Ontology and Semantic Interoperability - Personal websites at UB
Integration, Interoperability and Collaboration: Semantic Interoperability and Ontology Systems...
-
Upload
walter-marsh -
Category
Documents
-
view
237 -
download
0
Transcript of Integration, Interoperability and Collaboration: Semantic Interoperability and Ontology Systems...
Integration, Interoperability and Collaboration:Semantic Interoperability and Ontology Systems
Integration, Interoperability and Collaboration:Semantic Interoperability and Ontology Systems
Fifth Semantic Interoperability for E-government Conference
MITREMcClean, VA
October 11, 2006
Presented byMatthew K. Hettinger, CEO and Chief Architect
Mathet Consulting, Inc.
Integrated, Interoperable and Collaborative Systems: Architecture and Engineering (IICSAE™)
August 15, 2006 Copyright (C) 2006 Mathet Consulting, Inc. 2
AbstractAbstract
• Characterization of the semantic interoperability problem– Levels of semantic interoperability– Measures– Collaborations
• Characterization and utilization of ontology and semantic interoperability systems to address the semantic interoperability problem
– Open world semantics, business rules, innovation– Use and implementation of standards– Interoperability systems– Systems-based architecture
• Relationships between semantic interoperability , ontology, information quality, service-oriented architecture and (inter-) enterprise system architecture including reference models
– Semantic Interoperability as a service– System integration, interoperability and collaboration– System emergence and downward (and upward) causation
• Using ontology and semantic interoperability systems (e.g. in government)– Mapping law / statute to enterprise services– Strategic, Tactical, Operational Levels of use– Inter-enterprise Collaborations
Inte
grat
ion,
In
tero
pera
bilit
y, a
nd C
olla
bora
tion:
Sem
antic
Int
ero
pera
bilit
y an
d C
olla
bora
tion
Sys
tem
s
Inte
grat
ion,
In
tero
pera
bilit
y, a
nd C
olla
bora
tion:
Sem
antic
Int
ero
pera
bilit
y an
d C
olla
bora
tion
Sys
tem
s
This presentation will address the following:
August 15, 2006 Copyright (C) 2006 Mathet Consulting, Inc. 3
Background and ContextBackground and Context
• General Systems Theory and the ‘Modeling Discipline’
• (Inter-) Enterprise Systems Architecture and Engineering
• Integration, Interoperability and Collaboration
Inte
grat
ion,
In
tero
pera
bilit
y, a
nd C
olla
bora
tion:
Sem
antic
Int
ero
pera
bilit
y an
d C
olla
bora
tion
Sys
tem
s
Inte
grat
ion,
In
tero
pera
bilit
y, a
nd C
olla
bora
tion:
Sem
antic
Int
ero
pera
bilit
y an
d C
olla
bora
tion
Sys
tem
s
Semantic Interoperability and Ontology Systems are presented from the perspective of:
applied to e-Government.
This presentation is modified from an earlier presentation at the Aug 15th Summer Expedition Workshop. It overlaps a discussion of SOA and EA onto Semantic Interoperability.
Semantic Interoperability Problem Characterization
Semantic Interoperability Problem Characterization
Integration, Interoperability and Collaboration:Semantic Interoperability and Ontology Systems
Integration, Interoperability and Collaboration:Semantic Interoperability and Ontology Systems
August 15, 2006 Copyright (C) 2006 Mathet Consulting, Inc. 5
Semantic Interoperability – Problem CharacterizationSemantic Interoperability – Problem Characterization
Semantics Interoperability
SemanticInteroperability
Working together at the semantic level of communication(s)The ability for two or more systems to interoperate at the semantic level
Requires an alignment of intended meaning and interpretation among interacting systems
Requires a semantic systems layer underlying and supporting communications- Requires Logic -
Working together- The ability of two or more systems to communicate and work together to coordinate and execute their respective services.
(The study of) Meaning in Language-The assignment of meaning to symbol sets Intended Meaning The expression of meaning with languages-The extraction / discovery of meaning from symbol sets Interpretation in a formal way -(especially for machine-machine interactions)
Inte
grat
ion,
In
tero
pera
bilit
y, a
nd C
olla
bora
tion:
Sem
antic
Int
ero
pera
bilit
y an
d C
olla
bora
tion
Sys
tem
s
Inte
grat
ion,
In
tero
pera
bilit
y, a
nd C
olla
bora
tion:
Sem
antic
Int
ero
pera
bilit
y an
d C
olla
bora
tion
Sys
tem
s
August 15, 2006 Copyright (C) 2006 Mathet Consulting, Inc. 6
System of Interest
Modeling Systems
Modeled Systems
Model Systems Modeled
Systems
System theoretic modification ofthe Semiotic Triangle(the Meaning Triangle) Product
(a subset of modeled systems)(including KID)
Resources(including KID)
Di Do
Sets of symbols (including terms), Sets of definitionsSets of relations and functions between / among symbols, definitions, things.Sets of languages (including OWL, ODM standards)axioms, rules, models, etc reflects a semiotic (semantic) state-space at any given point in time, the system, S is in a given semiotic (semantic) state
Systems of Interest has•Mission / Vision•Goals and Objectives•Role
S
Task KID In(collaboration)
Task KID Out
e.g. any government organization / agency- A government agency may be considered an enterprise system
Inte
grat
ion,
In
tero
pera
bilit
y, a
nd C
olla
bora
tion:
Sem
antic
Int
ero
pera
bilit
y an
d C
olla
bora
tion
Sys
tem
s
Inte
grat
ion,
In
tero
pera
bilit
y, a
nd C
olla
bora
tion:
Sem
antic
Int
ero
pera
bilit
y an
d C
olla
bora
tion
Sys
tem
sSemantic Interoperability – Problem CharacterizationSemantic Interoperability – Problem Characterization
August 15, 2006 Copyright (C) 2006 Mathet Consulting, Inc. 7
Semantic Interoperability – Problem CharacterizationSemantic Interoperability – Problem Characterization
Modeling Systems
Modeled Systems
Model Systems
Modeled Systems
Di
Di
Di
DO
DO
DO
Di
DO
Each (Sub) System reflects a semiotic (semantic) state-space
At any given point in time, each (sub) system is in a given semiotic (semantic) state
Each (Sub) System of Interest has•Mission / Vision•Goals and Objectives•Role
Product(a subset of
modeled systems)(including KID)
Resources(including
KID)
Task KID In
Task KID Out
Di
DO
A
B
Inte
grat
ion,
In
tero
pera
bilit
y, a
nd C
olla
bora
tion:
Sem
antic
Int
ero
pera
bilit
y an
d C
olla
bora
tion
Sys
tem
s
Inte
grat
ion,
In
tero
pera
bilit
y, a
nd C
olla
bora
tion:
Sem
antic
Int
ero
pera
bilit
y an
d C
olla
bora
tion
Sys
tem
s
1
23
4
1
-Step in process of data flow in a static or dynamic collaboration among 3 (sub) systems offering their (sub) services-- The collaboration reflects a virtual subsystem of S offering a service
S1
S2
S3
S
August 15, 2006 Copyright (C) 2006 Mathet Consulting, Inc. 8
Semantic Interoperability – Problem CharacterizationSemantic Interoperability – Problem Characterization
Data {Language: syntax, vocabulary}
Sender
Context’
Reciever
SenderReciever
System MSystem M’
Role N’Role N
Context
Conduit / Channel / MediaReceive
Data
ProcessData
Make Decisions
Take ActionsReceive
Data
ProcessData
Make Decisions
Take Actions
Data {Language: syntax, vocabulary}
Sender Reciever
SenderReciever
System MSystem M’
Role N’Role N
Context
Conduit / Channel / MediaReceive
Data
ProcessData
Make Decisions
Take ActionsReceive
Data
ProcessData
Make Decisions
Take Actions
A
B
Role System(May be human or machine)
KID
KID
Has temporal / spatial, pragmatic components
Inte
grat
ion,
In
tero
pera
bilit
y, a
nd C
olla
bora
tion:
Sem
antic
Int
ero
pera
bilit
y an
d C
olla
bora
tion
Sys
tem
s
Inte
grat
ion,
In
tero
pera
bilit
y, a
nd C
olla
bora
tion:
Sem
antic
Int
ero
pera
bilit
y an
d C
olla
bora
tion
Sys
tem
s
Provide a service
Provide a service
August 15, 2006 Copyright (C) 2006 Mathet Consulting, Inc. 9
Semantic Interoperability – Problem CharacterizationSemantic Interoperability – Problem CharacterizationIn
tegr
atio
n, I
nte
rope
rabi
lity,
and
Col
labo
ratio
n:S
eman
tic I
nter
ope
rabi
lity
and
Col
labo
ratio
n S
yste
ms
Inte
grat
ion,
In
tero
pera
bilit
y, a
nd C
olla
bora
tion:
Sem
antic
Int
ero
pera
bilit
y an
d C
olla
bora
tion
Sys
tem
s
Behavior
Percepts
Language(s): L*
Internal Models: M*(e.g interpretations)
Reasoning
Context
Data
Data
Context DataReceive Data
Take Actions
Internal Storage
e.g. Send Data
Sensors
Effectors
His
tori
cal
Kno
wle
dge,
In
form
atio
n,
Dat
aF
acts
Bel
iefs
Interactio
n
Know
ledge, Inform
ation, D
ataF
actsB
eliefs
TemporaryInternal Storage
In
Out
Has intended / expected meaningHas intended / expected value
TaskKID
TaskKID
Has temporal / spatial, pragmatic
components
KID
KID
KID
May be an Ontology/Semiotic
System
Receive KID Process KID Make Decisions Take Actions (provide a service)
August 15, 2006 Copyright (C) 2006 Mathet Consulting, Inc. 10
Semantic Interoperability – Problem CharacterizationSemantic Interoperability – Problem Characterization
Has role / set of responsibilities
Context
ContextContext
Interaction Interaction
Task KID In
Task KID Out
Di
DO
Di Di
DiDi
DO DO
DODO
System A System B
System C
General FlowEach (sub) system has role / set ofresponsibilities (e.g. in collaborations)Each (sub) system has a semantic state-space and is in a particular semantic state
C1
C2
Semantic interoperabilitybetween A and C-C1
Semantic interoperabilitybetween C-C2 and B
Semantic interoperability between C-C1 and C-C2
Inte
grat
ion,
In
tero
pera
bilit
y, a
nd C
olla
bora
tion:
Sem
antic
Int
ero
pera
bilit
y an
d C
olla
bora
tion
Sys
tem
s
Inte
grat
ion,
In
tero
pera
bilit
y, a
nd C
olla
bora
tion:
Sem
antic
Int
ero
pera
bilit
y an
d C
olla
bora
tion
Sys
tem
s
August 15, 2006 Copyright (C) 2006 Mathet Consulting, Inc. 11
Semantic Interoperability – Problem CharacterizationSemantic Interoperability – Problem Characterization
The KID product and its’ quality attributes, produced as a service by System C at the request of and for System B, is a function of (with associated errors):
• Input data from System A (data supplier) and System B (data customer: KID requirements)– Task KID (has intended meaning)– Context KID (has intended meaning)
• The data capture mechanisms• Recognition processes (evaluation against a set of symbols, definitions, languages that are
known) physical / intrinsic attributes of data
• Level 1 Interpretation (by system) process (an Interpretation Function): KID in context– The selected / derived interpretation– Knowledge of context (e.g..... Pragmatics: spatial-temporal, linguistic)– Knowledge of KID in context (e.g..... expectation)From a data perspective Information defined as (data + Level 1 interpretation the data has
been attributed (extrinsic attribute) with meaning, as a function of context)From a knowledge perspective, this information is knowledge
• Evaluation of data + Level 1 interpretation against expectation (as a function of role for KID / KID requirements)
• Level 2 Interpretation process (an Interpretation function): System (interpreter) in context– The selected / derived interpretation– Knowledge of context (e.g..... Pragmatics: social, epistemic)– Knowledge of self in context (role, purpose, responsibilities, perspectives, focal points) –
significance of information to self and to super-systems – that data has been attributed (extrinsic attribute) with meaning.
– Knowledge (from data perspective) = information + level 2 interpretation (signification)• Evaluation / validation of interpretation against role
Inte
grat
ion,
In
tero
pera
bilit
y, a
nd C
olla
bora
tion:
Sem
antic
Int
ero
pera
bilit
y an
d C
olla
bora
tion
Sys
tem
s
Inte
grat
ion,
In
tero
pera
bilit
y, a
nd C
olla
bora
tion:
Sem
antic
Int
ero
pera
bilit
y an
d C
olla
bora
tion
Sys
tem
s
August 15, 2006 Copyright (C) 2006 Mathet Consulting, Inc. 12
Semantic Interoperability – Problem CharacterizationSemantic Interoperability – Problem Characterization
• Data product production (action taken) and quality is a function of: (output data is attributed with an intended meaning)
– Interpretation of input data requirements, and purpose (where appropriate) from receiver (e.g..... customer)
– Context of the data production system– Knowledge of the context (Level 1 interpretation of input data)– Knowledge of self (role, etc.) (Level 2 interpretation of input data)– Decision making (e.g..... what KID to produce and how to produce it)
• Time (any of the above may change with time, e.g..... Learning)
• For successful semantic interoperability:– The attributed intended meaning by the sender (a service provider), at
level 1 and level 2 interpretation levels, must “match” the attributed level 1 and level 2 interpretations of the receiver (a service requestor)
• Due to the nature of the nature of meaningful exchange of KID, especially between machines, logic is required
• To ensure semantic interoperability, there must be measures.• It is noted that the level 1 and level 2 interpretations are extrinsic attributes of KID as opposed
to intrinsic (those attributes that are interaction context (including interpreter) independent. The quality of KID is a function of the intrinsic and extrinsic attributes.
• Successful interaction requires level 1 semantic interoperability, successful collaboration (common purpose / goals / objectives, etc.) requires level 2 semantic interoperability
Inte
grat
ion,
In
tero
pera
bilit
y, a
nd C
olla
bora
tion:
Sem
antic
Int
ero
pera
bilit
y an
d C
olla
bora
tion
Sys
tem
s
Inte
grat
ion,
In
tero
pera
bilit
y, a
nd C
olla
bora
tion:
Sem
antic
Int
ero
pera
bilit
y an
d C
olla
bora
tion
Sys
tem
s
August 15, 2006 Copyright (C) 2006 Mathet Consulting, Inc. 13
• For any group of systems to collaborate, to interact with a purpose, to meet goals and objectives, and produce data products as a group, it is required that there be a shared understanding of what the data resources mean
• The shared understanding may be created dynamically during the interaction – semantic negotiation, mediation, arbitration during a task with the use of an ontology system
• The shared understanding may be created a priori and is “static” before an interaction – a shared ontology or ontologies with mappings between them
• How does one know there is a shared understanding? - Measures
Semantic Interoperability – Problem CharacterizationSemantic Interoperability – Problem CharacterizationIn
tegr
atio
n, I
nte
rope
rabi
lity,
and
Col
labo
ratio
n:S
eman
tic I
nter
ope
rabi
lity
and
Col
labo
ratio
n S
yste
ms
Inte
grat
ion,
In
tero
pera
bilit
y, a
nd C
olla
bora
tion:
Sem
antic
Int
ero
pera
bilit
y an
d C
olla
bora
tion
Sys
tem
s
August 15, 2006 Copyright (C) 2006 Mathet Consulting, Inc. 14
Semantic Distance / Similarity and Matching
Semantic Interoperability – Problem CharacterizationSemantic Interoperability – Problem Characterization
X = some KIDSupplier
InformationAnd
Management
CustomerInformation
And Management
X’1 = …..X’2 = …..X’3 = …..
E.g. Defense Logistics Agency – an Enterprise System
IndustrialService
Providers
-Information –- resources –
- money -
Customere.g.
War fighterSystem
X = (KID, intrinsic attributes, intended level 1 meaning, intended level 2 meaning)
X’ = (KID, intrinsic attributes, level 1 interpretation, level 2 interpretation)
A set of possible responsesBased on the set of interpretationsavailable (already known or derivable)
X = X’To what degree are these equivalent?Which alternative is best?Is there a match?
Note: Algorithms do not represent complete structures
Inte
grat
ion,
In
tero
pera
bilit
y, a
nd C
olla
bora
tion:
Sem
antic
Int
ero
pera
bilit
y an
d C
olla
bora
tion
Sys
tem
s
Inte
grat
ion,
In
tero
pera
bilit
y, a
nd C
olla
bora
tion:
Sem
antic
Int
ero
pera
bilit
y an
d C
olla
bora
tion
Sys
tem
s
August 15, 2006 Copyright (C) 2006 Mathet Consulting, Inc. 15
Semantic Interoperability – Problem CharacterizationSemantic Interoperability – Problem Characterization
Semantic Distance / Similarity and Matching
X = ‘Customer’
Sender (in context)Set of possible interpretations
Receiver (in context)Set of possible interpretations
Set of possible interpretations in common
Interpretations are a function of context
•Each set of possible interpretations is constrained by the semantic state of each system.•Each semantic state is a function of context.•As contexts increasingly overlap, the semantic states overlap and the set of possible common interpretations increases. •The conditions / situation / constraints of the interaction reduces the possible number common of interpretations
A measure of semantic distance / similarity is a measure of the semantic difference between sets of concepts (e.g.... X and X’).A semantic match is when distance is zero.
SupplierSystems
Customer Systems
X = (KID, intrinsic attributes, intended level 1 meaning, intended level 2 meaning)
X’ = (KID, intrinsic attributes, level 1 interpretation, level 2 interpretation)
Inte
grat
ion,
In
tero
pera
bilit
y, a
nd C
olla
bora
tion:
Sem
antic
Int
ero
pera
bilit
y an
d C
olla
bora
tion
Sys
tem
s
Inte
grat
ion,
In
tero
pera
bilit
y, a
nd C
olla
bora
tion:
Sem
antic
Int
ero
pera
bilit
y an
d C
olla
bora
tion
Sys
tem
s KID
August 15, 2006 Copyright (C) 2006 Mathet Consulting, Inc. 16
• Example Scenarios – There may be no common interpretations – failure– There may a set of common interpretations but the semantic
distance between those of the sender (e.g.... marketing) and those of the receiver may be greater than some criteria – failure
– There may be set of interpretations of the receiver that are acceptable to receiver and sender – workable
– There may be a match - success• Interactions may occur many times with differing results as the
interpretations may vary – one form of semantic variance.• The sender (service provider, e.g. industrial enterprise) and
receiver (service requestor, e.g. supplier management system) interact with N and M other systems respectively. The same kinds of results occur with these interactions. There may be Z (sub) systems. Variance at the system (e.g. enterprise) level is some function of variance on individual interactions – enterprise semantic variance.
• There are many semantic measures that may be implemented to ensure successful collaboration – see appendix
Semantic Interoperability – Problem CharacterizationSemantic Interoperability – Problem CharacterizationIn
tegr
atio
n, I
nte
rope
rabi
lity,
and
Col
labo
ratio
n:S
eman
tic I
nter
ope
rabi
lity
and
Col
labo
ratio
n S
yste
ms
Inte
grat
ion,
In
tero
pera
bilit
y, a
nd C
olla
bora
tion:
Sem
antic
Int
ero
pera
bilit
y an
d C
olla
bora
tion
Sys
tem
s
Characterization and Utilization of Ontology and Semantic Interoperability Systems to Address the
Semantic Interoperability Problem
Characterization and Utilization of Ontology and Semantic Interoperability Systems to Address the
Semantic Interoperability Problem
Integration, Interoperability and Collaboration:Semantic Interoperability and Ontology Systems
Integration, Interoperability and Collaboration:Semantic Interoperability and Ontology Systems
Relationships between semantic interoperability , ontology, information quality, service-oriented
architecture and (inter-) enterprise system architecture including reference models
Relationships between semantic interoperability , ontology, information quality, service-oriented
architecture and (inter-) enterprise system architecture including reference models
August 15, 2006 Copyright (C) 2006 Mathet Consulting, Inc. 18
Ontology DefinitionA logical theory accounting for the intended meaning of a formal vocabulary of a domain, i.e. it’s ontological commitment to a particular conceptualization (a system of categories) of the world (N. Guarino, 1995) [underline, italics, and ( ) added]
– From a (general) systems theoretic perspective of the Meaning Triangle, an ontology represents a semiotic (semantic) system state, which in turn may be used to represent a domain of discourse of a set of external systems.
– Both sender and receiver agree on the meaning of a KID element. This agreed upon meaning is formalized and stored in an ontology. The intended meaning of a KID element sent from a sender is expected to be equivalent to the interpreted meaning of the KID element of the receiver.
– The semantic heterogeneity / semantic interoperability problem is an inter-system specification (build-time) and communication (run-time) problem and is addressed at the specification - interpretation point (e.g. between creating a definition and the interpretation of that definition).
– Sentences are built from the vocabulary. A set of sentences (facts) pertaining to a domain (of discourse) is a KB of that domain. This is the basis for ontology-based knowledge management and engineering.
– These facts are not fixed due to enterprise learning – open world semantics– An enterprise systems that is architected to include an ontology-based knowledge
management system with open world semantics provides the foundation for a disciplined approach to innovation
Semantic Interoperability and Ontology SystemsSemantic Interoperability and Ontology SystemsIn
tegr
atio
n, I
nte
rope
rabi
lity,
and
Col
labo
ratio
n:S
eman
tic I
nter
ope
rabi
lity
and
Col
labo
ratio
n S
yste
ms
Inte
grat
ion,
In
tero
pera
bilit
y, a
nd C
olla
bora
tion:
Sem
antic
Int
ero
pera
bilit
y an
d C
olla
bora
tion
Sys
tem
s
August 15, 2006 Copyright (C) 2006 Mathet Consulting, Inc. 19
Semantic Interoperability and Ontology SystemsSemantic Interoperability and Ontology Systems
<X> SystemOperational
Administration and Monitoring
Tools
Stakeholder SystemsService Requester/
Service Provider
<X> System Services
<X> System APIs and Data Interchange
Product / ServiceDefinition / Specification
(Architects and Engineers)
Other <X> System Engines
(e.g. in a Federation)
<X> Systems<X> =
Ontology and Semiotic
<X> Interoperability-“the ability of two or more <X> engines to communicate and interoperate in order to coordinate and execute <X> service instances across those engines
May be remote
May be remote
<X> System Engine(“ a software service, or ‘engine’, that provides the run time execution environment for a <X>
System service instance”)
E.g. - real-time - monitoring
- analysis, prediction, optimization- self defining
- dynamic configuration- self managing, self-diagnosing, self-adapting
etc.
Modified from WfMC / OMB standardsModified from WfMC and OMG Standards
Interoperability
Ontologies
D
Inte
grat
ion,
In
tero
pera
bilit
y, a
nd C
olla
bora
tion:
Sem
antic
Int
ero
pera
bilit
y an
d C
olla
bora
tion
Sys
tem
s
Inte
grat
ion,
In
tero
pera
bilit
y, a
nd C
olla
bora
tion:
Sem
antic
Int
ero
pera
bilit
y an
d C
olla
bora
tion
Sys
tem
s
Services orientation viaEnterprise architecturegovernment of SOA
A
BC
August 15, 2006 Copyright (C) 2006 Mathet Consulting, Inc. 20
Semantic Interoperability and Ontology SystemsSemantic Interoperability and Ontology Systems
Sym
bols
-
Incl
udin
g te
rms
-
Log
ic L
angu
ages
O
WL,
Com
mon
Log
ic,
Def
easa
ble
Log
ic
Dom
ain
V
ocab
ula
ry-
Co
ntro
lled
Voc
abu
larie
s -
- D
ictio
narie
s -
- T
hes
aur
i -D
efin
ition
s
Ontolo
giesLog
ical theories accounting
for the intended m
eaning of a fo
rmal (shared) vo
cabulary
Context
Context / Meaning BrokerMediator / Negotiator / Arbitrator
Rel
atio
ns
Meaning Triangulation™
Meta-ontologie
sLog
ical theories that unify (m
erge / m
ap)
other logical theories that account for the intende
dm
eaning of associate
d formal vocabularies
Ena
bles merging / m
apping be
tween ontolog
ies
Map
ping
s
Mea
sure
s
Expression Engine
(Met
a) L
angu
age
s (s
ynta
x)
Things referred to
Agents
Ontology and Semiotic System Engine Components
Reasoning Engine(including interpretation functions)Induction / Deduction / Abduction
COTS products
Inte
grat
ion,
In
tero
pera
bilit
y, a
nd C
olla
bora
tion:
Sem
antic
Int
ero
pera
bilit
y an
d C
olla
bora
tion
Sys
tem
s
Inte
grat
ion,
In
tero
pera
bilit
y, a
nd C
olla
bora
tion:
Sem
antic
Int
ero
pera
bilit
y an
d C
olla
bora
tion
Sys
tem
s
OWL, ODM Standards
August 15, 2006 Copyright (C) 2006 Mathet Consulting, Inc. 21
Semantic Interoperability and Ontology SystemsSemantic Interoperability and Ontology Systems
Has role / set of responsibilities
Context
ContextContext
Interaction Interaction
Di Di
DiDi
DO DO
DODO
System AIndustrialSuppliers
System BWar fighter
System
System CE.g. Defense Logistics
Agency
General Flow
C1
C2
Ontology (and Semiotic) Engine
AB B
C1-C2 Shared Ontology (and Semiotic) Engine
Ontology (and Semiotic) Engine
DD
E.g. SupplierSystems
E.g. CustomerSystems
Semantic (semiotic) systems layer supporting communication(s)
Meta DataRepository
COTS productsw/ standards
Inte
grat
ion,
In
tero
pera
bilit
y, a
nd C
olla
bora
tion:
Sem
antic
Int
ero
pera
bilit
y an
d C
olla
bora
tion
Sys
tem
s
Inte
grat
ion,
In
tero
pera
bilit
y, a
nd C
olla
bora
tion:
Sem
antic
Int
ero
pera
bilit
y an
d C
olla
bora
tion
Sys
tem
s
May be a shared Ontology
August 15, 2006 Copyright (C) 2006 Mathet Consulting, Inc. 22
Semantic Interoperability and Ontology SystemsSemantic Interoperability and Ontology SystemsIn
tegr
atio
n, I
nte
rope
rabi
lity,
and
Col
labo
ratio
n:S
eman
tic I
nter
ope
rabi
lity
and
Col
labo
ratio
n S
yste
ms
Inte
grat
ion,
In
tero
pera
bilit
y, a
nd C
olla
bora
tion:
Sem
antic
Int
ero
pera
bilit
y an
d C
olla
bora
tion
Sys
tem
s
Data {Language: syntax, vocabulary}
Sender Reciever
SenderReciever
System MSystem M’
Role N’Role N
“Instance” of an emergent Contract Negotiation SystemProduces a service to all parties involved in negotiations
The parties are part of a group
Conduit / Channel / MediaReceive
Data
ProcessData
Make Decisions
Take Actions
ReceiveData
ProcessData
Make Decisions
Take Actions
Contract NegotiationRequires Deontic / Modal Logic
Requires Semantic interoperability
Output is a contract documentContract has a role- contract is passive, role is assigned (role may be different among the parties)The contract has value-
Negotiation StylesCompetitive, Cooperation, Collaboration,Compromising, Accommodating, Avoiding
Each negotiation style results in some level of utility on both the individual basis and at the systematic level (ecosystem levels)
There is an efficiency associated with each style to reach a given level of utility
There are 2..N members / parties in a negotiation group
Agents may be aware of enteprise history, goals, strategies, current state including existing contractsà contract negotiated must be consistent with all of above
Rights, obligations, permissions,
consequences, duties, violations, powers
(authority)
Contract
August 15, 2006 Copyright (C) 2006 Mathet Consulting, Inc. 23
Semantic Interoperability and Ontology SystemsSemantic Interoperability and Ontology SystemsIn
tegr
atio
n, I
nte
rope
rabi
lity,
and
Col
labo
ratio
n:S
eman
tic I
nter
ope
rabi
lity
and
Col
labo
ratio
n S
yste
ms
Inte
grat
ion,
In
tero
pera
bilit
y, a
nd C
olla
bora
tion:
Sem
antic
Int
ero
pera
bilit
y an
d C
olla
bora
tion
Sys
tem
s
Data {Language: syntax, vocabulary}
Sender Reciever
SenderReciever
System MSystem M’
Role N’Role N
“Instance” of an emergent / “programmed” Contract Fulfillment SystemProduces a service to all parties involved in the interaction
The parties are part of a group
The (business) services have value
Services are managedServices are governed. If a contract is in place, including rights, obligations, permissions, consequences, duties, violations, and powers (authority), then
the contract governs the interaction
Conduit / Channel / Media
ReceiveData
ProcessData
Make Decisions
Take Actions
ReceiveData
ProcessData
Make Decisions
Take Actions
Contract FulfillmentThe contract governs the interaction.Especially true with legal contracts-
Enterprise and Inter-Enterprise Architecture and Engineering
3 Examples Using Ontology Systems for Semantic Interoperability in Government
Enterprise and Inter-Enterprise Architecture and Engineering
3 Examples Using Ontology Systems for Semantic Interoperability in Government
Integration, Interoperability and Collaboration:Semantic Interoperability and Ontology Systems
Integration, Interoperability and Collaboration:Semantic Interoperability and Ontology Systems
August 15, 2006 Copyright (C) 2006 Mathet Consulting, Inc. 25
Enterprise and Inter-Enterprise SystemsEnterprise and Inter-Enterprise Systems
Data “Certification”- a service -
(based on meta-data, federated schema, rules,
ontology)
Internet data via Process /
Workflow and Web services
Relational / Object Databases
May be federated
Workflow Datastore
“Federated” Data System
Data Ware-house
Single Source Legal Record
Real-time / ETL
Hybrid
Read-Only
Quality Check
Feedback
Feedback
Business Transaction
Data
Trans-modal input
Data flow is on a Trans-modal
“Enterprise Service Bus”Partial View
Records Datastore
Document Datastore
Copyright © 2005 Mathet Consulting, Inc.
Content Datastore
Ontology and Information Quality in Enterprise / Inter-
Enterprise Architecture Partial View
Inte
grat
ion,
In
tero
pera
bilit
y, a
nd C
olla
bora
tion:
Sem
antic
Int
ero
pera
bilit
y an
d C
olla
bora
tion
Sys
tem
s
Inte
grat
ion,
In
tero
pera
bilit
y, a
nd C
olla
bora
tion:
Sem
antic
Int
ero
pera
bilit
y an
d C
olla
bora
tion
Sys
tem
s
August 15, 2006 Copyright (C) 2006 Mathet Consulting, Inc. 26
Agency Distributed
LOB System
Agency Process / WF
System
Agency BRSystem
Agency DW System
Architecture /
Engineering
State Legislature
Law
AgencyPolicy, Planning and Legislative
Analysis
Federated and Local Ontology
AgencyPolicies
ProceduresDirections
Service
Agency KM System
Business Process
X
Business Process
Y
Business Process
X
IRS Regulations
SocialSecurity
StateAdminRules
AttorneyGeneral
Court Decisions
Agency Board
StateNew /
RevisedStatutes
MetaDataData Quality System
Summary TablesHybrid “ETL” / Real-
time SubsystemsAnnotated with
Ontology
Embedded With MDA
=
Copyright © 2005 Mathet Consulting, Inc.
Ontology and Information Quality in
Enterprise / Inter-Enterprise Architecture
Partial View
BanksOther
Agencies
Money / Data
Economics OntologyLaw Ontology
Process OntologyRule Ontology
Etc.
A
A
B
Enterprise and Inter-Enterprise SystemsEnterprise and Inter-Enterprise SystemsIn
tegr
atio
n, I
nte
rope
rabi
lity,
and
Col
labo
ratio
n:S
eman
tic I
nter
ope
rabi
lity
and
Col
labo
ratio
n S
yste
ms
Inte
grat
ion,
In
tero
pera
bilit
y, a
nd C
olla
bora
tion:
Sem
antic
Int
ero
pera
bilit
y an
d C
olla
bora
tion
Sys
tem
s
All services (including ontology services)designed on top of unified and converged communications
Interoperable ontology engines
August 15, 2006 Copyright (C) 2006 Mathet Consulting, Inc. 27
Core and Foundation Boundaries captured in service level
agreements and IIC
Disaster Recovry
Service Event Analytics Repository
(System(s) state as represented
by state of managed services / components)
Availability
Capacity and Performance Service Levels
(Agreements)
Continuity
Disaster Recovry
Architecture /Engineering
Quality
Service Event Repository(System(s) state as
represented by state of managed
Services / components)
Availability
Capacity and Performance
Service Levels (Agreements)
Continuity
Service Event Server / Services(Real-time Monitoring and
management of system(s) state)
Availability
Capacity and Performance
Service Levels Continuity
Disaster Recovry
Managed Component Sensors for: Availability Service LevelsCapacity / Performance Service LevelsContinuity Service LevelsDisaster Recovery Service Levels
Sensors are “embedded” in systems (components / services)
Each system may be considered to be a
component offering a service
Sets of collaborating systems
(components / services) may
function a a single service
System 2
System 1
System N
System 4
System 3
Service Event Analytics Server / Services
(Monitoring and management of system(s) state)
Descriptive Statistics:
Means, Variance,
Histograms,(Cross-)
Correlations,etc.
ANOVA / MANOVA
Econometric Time Series
SimulationPredictionOptimization
Reporting
Copyright © 2005 Mathet Consulting, Inc.
Semiotics (Semantics)
Semiotics (Semantics)
Semiotics (Semantics)
Semiotics (Semantics)
Enterprise and Inter-Enterprise Architecture and EngineeringEnterprise and Inter-Enterprise Architecture and EngineeringIn
tegr
atio
n, I
nte
rope
rabi
lity,
and
Col
labo
ratio
n:S
eman
tic I
nter
ope
rabi
lity
and
Col
labo
ratio
n S
yste
ms
Inte
grat
ion,
In
tero
pera
bilit
y, a
nd C
olla
bora
tion:
Sem
antic
Int
ero
pera
bilit
y an
d C
olla
bora
tion
Sys
tem
s
Summary and ConclusionsSummary and Conclusions
Enterprise and Inter-Enterprise Information Quality through Semantic Interoperability and Ontology Systems
Enterprise and Inter-Enterprise Information Quality through Semantic Interoperability and Ontology Systems
August 15, 2006 Copyright (C) 2006 Mathet Consulting, Inc. 29
• Semantic interoperability – is pervasive and ubiquitous– between machines requires formal languages– requires objective measures based on measurement theory (a
model of a thing is required in addition to a definition)– is concerned with the differences in semantics between
interacting systems and the effects these differences have on interaction / collaboration
– is the driving force between a host of semantic measures that are useful for system-system interaction / collaboration (see glossary for partial list)
Summary and ConclusionsSummary and ConclusionsIn
tegr
atio
n, I
nte
rope
rabi
lity,
and
Col
labo
ratio
n:S
eman
tic I
nter
ope
rabi
lity
and
Col
labo
ratio
n S
yste
ms
Inte
grat
ion,
In
tero
pera
bilit
y, a
nd C
olla
bora
tion:
Sem
antic
Int
ero
pera
bilit
y an
d C
olla
bora
tion
Sys
tem
s
August 15, 2006 Copyright (C) 2006 Mathet Consulting, Inc. 30
• Semantic interoperability, KID Quality and KID Quality Systems
– Formal models of quality are required
– Quality KID / measures need to be cast in terms of semantic interoperability just like any other KID / measures that take part in system-system interaction / collaboration
– Semantic interoperability issues exist between KID quality systems as well as between any other sets of interacting systems
– Formally defining what things mean enables formal data definitions.
Summary and ConclusionsSummary and ConclusionsIn
tegr
atio
n, I
nte
rope
rabi
lity,
and
Col
labo
ratio
n:S
eman
tic I
nter
ope
rabi
lity
and
Col
labo
ratio
n S
yste
ms
Inte
grat
ion,
In
tero
pera
bilit
y, a
nd C
olla
bora
tion:
Sem
antic
Int
ero
pera
bilit
y an
d C
olla
bora
tion
Sys
tem
s
August 15, 2006 Copyright (C) 2006 Mathet Consulting, Inc. 31
• Semantic Interoperability, KID Quality and KID Quality Systems, and Ontology (and Semiotics) Systems
– As logical theories accounting for the intended meaning of a formal vocabulary, ontologies enable formal quality measures, including semantic measures to be defined and applied (meanings are formalized)
– Ontology systems provide tools to engineer new ontologies– Ontology systems provide a semantic reference point for semantic
interoperability between interacting / collaborating systems – conformance to an ontology.
– Ontology Systems stores background KID associated with the production of new KID
– A Quality Ontology may be constructed from quality standards and best practices, including CMMs, Continuous Process Improvement, ISO, QA, QC, Independent Validation and Verification, TQ(d)M, Six-Sigma
– An example of an Ontology language is OWL• Ontologies provide a vocabulary and definitions of rules for use by independently
developed resources, processes, services, systems, etc.• Ontologies enable agreements, e.g..... legal contracts, service-level agreements, to
be made among organizations sharing common services with regard to usage and meaning of relevant concepts that can be expressed unambiguously
• Independently developed systems, agents and services can work together to share information and processes consistently, accurately and completely by composing component ontologies, mapping ontologies to one another and / or mediating terminology among participating resources and services, Ontologies also facilitate conversations among agents to collect, process, fuse, and exchange information.
• Ontology (and Semiotic) Systems provide a common facility of semantic lifecycle mechanisms.
Summary and ConclusionsSummary and ConclusionsIn
tegr
atio
n, I
nte
rope
rabi
lity,
and
Col
labo
ratio
n:S
eman
tic I
nter
ope
rabi
lity
and
Col
labo
ratio
n S
yste
ms
Inte
grat
ion,
In
tero
pera
bilit
y, a
nd C
olla
bora
tion:
Sem
antic
Int
ero
pera
bilit
y an
d C
olla
bora
tion
Sys
tem
s
August 15, 2006 Copyright (C) 2006 Mathet Consulting, Inc. 32
• Enterprise and Inter-enterprise Systems Interaction and Collaboration– Ontology systems enable semantic interoperability
between enterprises where interacting enterprises have minimal shared context
– Interoperable ontology systems provide a semantic systems layer for semantic interoperability between interacting / collaborating systems
– Architectural Styles for addressing semantic interoperability include: brokered / mediated, common / merged, distinct / mapped
– Introduction of ontology / semiotic systems into enterprise systems elevates the enterprise to a higher systems capability level – symbol processing.
Summary and ConclusionsSummary and ConclusionsIn
tegr
atio
n, I
nte
rope
rabi
lity,
and
Col
labo
ratio
n:S
eman
tic I
nter
ope
rabi
lity
and
Col
labo
ratio
n S
yste
ms
Inte
grat
ion,
In
tero
pera
bilit
y, a
nd C
olla
bora
tion:
Sem
antic
Int
ero
pera
bilit
y an
d C
olla
bora
tion
Sys
tem
s
August 15, 2006 Copyright (C) 2006 Mathet Consulting, Inc. 33
• Semantic Alignment• Semantic Coupling• Semantic Cohesion• Semantic Conflict
Semantic Defect• Semantic Dependability• Semantic Difference• Semantic Discrepancy• Semantic Distance• Semantic Divergence• Semantic Drift • Semantic Error • Semantic Exception• Semantic Fault• Semantic Failure• Semantic Heterogeneity• Semantic Incident • Semantic Incident Prevention• Semantic Incident Detection• Semantic Incident Response• Semantic Incident Review
En
terp
rise
an
d I
nte
r-E
nte
rpri
se I
nfo
rmat
ion
Qu
alit
y th
rou
gh
Sem
anti
c In
tero
per
abil
ity
and
On
tolo
gy
Sys
tem
s
En
terp
rise
an
d I
nte
r-E
nte
rpri
se I
nfo
rmat
ion
Qu
alit
y th
rou
gh
Sem
anti
c In
tero
per
abil
ity
and
On
tolo
gy
Sys
tem
sAppendixAppendix
August 15, 2006 Copyright (C) 2006 Mathet Consulting, Inc. 34
• Semantic integration• Semantic Interoperability• Semantic Interoperability necessary and sufficient conditions• Semantic Issue• Semantic Match / Mismatch• Semantic Quality of Service• Semantic Problem• Semantic Proximity• Semantic Reliability• Semantic Risk• Semantic Service Levels • Semantic Stability• Semantic Stationarity• Semantic Similarity• Semantic Shift
Semantic Reliability• Semantic Robustness• Semantic Tolerance• Semantic Variance / Variability• Semantics• Semiotics• System Interoperability• System Failure • System Fault
En
terp
rise
an
d I
nte
r-E
nte
rpri
se I
nfo
rmat
ion
Qu
alit
y th
rou
gh
Sem
anti
c In
tero
per
abil
ity
and
On
tolo
gy
Sys
tem
s
En
terp
rise
an
d I
nte
r-E
nte
rpri
se I
nfo
rmat
ion
Qu
alit
y th
rou
gh
Sem
anti
c In
tero
per
abil
ity
and
On
tolo
gy
Sys
tem
sAppendixAppendix
August 15, 2006 Copyright (C) 2006 Mathet Consulting, Inc. 35
• Systems, S, of Interest
– A set of atomic symbols (e.g..... terms), which may be composed. For example, .A set of definitions (e.g..... connotations, denotations)
– A set of conceptual things – concepts / conceptual (sub) systems– A set of real-world things (subsystems), a subset of which are active– A set of relations between / among symbols, definitions, concepts, real-world things– A set of mappings between / among symbols, definitions, concepts, real-world things– A set of languages used to express symbols, definitions, concepts, real-world things,
relations and mappings– A set of allowable “configurations” / states between / among symbols, definitions,
concepts and real-world things constrained by the mappings / relations expressed with the languages
– A set of disallowed “configurations” / states between / among symbols, definitions, concepts and real-world things constrained by the mappings / relations expressed with the languages
– A set of relations between real-world / concept things (subsystems) and allowable system states, and disallowed system states
– A set of mapping functions between real-world / concept things (subsystems) and allowable, and disallowed system states
– A set of (sub) system boundaries, any one of which may be opened/closed, to some degree (e.g..... some set of inclusion / exclusion functions that operate on the above sets)
– (Sub) systems may be nested to some N levels– The whole is greater than the sum of the parts – new structure / behavior emerges from
interaction between the parts– Time – any of the above may change with time (i.e. is dynamic ; e.g..... learning,
adaptability, etc.)
Semantic InteroperabilitySemantic InteroperabilityIn
tegr
atio
n, I
nte
rope
rabi
lity,
and
Col
labo
ratio
n:S
eman
tic I
nter
ope
rabi
lity
and
Col
labo
ratio
n S
yste
ms
Inte
grat
ion,
In
tero
pera
bilit
y, a
nd C
olla
bora
tion:
Sem
antic
Int
ero
pera
bilit
y an
d C
olla
bora
tion
Sys
tem
s
Contact InformationContact Information
Mathet Consulting, Inc.Integrated, Interoperable and Collaborative Systems
MC
‘PMB 140041450 E. American LaneSchaumburg, IL 60173Office: 847-330-6375Cell: [email protected]