Wiki Page: Clarification of BFO Process Definition using a wide variety of use cases

Hi,

 

I have started a wiki page on this topic under:

 

http://esw.w3.org/topic/HCLS/OntologyTaskForce/BFOProcessDefinitionDiscussion

 

Please feel free to add various use cases, examples and definitions related to
"Process" on this web page.

 

Cheers,

 

---Vipul

 

=======================================

Vipul Kashyap, Ph.D.

Senior Medical Informatician

Clinical Informatics R&D, Partners HealthCare System

Phone: (781)416-9254

Cell: (617)943-7120

http://www.partners.org/cird/AboutUs.asp?cBox=Staff&stAb=vik

 

To keep up you need the right answers; to get ahead you need the right questions

---John Browning and Spencer Reiss, Wired 6.04.95

________________________________

From: William Bug [mailto:William.Bug@DrexelMed.edu] 
Sent: Monday, June 04, 2007 4:27 PM
To: bfo-discuss@googlegroups.com
Cc: Alan March; Boris Hennig; Pierre Grenon; michael.f.uschold@boeing.com;
Kashyap, Vipul; Alan Ruttenberg; Holger Stenzhorn
Subject: Re: [bfo-discuss] Re: Questions regarding processes

 

Hi Vipul,

 

I do think it is a good suggestion to have a wide variety of domains covered in
the Use Cases worked through using BFO.  Even when restricted to the overall
domain of biomedicine - as we know from work on W3C SW HCSL IG - different
global communities tend to model reality in widely divergent ways.  One need
only compare/contrast EMR models and the sort of models created specifically to
support clinical trials whose data model is somewhat prescribed by the
legislation that drives the clinical trials process.

 

We also have all been learning the hard way that "modeling" per se - in the OMG
& UML sense of the word - is a process that gives rise to artifacts that many
times do not provide the sort of coherent, well-founded, and
computationally-accessible structure many seek to achieve when constructing
ontological frameworks for representing biomedical reality.  It is precisely in
the more complex inter-relations - such as formally describing complex relations
between entities in time and space (e.g., mereotopological relations amongst
independent continuants constrained to specific sites in time, functional
inter-relations in space & time that collectively represent processes) - where
these different approaches - software engineering oriented modeling (i.e., UML)
and ontological representation - are often not commensurate and require
significant computational "glue" to make them interoperate.

 

Back to our topic of "process".  Thank you for the web references.  This helps
to make your questions a bit more clear.

 

I realize it's not fair to pillory a Wikipedia entry (especially on a page
clearly marked as "requiring cleanup to meet Wikipedia's quality standards"),
and that's not really my intent here; however, I believe the first two sentences
provide an excellent example of where some of this confusion arises between
software/hardware processes in the domain of computer science and the more
general occurrent "process" as is described in several foundational ontologies -
but in specific - as it is defined for bfo:process:

 

"In computing, a process is an instance of a computer program that is being
executed. While a program itself is just a passive collection of instructions, a
process is the actual execution of those instructions."

 

In the first sentence, "process" is an instance of an independent continuant (a
computer program - whether referred to in it's source, object/machine, or binary
code form).  In the second sentence, they go on to make a distinction between
this independent continuant instance, and the true, underlying "process" in the
bfo:process sense.  

 

As I see it, for this domain, the "computational_process" (subclass of
bfo:process) could simply be described as the:

          "a type of process where a series of machine-level instructions are
executed sequentially"

 

We could go into a lot more detail about branching, branch prediction, spawning
& reclaiming child processes, the details of multi-process semaphore signaling,
etc. - and would have to, if one was to truly represent what is going on in a
modern computer operating system; however, at the most basic level, I believe
that would be pretty close to the sort of bfo:process subclass required in this
domain.  One might argue about the "sequentially" portion of this definition,
but as nearly all of us are working with von Neumann-style Turing Machines, I'm
reasonably comfortable with that high-level definition as a reasonable first
pass.

 

Creating a consensus, human readable text definition of a computational_process
- one that hopefully is unambiguous and adopts a genera/species/diffentia style
- is very important.  This is the primary reason we are all spending time
discussing some of the definitions given in the current BFO OWL file.  Several
of them are problematic and require some work - and supporting examples.  This
is not news to any of the BFO developers**.  However, given the foundational
role BFO is intended to provide, such changes need to be made with proper
consideration of all the expected applications.  This is a part of what I
believe you are requesting when you suggest creating a Wiki on this topic.

 

However, it is really as we extend beyond the definition to work through clear
examples/use cases - IN OWL - that the true computational value of a
foundational ontology such as BFO becomes more clear.  For instance, one would
(or I would) really want to have a graphical depiction of how the various
independent and dependent continuants (e.g., CPU registers, program counters,
execution & branch units, instructions, instruction sets, execution state, I/O
lines, etc.) inter-relate to participate in a computational_process.

 

There has certainly been considerable, ongoing debate on this list (BFO-Discuss)
discussing bfo:process.  My sense is now would be an excellent for us to create
a summary of these discussions - we being both the BFO developers - and those
ontological engineers who've been trying to assemble bfo-derived representations
in a variety of domains - and to provide more fully-realized examples - both in
text and graphical form.  Having this in place first would greatly expedite
getting useful feedback on where BFO does - and does not - meet the needs of a
variety of ontological representation applications.  Some of these examples can
been culled from existing, published manuscripts.  Having read a good deal of
the existing manuscripts, I'm pretty convinced the only way to really expedite
this sort of effort would be to extract these examples - in their simplest form
- and supplemented by graphical depictions - to a web site or Wiki where
potential users of BFO would be able to greatly accelerate their understanding
of BFO and what it's designed to do - and decrease the steep and lengthy
learning curve required to to effectively use BFO now.  It would also expedite
our collective effort to vet BFO and determine where it needs adjustments,
extensions, or amendments.

 

Just my $0.02 on this issue.

 

Cheers,

Bill

 

** I added the lead BFO OWL developer - Holger Stenzhorn - to this thread, as
these are issues he's been giving lots of thought to.

          

 

 

On Jun 4, 2007, at 11:06 AM, Kashyap, Vipul wrote:





Bill and Barry,

 

Thanks for the various clarifications around different types of processes.

 

http://www.aiai.ed.ac.uk/project/enterprise/enterprise/ontology.html
<http://www.aiai.ed.ac.uk/project/enterprise/enterprise/ontology.html> 

 

http://en.wikipedia.org/wiki/Process%28computing%29
<http://en.wikipedia.org/wiki/Process%28computing%29> 

 

IMHO, I would consider the definitions from above as important use cases that

could help clarify and validate the BFO. Clearly the kinds of processes

enumerated above are different from the kinds of processes identified (as of
yet)

in biology.

 

At the same time, this may be one way of validating the BFO structure and maybe

refining the various notions of process etc.

 

Was wondering if there was some interest in articulating the definitions of a
process

from the viewpoints of distributed computation, biology and AI (planning) and
coming up

with a comprehensive framework for the same.

 

One interesting outcome of this could be a clear identification of points of
difference

between a biological process and other types of processes (distributed
computation, AI).

 

If there is interest in this, I could create a wiki page and get this going?
What do you think Bill?

 

Michael: I have included you in this e-mail because you worked with Austin Tate
on his Planning

Ontologies and think you could clarify for us the notion of a "Process" in the
context of Barry's work

for BFO.

 

Cheers,

 

---Vipul

 

Here some random thoughts.

BFO is divided into two types of ontologies (sometimes called SNAP, 

for the continuants, and SPAN, for the occurrents). SNAP ontologies 

are always indexed to a time (like snapshots); SPAN ontologies 

embrace processes taking place within an interval of time (for 

example the entire lifetime of the universe). From the SPAN 

perspective we view processes timelessly (or, equivalently, 

fourdimensionally). From the SPAN perspective, if a process is telic, 

then it is telic from the start (or, preferably, atemporally); thus 

the whole process type is already instantiated by the very first 

phase. This seems to me not problematic from the BFO perspective. It 

begins to appear problematic only if one applies to occurrents 

expectations appropriate to continuants.

BS

 

I believe this was exactly the misunderstanding at the root of the discussion on
the W3C SemWeb HCLS list last week re: process in the computational domain -
i.e., and OS-level process.  Such a "process" is really an independent
continuant (not even a relalizable entity), contains the pointers to a piece of
binary code and roles related to its history as a running process (e.g.,
"parent", "child", "orphan", "zombie", etc.).  In fact, it really would be a
non-sequitur to describe a bfo:process as "orphan" or "zombie".

 

 

 

 

At 12:33 AM 6/1/2007, Alan Ruttenberg wrote:

 

 

	On May 31, 2007, at 12:23 PM, Alan March wrote:

		- How would Henning's "telic" and "non-telic" processes fit into

		BFO? Or are

		they implied somewhere in BFO and thus redundant? Or are they
just

		plain

		incompatible with the BFO model as a whole?

	 

	A telic process would be the realization of a realizable entity.

	 

		- Is Henning's "procedure" (defined as a complex telic process)

		necessary or

		would it suffice to assume that a process could have other

		processes as its

		parts and such a class is unneccesary?

	 

	I think basically, that procedures are the same as his telic

	processes, and that he

	lumps realizable entities, (e.g. function) in with their realizations

	(e.g. processes).

	 

	Can elaborate if this is too telegraphic.

	 

	-Alan

	 

		 

		I admit that I may probably getting the whole picture wrong out
of

		ignorance

		of BFO's "big picture" (or weltanschaung, if I may) but I
thought this

		group's the best place to seek for an answer and would be most

		grateful for

		your advice.

		 

		Regards

		 

		Alan

		 

		 

			 

	 

	 

	 

 

 

--~--~---------~--~----~------------~-------~--~----~

You received this message because you are subscribed to the Google Groups "BFO
Discuss" group.

To post to this group, send email to bfo-discuss@googlegroups.com
<mailto:bfo-discuss@googlegroups.com> 

To unsubscribe from this group, send email to
bfo-discuss-unsubscribe@googlegroups.com
<mailto:bfo-discuss-unsubscribe@googlegroups.com> 

For more options, visit this group at
http://groups.google.com/group/bfo-discuss?hl=en
<http://groups.google.com/group/bfo-discuss?hl=en> 

-~----------~----~----~----~------~----~------~--~---

 

 

Bill Bug

Senior Research Analyst/Ontological Engineer

 

Laboratory for Bioimaging  & Anatomical Informatics

www.neuroterrain.org

Department of Neurobiology & Anatomy

Drexel University College of Medicine

2900 Queen Lane

Philadelphia, PA    19129

215 991 8430 (ph)

610 457 0443 (mobile)

215 843 9367 (fax)

 

 

Please Note: I now have a new email - William.Bug@DrexelMed.edu
<mailto:William.Bug@DrexelMed.edu> 

 

 






 

The information transmitted in this electronic communication is intended only
for the person or entity to whom it is addressed and may contain confidential
and/or privileged material. Any review, retransmission, dissemination or other
use of or taking of any action in reliance upon this information by persons or
entities other than the intended recipient is prohibited. If you received this
information in error, please contact the Compliance HelpLine at 800-856-1983 and
properly dispose of this information.

 

Bill Bug

Senior Research Analyst/Ontological Engineer

 

Laboratory for Bioimaging  & Anatomical Informatics

www.neuroterrain.org

Department of Neurobiology & Anatomy

Drexel University College of Medicine

2900 Queen Lane

Philadelphia, PA    19129

215 991 8430 (ph)

610 457 0443 (mobile)

215 843 9367 (fax)

 

 

Please Note: I now have a new email - William.Bug@DrexelMed.edu

 

 





 






The information transmitted in this electronic communication is intended only for the person or entity to whom it is addressed and may contain confidential and/or privileged material. Any review, retransmission, dissemination or other use of or taking of any action in reliance upon this information by persons or entities other than the intended recipient is prohibited. If you received this information in error, please contact the Compliance HelpLine at 800-856-1983 and properly dispose of this information.