Transcript Chapter 6: Modeling and Representation

Chapter 6:
Modeling and Representation
Service-Oriented Computing: Semantics, Processes, Agents
– Munindar P. Singh and Michael N. Huhns, Wiley, 2005
Highlights of this Chapter







Chapter 6
Integration versus Interoperation
Common Ontologies
Knowledge Representations
Relationships
Hierarchies
Modeling Fundamentals
Unified Modeling Language (UML)
Service-Oriented Computing: Semantics, Processes, Agents - Munindar Singh and Michael Huhns
2
Integration versus Interoperation
Integration
EDI
XML
Application
Application
Application
Application
Application
Transform
Transform
Transform
Transform
Transform
Transformation
Transformation
Transform
Transform
Transform
Transform
Transform
Application
Application
Application
Application
Application
API
solution
Standard Data
Format
Tight coupling
Chapter 6
Portal + Workflow Solution
XML-Based Data Distributed application: XML +
Exchange Format Web services + workflow
Loose coupling
Service-Oriented Computing: Semantics, Processes, Agents - Munindar Singh and Michael Huhns
3
Modeling and Composing Services
Universe of
Discourse
Chapter 6
Requirements
Analysis
Conceptual
Schema
Design
Service
Implementation
INTEROPERATE
MAP
Background
Knowledge
Conceptual
Schema
COMPREHEND
Universe of
Discourse
Requirements
Analysis
Service
Interface
Service
Interface
Design
Service
Implementation
Service-Oriented Computing: Semantics, Processes, Agents - Munindar Singh and Michael Huhns
4
Dimensions of Abstraction: 1
Abstractions of information resources capture
knowledge that is relevant for interoperation;
think of these as constraints that must be
discovered and represented
 Data



Chapter 6
Domain specifications
Value ranges, e.g., Price >= 0
Allow/disallow null values
Service-Oriented Computing: Semantics, Processes, Agents - Munindar Singh and Michael Huhns
5
Dimensions of Abstraction: 2

Structure






Chapter 6
Taxonomic representations and relationships such
as in schemas and views, e.g., securities are
stocks
Specializations and generalizations of domain
concepts, e.g., stocks are a kind of liquid asset
Value maps, e.g., S&P A+ rating corresponds to
Moody’s A rating
Semantic data properties, sufficient to characterize
the value maps, e.g., some stock price databases
consider daily averages; others closing prices
Cardinality constraints
Integrity constraints, e.g., each stock must have a
unique SEC identifier
Service-Oriented Computing: Semantics, Processes, Agents - Munindar Singh and Michael Huhns
6
Dimensions of Abstraction: 3

Process
 Procedures, i.e., how to process information, e.g.,
how to decide what stock to recommend
 Preferences for accesses and updates in case of
data replication (based on recency or accuracy of
data)
 Preferences to capture view update semantics
 Contingency strategies, e.g., whether to ignore,
redo, or compensate
 Contingency procedures, i.e., how to compensate
transactions
 Flow, e.g., where to forward requests or results
 Temporal constraints, e.g., report tax data every
quarter
Chapter 6
Service-Oriented Computing: Semantics, Processes, Agents - Munindar Singh and Michael Huhns
7
Dimensions of Abstraction: 4

Policy



Chapter 6
Security, i.e., who has rights to access or update
what information? (e.g., customers can access all
of their accounts, except blind trusts)
Authentication, i.e., a sufficient test to establish
identity (e.g., passwords, retinal scans, or smart
cards)
Bookkeeping (e.g., logging all accesses)
Service-Oriented Computing: Semantics, Processes, Agents - Munindar Singh and Michael Huhns
8
Value Maps: 1
A value map relates the values expressed
by different services
Key properties
 Totality
 Order preservation
 Consistent inversion
Chapter 6
Service-Oriented Computing: Semantics, Processes, Agents - Munindar Singh and Michael Huhns
9
A.M. Best
Moody’s
A.M. Best
Aaa
Aaa
Aa1
A++
Value
Maps:
2
A+
A
Aa2
Aa3
A1
A-
A2
B+
B
C++
C+
C
Baa1
Baa2
Baa3
B-
Ba1
Ba2
Ba3
B1
B2
A+
A
A-
Aa2
Aa3
A1
A2
A3
B++
B+
B
BC++
C+
C
Baa1
Baa2
Baa3
Ba1
Ba2
Ba3
B1
B2
C-
B3
C-
B3
D
Caa
D
Caa
E
Ca
E
Ca
F
C
F
C
(a) A consistent value map
Chapter 6
Aa1
A++
A3
B++
Moody’s
(b) A value map that violates
consistent inversion
Service-Oriented Computing: Semantics, Processes, Agents - Munindar Singh and Michael Huhns
10
Ontology

A specification of a conceptualization or a set of
knowledge terms for a particular domain, including





Chapter 6
The vocabulary: concepts and relationships
The semantic interconnections: relationships among
concepts and relationships
Some simple rules of inference and logic
Some representation languages for ontologies:
 Uniform Modeling Language (UML)
 Resource Description Framework Language
Schema (RDFS)
 Web Ontology Language (OWL)
Some ontology editors: Protégé, Webonto, OilEd
Service-Oriented Computing: Semantics, Processes, Agents - Munindar Singh and Michael Huhns
11
Common Ontologies

A shared representation is essential to
successful communication and interoperation



For humans: physical, biological, and social world
For computational agents: common ontology
(terms used in communication)
Representative efforts are



Cyc (and Opencyc)
WordNet (Princeton); LDOCE; OED
Several upper-level ontologies, including by IEEE

Chapter 6
Mostly stable concepts such as space, time, person,
which can be used within various domains
Service-Oriented Computing: Semantics, Processes, Agents - Munindar Singh and Michael Huhns
12
Ontologies and Articulation Axioms
Travel Agent Service
User’s Agent
nonNegativeInteger
range
seats
Airliner
Seating
Arrangement
equipment
Airplane
Flight
numpassengers
to
Mapping by
hand, but
with tool
support
from
nonNegativeInteger
JumboJet
Common
Ontology
Airport
Transportation
Device
Itinerary
uses
Commercial
Transportation
Device
*
1
to
Location
Chapter 6
Leg
from
class
Class of
Service
Boeing
777
Developing a
common ontology:
• All at once
• Incrementally via
consensus
Public
Transportation
Device
Service-Oriented Computing: Semantics, Processes, Agents - Munindar Singh and Michael Huhns
13
Knowledge Representation


Expressive power
Procedural (how) versus declarative (what)



Chapter 6
Declarative pros: enables standardization,
optimization, improved productivity
Declarative cons: nontrivial to achieve and causes
short-term loss of performance
Trade-offs shifted by Web to favor declarative
modeling
Service-Oriented Computing: Semantics, Processes, Agents - Munindar Singh and Michael Huhns
14
Frames versus Descriptions
Frame-based approaches are intuitive
but rely on names of classes and
properties to indicate meaning
 Description logics provide a
computationally rigorous means to
represent meaning; difficult for people
Managing this trade-off is a major
challenge for Knowledge Representation

Chapter 6
Service-Oriented Computing: Semantics, Processes, Agents - Munindar Singh and Michael Huhns
15
Exercise: Which Conceptualization is Most
Expressive and Flexible?





awg22SolidBlueWire(ID5)
blueWire(ID5, AWG22, Solid)
solidWire(ID5, AWG22, Blue)
wire(ID5, AWG22, Solid, Blue)
wire(ID5)^size(ID5, AWG22)^type(ID5,
solid)^color(ID5, Blue)
Chapter 6
Service-Oriented Computing: Semantics, Processes, Agents - Munindar Singh and Michael Huhns
16
Mappings among Ontologies

Term-to-term (one-to-one), e.g.,
hookupWireO1 ´ wireO2

Many-to-one, e.g.,
solidWireO1(x, size, color) Æ strandedWireO1(x, size, color)
´ wireO2(x, size, color, (Stranded|Solid))

Many-to-many, e.g.,
solidBlueWireO1(x, size) Æ
solidRedWireO1(x, size) Æ
strandedBlueWireO1(x, size) Æ
strandedRedWireO1(x, size)
´
solidWireO2(x, size, (Red|Blue)) Æ
strandedWireO2(x, size, (Red|Blue))
Chapter 6
Service-Oriented Computing: Semantics, Processes, Agents - Munindar Singh and Michael Huhns
17
Unified Modeling Language (UML) for Ontologies
PurchaseOrder
Class diagrams
without members
1
*
1
Chapter 6
*
*
1
ItemDetail
1
ShippingDetail
BillingDetail
*
*
Seller
Buyer
Organization
Person
1
1
*
Account
Service-Oriented Computing: Semantics, Processes, Agents - Munindar Singh and Michael Huhns
18
Comparison of Modeling Languages
Chapter 6
Service-Oriented Computing: Semantics, Processes, Agents - Munindar Singh and Michael Huhns
19
Chapter 6 Summary

Shared models are essential for
interoperation





Chapter 6
Based on shared ontologies or conceptualizations
Good models must accommodate several
important considerations
Modeling requires several subtle considerations
Declarative representations facilitate
reasoning about and managing models
Formalization enables ensuring correctness of
models and using them for interoperation
Service-Oriented Computing: Semantics, Processes, Agents - Munindar Singh and Michael Huhns
20