Re: Closing of open (user story) issues

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

One doesn't have to use anonymous classes.  If there is a named class
involved then that can be part of the axiom.  However, any if-then property
relationship is going to involve some sort of if-then syntax, as far as I
can see, and the axiom I gave is just about the simplest imaginable.

You are correct that inter-property value relationships are not part of OWL
so you can't require drop-off dates before pirk-up dates.  That's an
expressivity issue. It would be easy to augment the language to add in these
sorts of things.  (Implementation might be different, but the whole idea
requires an entire new kind of facility so the implementation effort is not
going to be easy in any case.)  Differing security digits is already
possible via axioms like
   AmExCreditCard <= securityCode<=10000

You seem to be finally asking for some new JSON data structure that encodes
these constraints.  I don't see how this is any simpler than an axiom- (or
constraint-) based solution.

peter


On 02/03/2015 07:01 PM, Irene Polikoff wrote:
> Yes, I suppose one could say that an anonymous class of all things that
> have at least one foo is equivalent to an anonymous class of all things
> that have at least one bar and then, somehow use this information for the
> form. It is a rather complex way of communicating that foo and bar are
> "co-dependent".
> 
> There are other constraint-type bits of information that one may need to
> make available to a form e.g., you can't have a drop of date that is
> before a pick up date or that if your payment method is American Express,
> the length of the security code must be 4 digits, but for other credit
> cards it is 3 digit long, etc.
> 
> Even if every one of these things was somehow expressible in OWL, when
> ways in which they are expressible is too complex, it stops being
> practically usable. What is probably expected here is a single JSON
> structure that expresses the constraints built into the form in a
> straight forward way.
> 
> This is a matter of judgment, of course.
> 
> Irene
> 
> -----Original Message----- From: Peter F. Patel-Schneider
> [mailto:pfpschneider@gmail.com] Sent: Tuesday, February 03, 2015 8:26 PM 
> To: Irene Polikoff; 'Simon Steyskal'; 'Public-data-shapes Wg' Subject:
> Re: Closing of open (user story) issues
> 
> Both S19 and S20 involve tools analyzing the requirements, not tools
> enforcing the requirements.  If these two stories are accepted then the
> working group should specify how these analysis tools will work.  I'm not
> convinced that this should be part of the working group's output.
> 
> Further, such analysis tools don't need to be run on constraint / shape
> documents.  They can be run on pure open-world ontologies.  For example,
> the OWL axiom
>> = 1 foo  <=  >= 1 bar
> (everything that has at least one foo also has at least one bar) could be
> interpreted in the way that you suggest.
> 
> There could even be two ontologies, as suggested in the story - one for
> the input information and one for the output information.
> 
> 
> peter
> 
> 
> 
> On 02/03/2015 01:20 PM, Irene Polikoff wrote:
>> S19's connection to constraints is not clear.
> 
>> With S20 it is clearer because it is about creating data which needs to
>> conform to constraints.
> 
>> In this context, one example of a constraint could be "if there is 
>> value in this field, then there must be value in that field" such as if
>> a person enters or selects a program name for the "rewards program" 
>> field, they must enter their participant's ID number for the program. 
>> And vice-versa.
> 
>> The form may need to have this information so that it can 
>> enforce/encourage correct data entry without sending data to the server
>> and, in fact, fields may appear dynamically on the form - if a reward
>> program is selected, then the participant's ID field is shown. You
>> can't express such co-dependence of properties in OWL.
> 
>> Irene
> 
>> -----Original Message----- From: Peter F. Patel-Schneider 
>> [mailto:pfpschneider@gmail.com] Sent: Tuesday, February 03, 2015 2:53 
>> PM To: Simon Steyskal; Public-data-shapes Wg Subject: Re: Closing of
>> open (user story) issues
> 
>> You can check to see who raised an issue by looking at its page in 
>> tracker. I am the person who raised all these issues.  There has been 
>> no notification that I have noticed indicating that any change has been
>> made to any of these user stories.
> 
> 
> 
>> I do think that some of the user stories may have been changed.  I'll 
>> summarize my current thinking of the status of each of the relevant 
>> user stories here.
> 
>> ISSUE-8 User story S6
> 
>> This may have changed somewhat.  It appears to be asking for partial 
>> ontology import.  There is still no connection to constraints or
>> shapes.
> 
>> ISSUE-9 User story S7
> 
>> The creator of this user story agrees that it is a repeat of S4, and 
>> can be removed.  I have updated the status.
> 
>> ISSUE-11 User story S9
> 
>> The continuing problem with user story S9 is that it asks for something
>> to exist but not be specified.  It is unclear as to what that means. 
>> Discussion on the user story may have cleared up the confusion, but the
>> beginning of the story is still unclear.
> 
>> ISSUE-12 User story S10
> 
>> The description of the story is still very limited.
> 
>> ISSUE-13 User story S12
> 
>> This user story still contains no details as to what is supposed to be
>>  happening.
> 
>> ISSUE-14 User story S14
> 
>> This user story still has unresolved discussion.
> 
>> ISSUE-15 User story S17
> 
>> This story is about referring to part of a data set.  The connection to
>> constraints is unclear, even though it talks about shapes.
> 
>> ISSUE-16 User story S18
> 
>> This story is about exporting part of an RDF graph.  It appears to be 
>> very similar to S17.  The connection to constraints is similarly 
>> unclear.
> 
>> ISSUE-17 User stories S19 and S20
> 
>> User story S19 is about querying to find out what should be in some 
>> data as opposed to constraints on what is in the data.  It is unclear 
>> what role constraints have in this story.
> 
>> User story S20 is very similar to S19.  It is similarly unclear as to 
>> what role constraints have in this story.
> 
> 
> 
>> peter
> 
> 
> 
>> On 02/03/2015 10:59 AM, Simon Steyskal wrote:
>>> Hi!
> 
>>> May I ask the creators of issue:
> 
>>> ISSUE-8 ISSUE-9 ISSUE-11 ISSUE-12 ISSUE-13 ISSUE-14 ISSUE-15
>>> ISSUE-16 ISSUE-17
> 
>>> to check whether their issues were addressed and if so, if those 
>>> issues can be closed.
> 
>>> thx, simon
> 
> 
> 
> 
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1

iQEcBAEBAgAGBQJU0gfnAAoJECjN6+QThfjzWPMIAIKYmKkX2ojyFavn8gAWme1O
/fxCOnFsQooZR8vDOdmNj+Hzk1yZQRjv6i1r1G0jjblUEbUn0ehBq30tC1dfLh05
lfo9FP9+d0+cwTf+3QydCVgUpAsngkh8iSnG14Wn5vHaP4QEuB4pa07CDYBNXCqP
1u9UrXOYtZ1tZ0Vapm65MVnTCXExbpbN0Zs9HGyk80Bw5iwiomI6/ejFzFKXfBLz
tOBXjNvgAfZbNmew9INRYGJEiZmxt9v6xsVzriPsX9x+wlTq5cnY0KX4k4ml0HYb
ZjXfxRZsDTH31Vb3orzQ4R2j5QzJB5H+j7RruNVi+pNF/dJ330EZSFwV/oe/E24=
=+yqD
-----END PGP SIGNATURE-----

Received on Wednesday, 4 February 2015 11:52:37 UTC