Hi:
This is
a later version of a portion of what was previously posted to Cory C’s
semantic core group on Yahoo.
By the
way, I enjoyed the Ontology Summit Workshop yesterday, and find that the
following GEM approach can probably be used to support item 8 of the Upper
Ontology Summit Joint Communique (15 March, 2006) by providing a “principled
means of articulating the relationships (including differences)” among
the various upper ontologies presented at the summit and those to be introduced
later. As described in previous ONTAC forum threads, it is my perception
that the ontology I’m describing below could serve as an orthogonal “backplane”
or “common-ground” for the upper, middle, and lower/detail ontologies
representing the various “world views” documented/architected as
ontologies. I know my “approach” is not “formal”
enough for some in the Ontolog community, but it can be with a bit of work on
the part of these communities. I’m not trained to achieve the “formalism”,
so any help is appreciated.
Note
that my approach is available under the “Creative Commons”
ShareAlike, Attributed, NonCommercial License, as are some of the other “public
ontologies” presented at the Summit, such as is, I believe, OpenCyc.
There is also a commercial version, also as is the case with Cyc. My approach
is currently licensed for share-alike, attributed, non-commercial use under the
above Creative Commons License to the U.S. Federal Executive Branch as a result
of using my approach in the previous Federal Continuity Communications
Enterprise Architecture (CCEA) project. My approach also provides what
has been sought by the larger Federal EA community – an “operational
enterprise architecture” with the capability of providing “virtual”
or “composite” applications from our architecture/ontology/knowledge
repository.
Roy
=================
From my
30+ years experience as a manager and management analyst with government and
then as a government contractor, I have discerned that everyone in and around
an endeavor needs to be able to ask and answer a whole series of basic
questions: what, how, where, why, who, when, with what, how many, etc. This
“world model” was used to build up a collection of basic, and
highly abstract, question and answer sets that eventually jelled into the
following subject types, representing the basic categories of
“subjects” of interest to managing any endeavor.
1.
Location subjects (i.e., where physically, conceptually, and
virtually are parts relevant in the cause-effect (or sequence, value-chain,
flow), categorization, containment, change, variance, equivalence, and
descriptive relations to the endeavor)
2.
Organization subjects (i.e., who is responsible for the
endeavor or parts within and around it)
3.
Organization unit subjects (i.e., who performs the work of
the endeavor or parts within and around it)
4.
Function (i.e., what work is performed by the endeavor or
parts within and around it)
5.
Process (i.e., the actions taken by an organization unit to
perform some function within a flow (i.e., sequence, customer-value-chain) of
functions intended to produce a specific resource as output, using specific
resources as input, constrained by other resources, and enabled by others, as
in the IDEF0 ICOM model)
6.
Resource (i.e., parts accounted for, i.e., controlled, by
the endeavor)
7.
Requirement (i.e., the quantity, qualities, and
schedule of input, constraint, and enabling resources required by the
endeavor’s organization unit and parent organization at a location to
perform a function’s process, to produce the endeavor’s functional
output-resource on schedule, on budget, and to the specified quality)
All the
above led me to develop the following endeavor management metaschema (with significantly
more detail within the four enterprise subcategories below):
- Enterprise (i.e., a purposeful enterprise)
- Catalogs (of Types and Instances of world parts)
(documented at OMG MDA Layer 2)
i.
Location
ii.
Organization
iii.
Organization Unit
iv.
Function (with a common strategic/operation/tactical
management metaschema and methodology documented here at the OMG MDA Layer 1)
v.
Process
vi.
Resource
vii.
Requirement
- Relation Types categorizing all specific relations
between all subjects-types and specific subjects (documented at the OMG
MDA Layer 3)
i.
Categorization
ii.
Containment
iii.
Sequence (with a 4d view of time/space flow represented here,
and further elaborated the “flow roles” below)
iv.
Change
v.
Variance
vi.
Equivalence
vii.
Description
- Flow (Sequence) Roles
i.
Customer
ii.
Performer
iii.
Supplier
iv.
Authority
v.
Subordinate
vi.
Outsource
vii.
Public
- Rules (Axioms) (bi-directional relations
implied) (Logic still under development)
i.
Location contains Organization
ii.
Organization forms Organization Unit
iii.
Organization Unit performs Function
iv.
Function applies Process
v.
Process Consumes Resources as Input
vi.
Process Produces Resources as Output
vii.
Process Is Constrained by Resources
viii.
Process Is Enabled by Resources
ix.
Resources Fulfill Requirements
x.
Requirements Are Endeavors
xi.
Mission Capabilities contain resources required on time, at
quantity, with qualities, for processes, to support functions, which are
assigned to organization units as functional missions, or are assumed/taken by
organizations as enterprise missions, across some location/area
This
then forms my idea of what is meant by a “semantic core”, or as I
described it in the various ONTAC fora as a “general ontology” that
describes how “the world works”, and as I previously called
it – a “general enterprise management (GEM) ontology” and a
“GEM architecture”. (I consider an
“architecture”, which identifies the “parts of a thing and
their relationships”, to be the information structure, or metaschema, of
an ontolog,y which I consider to be a newer form of a process model.)
I
believe that this particular model of a “semantic core” will
succeed in implementation and operation because it can be shown to encompass all
other ontologies and architectures because it begins with the premise that
there is only “one thing”, meaning there is no need for
“integration” of this semantic core with others because it
“includes” them (although their authors may not realize or accept
this as the case).
Roy