fellow - Operational Infrastructures
Met Office  FitzRoy Road  Exeter  Devon  EX1 3PB  United Kingdom
Tel: +44(0)1392 886278  Fax: +44(0)1392 885681  Mobile: +44(0)7753 880514
E-mail: chris.little@metoffice.gov.uk<mailto:chris.little@metoffice.gov.uk>  http://www.metoffice.gov.uk


I am normally at work Tuesday, Wednesday and Thursday each week