obi-protocol-application-branch Mailing List for OBI

SourceForge Headquarters 225 Broadway Suite 1600 San Diego, CA 92101 +1 (858) 422-6466

Bjoern, I have highlighted the queries you can make and the queries you
cant. If from you conclusion you infer that I am trying to be explicit, then
thanks, this is what defining ontologies is all about.

As Melanie points out as well, the definition, the domain and range and the
use of the realizes relation do not match. This needs to be addressed for
the 1.0 release. I have added this issue to the tracker.

Frank

On Fri, Jul 31, 2009 at 4:42 PM, Bjoern Peters <bp...@li...> wrote:

> The 'capacity' in 'animal feed capacity' was only part of the label, in
> order to identify the 'realizable entity' parent of 'animal feed role' and
> 'animal feed function'.
>
> If you want to get anal about the inconsistencies in the RO definition of
> realizes and our extension of it to other realizable entities (which we have
> done for years), then I do give up now.
>
> Frank Gibson wrote:
>
>>
>>
>> On Fri, Jul 31, 2009 at 12:00 AM, Melanie Courtot <mco...@gm...<mailto:
>> mco...@gm...>> wrote:
>>
>>
>>    I think I don't get the difference between the proposed
>>    planned_realization_of and realizes.
>>
>>
>> I would agree with you here Melanie. We seem to be on slightly differenty
>> wavelengths here Bjoern. We can only build against the version of BFO that
>> we have at the minute - to do anything else is a recipie for disaster. These
>> have specfic defintions for what a realizable_entiy is and how is is
>> realized. You seem to be talking about advanced versions of BFO that inlcude
>> capacity - I have no idea what capacity is. What I describe above is how to
>> model the example currently, and what we can not do currently. I think we
>> need these boundaries to get version 1 and the paper out.
>>
>>
>>
>>
>>    definition of realizes: "Relation between a process and a function,
>>    where the unfolding of the process requires the execution of the
>>    function. Class level: P realizes F iff: given any p that instantiates
>>    P, there exists some f, t such that f instantiates F at t and p
>>    *realizes* f. Here, *realizes* is the primitive instance level
>>    relation [GOC:cjm]"
>>
>>    This probably needs to be updated to say function and roles, which
>>    would match the range realizable_entity.
>>
>>
>> Or all realizable entities? As Melanie points out here, some thing is not
>> correct here, either the definition or the domain and range.
>>
>>
>>
>>
>>
>>
>>    By using this relation with planned processes we would get the plan
>>    specification for free, as planned_process is defined as realization
>>    of concretization of a plan.
>>
>>
>> Which is not the current textual definition of the realizes relation.
>>
>>
>>
>>    Or were you suggesting to create a new relation to restrain the
>>    current domain of realizes (i.e. process) to only planned processes?
>>
>>
>> I think we have overloaded the realizes relation here, we need some
>> clarification on what is actually is for, improving the textual def and the
>> domain and range. Keeping in mind we already have the inheres_in relation
>> which links realizable entities with their bearers - maybe the realizes
>> relation should have the domain realizable_entity_bearer (some material
>> bearing a realizable entity) and the range process. Or something else
>> completely different........
>>
>>
>>    ps: I think Phil's proposal was to assert directly under realizable
>>    entities.
>>
>>
>> It was
>>
>>
>>
>> Frank
>>
>>
>>
>>
>>
>>
>>
>>
>>    On 30-Jul-09, at 3:45 PM, Bjoern Peters wrote:
>>
>>    > My proposal would have been to just use the 2nd relationship for
>>    > simplicity, even if it is not clear that it is referenced by a plan.
>>    >
>>    > Just now I am thinking:
>>    > How about a relation planned_realization_of' (crappy label for
>>    > now)=def: a relationship between a planned process and a realizable
>>    > dependent continuant for which the planned process realizes the
>>    > realizable dependent continuant, and for which the plan
>>    > specification associated with the planned process specifies the
>>    > realization of the dependent continuant.
>>    >
>>    > Then we can say:
>>    >
>>    > animal feeding planned_realization_of some feed capacity
>>    >
>>    > - Bjoern
>>    >
>>    >
>>    >
>>    >
>>    >
>>    > ----- Original Message -----
>>    > From: "Philippe Rocca-Serra" <ro...@eb...
>>    <mailto:ro...@eb...>>
>>    > To: "Bjoern Peters" <bp...@li... <mailto:bp...@li...>>
>>    > Cc: "Frank Gibson" <fg...@gm...
>>    <mailto:fg...@gm...>>, "OBI Developers"
>>    <obi...@li...
>>    <mailto:obi...@li...>
>>    > >, "Protocol App Branch"
>>    <obi...@li...
>>    <mailto:obi...@li...>
>>    > >
>>    > Sent: Thursday, July 30, 2009 10:30:32 AM GMT -08:00 US/Canada
>>    Pacific
>>    > Subject: Re: [Obi-protocol-application-branch] Protocol Application
>>    > review: consistency/style of restriction
>>    >
>>    > That was my point.
>>    > We need both statement to fully qualify the process we are
>>    describing.
>>    > If we all agree on this, (please chip in if you disagree) then
>>    we can
>>    > start correcting /adjusting the restrictions of a number of
>>    > processes (I
>>    > have already identified 12 classes and was not even half way
>>    > through ).
>>    >
>>    > Cheers
>>    >
>>    > P
>>    >
>>    >
>>    >
>>    >
>>    > Bjoern Peters wrote:
>>    >> You cannot infer from an entity bearing a role during a process
>>    that
>>    >> the role is realized in that process, at least according to Barry's
>>    >> definitions. Anyway, I was ignoring role / function differences in
>>    >> this discussion, in anticipation of replacing it with the solution
>>    >> presented by Phil at the workshop. 'feed role' here is a good
>>    example
>>    >> why we need that, as many of the things being fed to lab
>>    animals were
>>    >> manufactured for that purpose, and therefore would have a 'feed
>>    >> function' from the start of their existence. I forgot if we
>>    agreed on
>>    >> how to call the parent entities of function / role , so I will
>>    >> rewrite
>>    >> using 'capacity'
>>    >>
>>    >> animal feeding has_specified_participant some material entity and
>>    >> bearer_of some feed capacity
>>    >> AND
>>    >> animal feeding realizes feed capacity borne by some material entity
>>    >>
>>    >> What we can't say in OWL is that the same material entity is
>>    referred
>>    >> to. Even if this may not be a problem for role if you believe every
>>    >> time a material has a role it also realizes that role, it is a
>>    >> problem
>>    >> for functions. A process realizing the plann to use a bag of
>>    >> manufactured animal feed as a door stopper would fulfill the first
>>    >> statement above. The second statement avoids that, but we don't
>>    get a
>>    >> reference to the plan.
>>    >>
>>    >> - Bjoern
>>    >>
>>    >>
>>    >>
>>    >> Frank Gibson wrote:
>>    >>>
>>    >>>
>>    >>> On Thu, Jul 30, 2009 at 4:55 PM, Bjoern Peters
>>    <bp...@li... <mailto:bp...@li...>
>>    >>> <mailto:bp...@li... <mailto:bp...@li...>>> wrote:
>>    >>>
>>    >>>    Frank: that is not the point. The individual statements are
>>    >>>    trivial. The problem is
>>    >>>    to logically exactly say that an entity is a specified
>>    >>> participant
>>    >>>    of a process AND the role it realizes in that process..
>>    >>>
>>    >>>
>>    >>> You cant - and it should not be something you want to do. The
>>    >>> material doesnt realise a role, it bears it (has_role), for the
>>    >>> duration of the process. Becasue is bears the role and the process
>>    >>> complete the role has been realized by the process, which includes
>>    >>> the participants. I admit the BFO defintion of role is not very
>>    >>> helpful - there is a surprise...
>>    >>>
>>    >>> You can ask what role is it bearing for the duration of the
>>    process
>>    >>> and infer it was realised. You cant ask your statement above.
>>    >>>
>>    >>> Frank
>>    >>>
>>    >>>
>>    >>>
>>    >>>
>>    >>>
>>    >>>
>>    >>>
>>    >>>    ----- Original Message -----
>>    >>>    From: "Frank Gibson" <fg...@gm...
>>    <mailto:fg...@gm...>
>>    >>> <mailto:fg...@gm... <mailto:fg...@gm...>>>
>>    >>>    To: "Philippe Rocca-Serra" <ro...@eb...
>>    <mailto:ro...@eb...>
>>    >>> <mailto:ro...@eb... <mailto:ro...@eb...>>>
>>    >>>    Cc: "Bjoern Peters" <bp...@li...
>>    <mailto:bp...@li...> <mailto:bp...@li...
>>    <mailto:bp...@li...>>>,
>>    >>>    "OBI Developers" <obi...@li...
>>    <mailto:obi...@li...>
>>    >>>    <mailto:obi...@li...
>>    <mailto:obi...@li...>>>, "Protocol App Branch"
>>    >>>    <obi...@li...
>>    <mailto:obi...@li...>
>>    >>>       <mailto:
>> obi...@li...
>>    <mailto:obi...@li...>>>
>>    >>>    Sent: Thursday, July 30, 2009 8:29:48 AM GMT -08:00 US/Canada
>>    >>> Pacific
>>    >>>    Subject: Re: [Obi-protocol-application-branch] Protocol
>>    >>>    Application review:  consistency/style of restriction
>>    >>>
>>    >>>
>>    >>>    On Thu, Jul 30, 2009 at 4:11 PM, Philippe Rocca-Serra <
>>    >>>    ro...@eb... <mailto:ro...@eb...>
>>    <mailto:ro...@eb... <mailto:ro...@eb...>> > wrote:
>>    >>>
>>    >>>
>>    >>>    Frank, Bjoern
>>    >>>
>>    >>>    +1
>>    >>>
>>    >>>    Frank, I believe we still need to state process 'animal feed'
>>    >>>    realizes feed_role
>>    >>>
>>    >>>    I think you need to have this relation on the role so the.
>>    >>>    feed_role is_realized_by some animal_feeding_process
>>    >>>
>>    >>>    Frank
>>    >>>
>>    >>>
>>    >>>
>>    >>>
>>    >>>
>>    >>>
>>    >>>    since the has_specified_input relation does not imply anything
>>    >>>    about the realization of a role born by the continuant being
>>    >>> specified
>>    >>>
>>    >>>    if this assertion is wrong then the has_specified_input
>>    relation
>>    >>>    needs to be clarified and an example supplied
>>    >>>
>>    >>>    Cheers
>>    >>>
>>    >>>    P
>>    >>>
>>    >>>    Frank Gibson wrote:
>>    >>>
>>    >>>
>>    >>>
>>    >>>
>>    >>>
>>    >>>
>>    >>>
>>    >>>    On Thu, Jul 30, 2009 at 3:46 PM, Bjoern Peters <
>>    bp...@li... <mailto:bp...@li...>
>>    >>>    <mailto:bp...@li... <mailto:bp...@li...>>
>>    <mailto: bp...@li... <mailto:bp...@li...>
>>    >>>    <mailto:bp...@li... <mailto:bp...@li...>> >> wrote:
>>    >>>
>>    >>>
>>    >>>> *animal feeding* has currently the following restriction:
>>    >>>    realizes some
>>    >>>> 'feed role' , which should probably be replaced by:
>>    >>>>
>>    >>>    I believe the above is the best we can do
>>    >>>
>>    >>>
>>    >>>    Ok, so the class as it is is wrong, it probably shoudl be
>>    >>>
>>    >>>    animal feeding
>>    >>>    has_specified_input some organism
>>    >>>    has_specified_input some (material has_role feed_role)
>>    >>>    has_specified_ouput some fatter_organism
>>    >>>
>>    >>>    The output here has to indicate that the organims has been fed.
>>    >>>
>>    >>>    Frank
>>    >>>
>>    >>>
>>    >>>
>>    >>>
>>    >>>
>>    >>>> realizes some ('feed role' and (inheres_in some
>>    >>>    'material_entity'))....
>>    >>>>
>>    >>>>
>>    >>>    Not sure what the .. is. If you want to connect it to
>>    >>>    'is_specified_participant', that won't work.
>>    >>>> or should it be
>>    >>>>
>>    >>>> has_specified_input some (material_entity and (realizes some
>>    >>>    'feed role'))
>>    >>>>
>>    >>>>
>>    >>>    This does not work, as 'realizes' needs a process on the left
>>    >>> hand
>>    >>>    side, not a material entity. What does work is 'has_role'.
>>    >>>> Classifier likes either but I guess it is more than a simple
>>    >>>    question of style. Bjoern had pointed to frequent abuse of
>>    >>>    'has_specified_input' relation.
>>    >>>> I have found myself confronted with this issue with several
>>    >>>    classes already (e.g. elution...)
>>    >>>> s already (e.g. elution with a restriction such as
>>    >>>    has_specified_input some ('{molecular entities, molecular
>>    >>> entity}'
>>    >>>    and (has_role some 'solvent role')))
>>    >>>>
>>    >>>>
>>    >>>
>>    >>>    I agree that we need to have one consistent way of modeling
>>    >>> this. I
>>    >>>    don't believe we can logically exactly say that an entity is a
>>    >>>    specified
>>    >>>    participant of a process and the role it realizes.
>>    >>>>
>>    >>>    --
>>    >>>    Bjoern Peters
>>    >>>    Assistant Member
>>    >>>    La Jolla Institute for Allergy and Immunology
>>    >>>    9420 Athena Circle
>>    >>>    La Jolla, CA 92037, USA
>>    >>>    Tel: 858/752-6914
>>    >>>    Fax: 858/752-6987
>>    >>>    http://www.liai.org/pages/faculty-peters
>>    >>>
>>    >>>
>>    >>>
>>    >>>
>>
>>  ------------------------------------------------------------------------------
>>    >>>
>>    >>>    Let Crystal Reports handle the reporting - Free Crystal Reports
>>    >>>    2008 30-Day
>>    >>>    trial. Simplify your report design, integration and
>>    deployment -
>>    >>>    and focus on
>>    >>>    what you do best, core application coding. Discover what's new
>>    >>> with
>>    >>>    Crystal Reports now. http://p.sf.net/sfu/bobj-july
>>    >>>    _______________________________________________
>>    >>>    Obi-protocol-application-branch mailing list
>>    >>>    Obi...@li...
>>    <mailto:Obi...@li...>
>>    >>>       <mailto:
>> Obi...@li...
>>    <mailto:Obi...@li...>>
>>    >>>    <mailto:
>>    Obi...@li...
>>    <mailto:Obi...@li...>
>>    >>>       <mailto:
>> Obi...@li...
>>    <mailto:Obi...@li...>> >
>>    >>>
>>    >>>
>>    >>>
>>
>> https://lists.sourceforge.net/lists/listinfo/obi-protocol-application-branch
>>    >>>
>>    >>>
>>    >>>
>>    >>>
>>    >>>
>>    >>>    --
>>    >>>    Frank Gibson, PhD
>>    >>>    http://peanutbutter.wordpress.com/
>>    >>>
>>    >>>
>>    >>>
>>    >>>
>>    >>>
>>    >>>    --
>>    >>>    Philippe Rocca-Serra, PhD
>>    >>>
>>    >>>    Technical Coordinator
>>    >>>    www.ebi.ac.uk/net-project
>>    <http://www.ebi.ac.uk/net-project> <http://www.ebi.ac.uk/net-project>
>>    >>>
>>    >>>    The European Bioinformatics Institute email:
>>    ro...@eb... <mailto:ro...@eb...>
>>    >>>    <mailto:ro...@eb... <mailto:ro...@eb...>>
>>    >>>    EMBL Outstation - Hinxton direct: +44 (0)1223 492 553
>>    >>>    Wellcome Trust Genome Campus fax: +44 (0)1223 492 620
>>    >>>    Cambridge CB10 1SD, UK room: A3-141
>>    >>>    --
>>    >>>
>>    >>>
>>    >>>
>>    >>>
>>    >>>    --
>>    >>>    Frank Gibson, PhD
>>    >>>    http://peanutbutter.wordpress.com/
>>    >>>
>>    >>>
>>    >>>
>>    >>>
>>    >>> --
>>    >>> Frank Gibson, PhD
>>    >>> http://peanutbutter.wordpress.com/
>>    >>
>>    >>
>>    >
>>    >
>>    > --
>>    > Philippe Rocca-Serra, PhD
>>    >
>>    > Technical Coordinator
>>    > www.ebi.ac.uk/net-project <http://www.ebi.ac.uk/net-project>
>>    >
>>    > The European Bioinformatics Institute   email:  ro...@eb...
>>    <mailto:ro...@eb...>
>>    > EMBL Outstation - Hinxton               direct: +44 (0)1223 492 553
>>    > Wellcome Trust Genome Campus            fax: +44 (0)1223 492 620
>>    > Cambridge CB10 1SD, UK                  room: A3-141
>>    > --
>>    >
>>    >
>>    >
>>
>>  ------------------------------------------------------------------------------
>>    > Let Crystal Reports handle the reporting - Free Crystal Reports 2008
>>    > 30-Day
>>    > trial. Simplify your report design, integration and deployment - and
>>    > focus on
>>    > what you do best, core application coding. Discover what's new with
>>    > Crystal Reports now.  http://p.sf.net/sfu/bobj-july
>>    > _______________________________________________
>>    > Obi-devel mailing list
>>    > Obi...@li...
>>    <mailto:Obi...@li...>
>>    > https://lists.sourceforge.net/lists/listinfo/obi-devel
>>
>>    ---
>>    Mélanie Courtot
>>    TFL- BCCRC
>>    675 West 10th Avenue
>>    Vancouver, BC
>>    V5Z 1L3, Canada
>>
>>
>>
>>
>>
>>
>>  ------------------------------------------------------------------------------
>>    Let Crystal Reports handle the reporting - Free Crystal Reports
>>    2008 30-Day
>>    trial. Simplify your report design, integration and deployment -
>>    and focus on
>>    what you do best, core application coding. Discover what's new with
>>    Crystal Reports now.  http://p.sf.net/sfu/bobj-july
>>    _______________________________________________
>>    Obi-protocol-application-branch mailing list
>>    Obi...@li...
>>    <mailto:Obi...@li...>
>>
>> https://lists.sourceforge.net/lists/listinfo/obi-protocol-application-branch
>>
>>
>>
>>
>> --
>> Frank Gibson, PhD
>> http://peanutbutter.wordpress.com/
>>
>
>
> --
> Bjoern Peters
> Assistant Member
> La Jolla Institute for Allergy and Immunology
> 9420 Athena Circle
> La Jolla, CA 92037, USA
> Tel: 858/752-6914
> Fax: 858/752-6987
> http://www.liai.org/pages/faculty-peters
>

-- 
Frank Gibson, PhD
http://peanutbutter.wordpress.com/

2007	Jan	Feb (6)	Mar (76)	Apr (28)	May (48)	Jun (28)	Jul (11)	Aug (15)	Sep (10)	Oct (7)	Nov (16)	Dec (12)
2008	Jan (11)	Feb (25)	Mar (18)	Apr (35)	May (51)	Jun (7)	Jul (11)	Aug (12)	Sep (11)	Oct (52)	Nov (64)	Dec (25)
2009	Jan (26)	Feb (23)	Mar (3)	Apr (5)	May (2)	Jun	Jul (39)	Aug (1)	Sep	Oct	Nov	Dec

obi-protocol-application-branch Mailing List for OBI

obi-protocol-application-branch — Discussion of terms that are in the Protocol Application branch of OBI.