inf5120 – model-based system development · telecom and informatics 3 inf5120 - lecture plan -...

Telecom and Informatics

INF5120 – Model-based SystemDevelopment

Lecture 12: System Architecture andInformation/Ontology Modeling

April 15th, 2013Arne J. Berre, SINTEF ICT


Content

n Lecture plan 2013n Oblig 2 and 3n System Architecturen Cloud computingn Java 6, JEEn Conceptual Modelingn Ontologiesn Information Modelingn UML Class Modelsn ISO 19103n UML Profiles

2

Telecom and Informatics 3

INF5120 - Lecture plan - 2013

n 1 (14/1): Introduction – overview Enterprise Architecture with UML and BPMN and DSLsn 2 (21/1): Service Innovation and Design, AT ONE method/workshop – myServiceFellow

(Marika Lüders)n 3: (28/1): Value Networks/VDML BPMN, vs. UML Activity diagrams - Oryxn 4 (4/2): User experience and Touchpoints/UI Design – Balsamiq – (Amela Karahasanovic)n 5 (11/2): UML and Req.Modeling –Agile User stories versus Use casesn 6 (18/2): Business-SoaML, Requirements Modeling, Goal Modeling, BMM, and Non

Functional requirementsn 7 (25/2): Model driven engineering – Metamodels, DSL, UML Profiles etc. NAV on NFR

n 8 (4/3): UML Service Modeling – SoaML, UML 2.0 Service composition, MagicDrawn 9 (11/3): Method Engineering, SW Process frameworks , SPEM/EPF, ISO 24744,

FACESEM/ESSENCE (Brian Elvesæter)n 10(18/3): Model driven engineering, transformation technologies,n 11(8/4): MDE and DSL in practice, with ThingML example – Franck Fleureyn 12(15/4): System Architecture and Information/Ontology modeling, UML, ISO 19103n 13(22/4): System Architecture and Service models, semantic modelsn 14(29/4): BPMN and Business Process Management and UI modelsn 15(6/5): Platform models for the Cloud, with CloudML, Alessandro Rossinin 16(13/5): Conclusion and Summary for INF5120 - Preparation for Exam

n Exam: Monday June 3rd, 2013, (4 hours)


INF5120 – Oblig plan - 2013n 1 (14/1): Introductionn 2 (21/1): myServiceFellow/Service Designn 3: (28/1): Oryxn 4 (4/2): Balsamiqn 5 (11/2): Use cases 2.0n 6 (18/2): Oblig 1 – Group workn 7 (25/2): EMF and Eclipse – Group (3) presentation – Business Model, Business-SoaMLn 8 (4/3): Oryx– Group (3) presentation - User stories/use cases, UI, NFRn 9 (11/3): Group presentation(s) on Oblig 1: User stories/use cases, UI/Balsamiqn 10(18/3): Group presentation on Oblig 1 – NFR 1 – Parallell Start on Oblig 2 (individual)

and Oblig 3 (group)

n - 22/3 : Delivery of Oblig 1n EASTER

n 11(8/4): Walk through of Oblig 1 – Questions on Oblig 2 and 3n 12(15/4): Question on Oblig 2 – “Value Network editor in Eclipse”, description of Oblig 3 –

including references to MagicDraw – Cameo Enterprise Architecturen 13(22/4): Group work, Oblig 2/3 – Group presentation/discussionn 14(29/4): Group work, Oblig 2/3 – Group presentation /discussion – delivery Oblig 2n 15(6/5): Delivery of Oblig 3 – “Concierge – SA Model”n 11/5: Delivery of Oblig 3n 16(13/5): Walk trough of Oblig 3 - Preparation for Examn Exam: Monday June 3rd, 2013, (4 hours)


Oblig 3 – Concierge Model-basedSystem Architecturen System Architecture for Conciergen Basis for Cloud realisationn Use of Java, JEE 6.xn MagicDraw Cameo Enterprise Architecture 17.0.3

n Information Model – UML Class modeln Service Model – SoaML modeln Process/Behaviour Model – BPMN/Seq.diagrams, …n UI Model – (from Balsamiq)

5


3-tier System Reference architecture

DataLayer:

ApplicationLayer:

BusinessLayer:

UserService

UIC

Legacy

Service Infrastructure

Service

Entity

DataService

6


Work productsRequirements ModelBusiness Model

Service Model

Platform Realization Model

Business ProcessModel

Services Architecture Service Contract

Service Detailed designService design

BusinessInformation

Modell

SystemBoundary

Cloud, Web Services, JEE, MAS, P2P/Grid, SWS

uc MODUS Studio

MODUS L ang uag e En ginee ri ng Wo rkben ch

La nguageEngine ering

Langua geEnginee rChec k

LanguageQuality

IDE Dev elope rDev e lop IDE

Rich Picture

Use casescenario. Modeling artefacts at different

levels of abstracionand from different viewpoints

Inception

Elaboration

& Construction phase focus

NFR andother

requirements

Construction

& Transition phase focus

& Elaboration phase focus

7

Business Architecture

System Architecture


Key conceptsn Process Phases:

n Four phases are defined (related to UP phases). The key concept is iterations ineach phase.

n Disciplines:n Five core disciplines are defined where each has a specific goal and includes

some practices to achieve the goal.n Business modelling: understanding the business and providing information for

developing a system that satisfies business goals.n Requirement modelling: eliciting stakeholders’ requirements on what the

system should do and the constraints.n Service modelling: Developing a service-oriented architecture and related

components that fulfil the requirements.n Platform realization: realization on the specific platform.n Testing: validation and verification of the system.

n Each discipline includes a set of practices that are performed in one orseveral tasks. The output are work products (models and others).

8


Use case driven

n All process components, from requirements capture totest, are driven by use cases/scenarios

9

AnalyzeCapture, Clarifyand Validate the

Use Cases

Design andImplementation

Realize the UseCases

TestVerify that the Use Cases

are fulfilled

Use Cases bind these work steps together


Architecture centric (SOA elements)

10

(Dirk Krafzig, Karl Banke, and Dirk Slama.Enterprise SOA. Prentice Hall, 2005)

Code Modules Offered asservice with description for

• Discovery• Utilisation (composition)


Different architectural views

n Application architecturen This is the architecture of an

application that consumes servicesand integrates them into businessprocesses in order to provide someservice of value to the customers.

n Services architecturen This view shows services that are

available for use through theirinterfaces. This view hides theimplementation from theconsumers.

n Component architecturen This view shows the components

implementing the services.

11

ComponentA

ComponentB

ComponentC

ServiceA

ServiceB

ServiceC

User interface

Business logic

Data businessprocesses

services(interfaces)

components(implementation)

application

(mediators,wrapping)


SOA and Cloud Computingn One can consider cloud

computing the extension ofSOA out to cloud-deliveredresources, such as storage-as-a-service, data-as-a-service, platform-as-a-service-- you get the idea.

n Provide on demand servicesn Platform/infrastructure

services forrunning/outsourcingapplication services

12


Cloud vs SOA

13

NewAccountsCommission

Calculation

Sales

DataCleaning

Sales OrderUpdate

SOA

Some (more and more) include SOA andSaaS as part of the cloud computing

concept (the cloud include it all)

Finance/Operations

Cloud is the enabler• Platform a service• Storage as a service• Computation as a service• …SaaS


Infrastructure-as-a-ServiceComputation-as-a-Service

Security-as-a-S

ervice Storage-as-a-Service

Integration-as-a-Service Database-as-a-Service

Information-as-a-Service

Process-as-a-Service

Everything as a servicePlatform

-as-a-Service

Software/Application-as-a-Service

Management/Governance-as-a-Service

Testing-as-a-Service

14


Different cloud vendors differentperspectives

VirtualServers

VirtualServers

VirtualServers

VirtualServers

VirtualServers

PaaS for theInquiring

Developer

“Servers as aService”

PaaS as anApplicationFramework

Integration as a Service

Logic as a Service

UI as a Service

Infrastructure as a ServiceElastic Computing as a

service

Python App Server

~Familiar DeveloperModel

Rapid Scalability

b OfferingInnovativeTechnology

Supports Large-Scale SaaSDeep-Dyed

Multitenancy

Storage/DB as a Service


service



service


Google appsGmail, Docs…

15


Java Enterprise Edition (JEE)

16


References

n SiSaS Methodology Wikin http://sisas.modelbased.net/

n JEE 6.0n First cup of JEE 6

n http://docs.oracle.com/javaee/6/firstcup/doc/n Official JEE 6 tutorial

n http://docs.oracle.com/javaee/6/tutorial/doc/

17


JEE: Java Enterprise Edition

n Java for 3 tiersarchitectures

n Includesn Java Libraries (Mail,

Protocols, etc.)n Tools, compilers, code

generators,n Application Server

n JEE 6.x

18


Java Server Faces

n Presentation (XHTML)<html>

<body><f:view>

<h:form formName="loginForm"><h:panelGrid columns="3">

<f:facet name="header"><h:outputText value="Login"/>

</f:facet><h:outputLabel value="Login"/><h:inputText value="#{login.login}"/><h:outputLabel value="Password:"/><h:inputSecret value="#{login.password}"/><h:commandButton type="reset" value="Reset"/><h:commandButton type="submit"

action="#{login.doLogin}"/></h:panelGrid>

</h:form></f:view>

</body></html>

n Controller (Java)

@ManagedBean(name=login)@SessionScopedpublic class LoginBean {

@EJBprivate LoginService delegate;

private String login;private String password;

public String getLogin() { … }public void setLogin(String login){ … }

public String getPassword() { … }public void setPassword(String password) { … }

public boolean doLogin(String login, String password) {delegate.login(login, password);

}

}

19

Login:

Password:

Authentication:


Web Services (With JEE 3.x)

n Expose methods on anetworkn XML/SOAPn HTTP

nWSDL descriptorsn Generated by A.S.n From annotations

20

import javax.jws.WebService;import javax.jws.WebMethod;

@WebServicepublic class Hello {

private String message = new String("Hello, ");

public void Hello() {

}

@WebMethodpublic String sayHello(String name) {

return message + name + "."; }

}


Persistence (With JEE 6.x)

n Store persistent datan How to transform

objects into tables?

nMapping generatedfrom annotationn What attribute are ID,n Associations

realizations

21

Customer

- id: String- firstname: String- lastname: String- email : String

Account

- id: String- name: String- /balance: double

Transaction

- amount: double- date: Date {readOnly}- description: String

User

- login: String- password: String

Agent

Request

- firstname: String- lastname: String

+accounts 1..*

+owner 1

+withdraw 0..*

+source1

+deposi t 0..*

+target1

+pending 0..*

+assignee 1

create table Customers ( id integer, firstname varchar(100), lastname varchar(100), email varchar(100) unique,

constraint primary key (id))

create table Accounts( id integer, owner integer,

constraint primary key (id),constraint foreign key owner references Customer(id)

)

create table Transactions ( id integer, description varchar(50), amount double,

constraint primary key (id),constraint foreign key source references Account(id),constraint foreign key target references Account(id)

)


Model based system development

nWhat does MBSDbrings?n Capitalize on key

developmentprocesses

n Automationn Model transformations

22

IBPModels

ModelTransformations

MD

E To

ols

Engine

Java Code

Clouddeployed


Internet Bank Example

n 3-Tiers Architecturen Presentation

n JSF + web servern Business

n WS + Application Servern Persistence

n EJB3.0 + ApplicationServer

23

BankCustomer

BankAgent

Browser BrowserCLIENT-SIDE

SERVER-SIDE

Java Server Faces(JSF)

Java Enterprise Edition(EJB 3.0)

Login Sign-UpNew

Account Transfer

MySQL

Services

BUSINESS SERVICES

PERSISTENCE

PRESENTATION


IaaS: Deploying a Glassfish server

n AWS-EC2: AmazonInfrastructure Servicen Provides virtual

Machinen With SSH connection

n Turn-key VMn Nothing installed!

n JREn Databasen Glassfish

24

DEMO


IaaS: Deploying a Glassfish server

nGlassfishn Provide services to

deploy services/appsremotely

25


Cordys Business Operations Platform

26


OMG Modeling languages andZachman Framework

27

Data(What)

Function(How)

Network(Where)

People(Who)

Time(When)

Motivation(Why)

Scope(Contexts)

Business(Concepts)

System(Logic)

Technology(Physics)

Component(Assemblies)

List of thingsimportant

to business

List of processesthat the business

performs

List of locationswhich the business

operates

List of organizationsimportant to the

business

List of events/cyclesimportant to the

business

List of businessgoals/strategies

Semantic ModelBusinessProcessModel

BusinessLogisticsSystem

WorkflowModel

MasterSchedule

BusinessPlan

Logical Data Model ApplicationArchitecture

DistributedSystem

Architecture

HumanInterface

Architecture

ProcessStructure

Business RuleModel

Physical Data Model System Design TechnologyArchitecture

PresentationArchitecture

ControlStructure

RuleDesign

Data Definition Program NetworkArchitecture

SecurityArchitecture

TimingDefinition

RuleDefinition

Operation(Instances)

Data Function Network Organization Schedule Strategy

BMM

SBVR

VDM OSMSBVR

DTFV

BPMN

UMLIMM(CWM)

CMPM

SoaML

ODM

UML

SoaML

BPMN


Context and Goals

Interactions

Interface

Channels

Roles

Actors

Resources

Functions

Tasks

Executors

Processes

Orchestra-tion

Workflows

Information

Data

Stores andMessages

EFAExtra

FunctionalAspects

QoSSLA

Monitoring,adaptation

Inte

ract

ion

Requirements

Func

tion

Coo

rdin

atio

n

Info

rmat

ion

Qua

lity

Design

Implementation

Infrastructure

Stru

ctur

e

RoleModels

UMLClass

Ontologies

Goal orientedUse cases/stories

UIIFMLWebML

OCL

collaboration

withINF5120Modelinglanguages

Model Driven Architecture/MDE

SoaML BPMN


PIM-K

CIM-KOntologies Bus.Process Bus.Rules Goals NFA/Qualities Org

BPDM, SBVR,EDOC,UPMS,

PIM4SOA, ODM

PSM-K

BPMN, POP*, ARIS,ArchiMate,

GERAM, GRAI, Zachman,UEML, B.Rules. ..

Technologies/Realisation-K

UML profiles andmetamodels forBPEL, WSDL, XML,

XPDL,OWL-S, WSML, WSDL-S

TechnologyReliastion

Code

Information Process Services Rules NFA UI

Data Wflow/Comp Interfaces Rules NFA UI

XML, BPEL/XPDL, WSDL, SWRL, Security, AJAXOWL, OWL-S/WSML WSDL-S, Induction, … QoS

Legacy and New systems/services, ERPs/ESAs


Information modeling and OntologyEngineering – Semantic technologiesn Conceptual Modeling – ISO TR 9001n ER Modeling, NIAM Modeling, EXPRESS Modelingn UML Modelingn ISO 19103 – Conceptual Schema with UMLn Semantic Web Modelingn Ontology Engineeringn Logic Formalisms – First order logicn XML Schema, RDF, Topic Maps, OWLn ODM - Ontology Definition Metamodeln Open Linked Data, Open Linked Services

30


Conceptual Modeling

n Conceptsn Conceptual modeln Conceptual schema languagen Conceptual schema

n Why conceptual modeling in 19100?n to provide a rigorous definition of geographic information and

geographic information servicesn to standardize the definition of conceptual schemas


Real World

Conceptualformalism

Conceptual Schema Language(s)

Lexical Languages

Graphical Languages

basis for one or more

ConceptualModel

Universe ofDiscourse

defined in

Conceptual Schema

formally represented in

provides formallanguage forrepresenting

provides conceptsfor describing

From reality to conceptual schema


ConceptualModel


ISO 100% Principle- static aspects- dynamic aspects

InformationSystem

Real World- Problem

domain

Solutiondomain

Basis for

Information modeling Data modeling

1) 2) Conceptualisation principle- exclude physical representation

3) Helsinki principle- agree set of rules

4) Concrete syntax- UML

5) Self description

Principles


ConceptualModel


ISO 100% Principle- static aspects- dynamic aspects

InformationSystem

Real World- Problem

domain

Solutiondomain

Basis for

Information modeling Data modeling

What is relevant ?Information,

Rules, constraintsBehaviour/Dynamics – services, process


Different kind of models

n Conceptual models

n Specification models

n Implementation models


PurchaseProcess2

OrderInfo :OrderInfo

ProceedPurchase2

Handle invoice2

Order Process2

confInfo

DeliveryProcess2

invoiceInfo

Seller2 : ISellerCustomer2 : ICustomer

Activity model

Information model

Interaction model

UML Activity model (or BPMN)

UML component diagram(enhanced in UML 2.0), SoaML

UML Class diagram

LTB HPL

ICustomer ISeller

IProdOwn

LHH

IWareHouse

TRA

ITransportServ

OrderLine+ nrOfProd : Int

<<entity>>

Invoice- date : Date- amount : Currency

<<entity>>

SellerEntity- name : String

<<entity>>

Product+ prodNr : Int+ descript ion : String

<<entity>>

Shipment+ delivAddress : Address

<<entity>>

CustomerEntity+ name : String+ address : String+ phonenr : String

<<enti ty>>

PurchaseOrder- poNr : int- date : Date

<<entity>>

Order+ orderNr : Int+ date : Date

<<entity>>

1. .*1. .*

Informationmodeling

SemanticModels


Parts of UMLMetamodel

ClassAttribute

initialValue : Expression

DataTypeOperation

SpecificationisPolymorphic : Booleanconcurrency : CallConcurrencyKind

GeneralizableElement

Feature

BehaviouralFeatureStructuralFeature

Association

Classifier

0..*

+features

0..*

AssociationEnd2..n2..n

0..*0..*

0..* 1

0..*

1

0..*

0..*


Different kind of models

Idea Systemguide thethoughtprocess

semantics,algorithms,datastructures,performancemeasures, ...

focus onsemanticintent


Information Modeling Guidelines

n Process steps

n Modeling phases

n Modeling harmonization


Process Steps

n A) Problem definition - Enterprise (Analysis)n 1) Identify Area of Concern/Problem definitionn 2) Goals (Use cases)n 3) Interaction scenarios (sequence / collaboration diagrams)n 4) High level domain model / concepts (Class diagrams)

n B) Domain model - Informationn C) Architectural Model - Servicesn D) Implementation modelsn E) Implementationn F) Maintenance


Phase 0: Identify scope and context

Phase 1: Identify basic entities

Phase 1b: Is the modeling approach consistentwith the approach described in the rules forapplication schema document

Phase 2: Specify relationships, attributes andpossibly operations

Phase 3: Completion of constraints

Phase 4: Model definition harmonisation –withsub-models and other work items

Information Modeling


P1 Identify basic entities

n Refine objects/classes from enterprise viewpointn Tasks

n 1) Identify generic and specific conceptsn 2) describe each of the entititesn 3) Watch out for synonyms and homonymsn 4) Create a data dictionary (glossary of terms)

n Questionsn Are entities according to scope?n Can we categorize entities?’n What are the attributes and constraints?

n 1b) Consistent with rules for application schemas?


P2 Specify relationships, attributesand operationsn The entities of interest -

corresponding to units ofinformation

n Any generalizationrelationships betweenentitites

n Other relationships betweenentities

n Check if the existence of anentity depends on theexistence of another entity

n What consistency rules need tobe satisfied in order for aninstance of the entity, or modelas a whole, to be valid.

n The semantics, if any, of thetypes used for attributes

n Which attributes uniquely identifyan entity

n The attributes of the entities andwhether any of them can bederived from other attributes

n Homonyms and synonyms -create a glossary of terms


Checklist

1) Readability2) Scope3) nym principle4) context independence5) implementation independence6) invariance7) constraints8) reality9) redundancy10) concepts11) hierarchies12) simple types

Checklist

1) Readability2) Scope3) nym principle4) context independence5) implementation independence6) invariance7) constraints8) reality9) redundancy10) concepts11) hierarchies12) simple types

Before harmonization

n Review your work criticallyn RATIONALIZE

n Remove redundancy:n Ensure that each relationship

is only specified once.n Ensure that each constraint

is captured only once, andthat multiple constraints areseparated.

n Ensure that if one entity isdependent on another, thenthe relationship is specifiedfrom the dependent to theindependent.


Model harmonization

n Are there redundancies or conflicts?

n Are there any ambiguities?

n Are the models complete?

n Do the abstraction levels match?


More semantics using ontologies

n Models can be annotatedn Creating a connection between the models and a formal ontology

n The annotations can be used in order to ”automate”mapping definitionsn Not complete automationn Can be used for aiding mapping designer through tool support

n The reasoning power from the world of ontologies comesinto play heren Query the ontology system for things like

n ”Give me a list of elements in Model B that are annotated with similarontology concepts as element X in Model A”


Ontologies

n Ontologies provide a shared understanding of a domainn They provide background knowledge to systems to automatize certain

tasksn By the process of annotation, knowledge can be linked to ontologies

n Example: “Angelina Jolie” (Text) linked to concept Actressn In our ontology we also know that an actress always is female and a

personn Ontologies allow the creation of annotations à machine-readable and

machine-understandable contentn If machines can understand content, they can also perform more

meaningful and intelligent queriesn Distinction of Jaguar the animal and the carn Combination of information that is distributed on the Web

47


Definitions

n An ontology defines the basic terms and relations comprising thevocabulary of a topic area, as well as the rules for combining termsand relations to define extensions to the vocabulary

n An ontology is a hierarchically structured set of terms for describing adomain that can be used as a skeletal foundation for a knowledgebase

n An ontology provides the means for describing explicitly theconceptualization behind the knowledge represented in a knowledgebase

n An ontology is a formal, explicit specification of a sharedconceptualization


Examples

49


Features of an ontology

n Modelled knowledge about a specific domainn Defines

n A common vocabularyn The meaning of termsn How terms are interrelated

n Consists ofn Conceptualization and implementation

n Containsn Ontological primitives

50


Catalog/ID

Thessauri“narrower term”

relationFormal

is-aFrames

(properties)

GeneralLogical

constraints

Terms/glossary

Informalis-a

Formalinstance

ValueRestrs.

Disjointness,Inverse, part-Of ...

Classifications of ontologies

51

Lassila O, McGuiness D. The Role of Frame-Based Representation on the Semantic Web.Technical Report. Knowledge Systems Laboratory. Stanford University. KSL-01-02. 2001.


Languages for building ontologies

n Ontologies can be built using various languages with variousdegrees of formalityn Natural languagen UMLn ERn OWL/RDFSn WSMLn FOLn ...

n The formalism and the language limit the kind of knowledgethat can be represented

n A domain model is not necessarily a formal ontology onlybecause it is written in OWL

52


Applications of ontologies

n Knowledge representationn Ontology models domain knowledge

n Semantic annotationn Ontology is used as a vocabulary, classification or indexing

schema for a collection of itemsn Semantic search

n Ontology is used as a query vocabulary or for query rewritingpurposes

n Configurationn Ontology defines correct configuration templates

53


Principles for the design of ontologies

n Clarityn To communicate the intended meaning of defined terms

n Coherencen To sanction inferences that are consistent with definitions

n Extendibilityn To anticipate the use of the shared vocabulary

n Minimal Encoding Biasn To be independent of the symbolic level

n Minimal Ontological Commitmentsn To make as few claims as possible about the world

54

Gruber, T.; Towards Principles for the Design of Ontologies.KSL-93-04. Knowledge Systems Laboratory. Stanford University. 1993


Ontology engineeringn The set of activities that concern the ontology development process,

the ontology life cycle, and the methodologies, tools and languages forbuilding ontologies

n Ontology life cycle

55

Domain analysismotivating scenarios, competency questions, existing solutions

Conceptualizationconceptualization of the model, integration and extension of

existing solutions

Implementationimplementation of the formal model in a representation language

Usage/Maintenanceadaptation of the ontology according to new requirements

Know

ledge acquisition

Evaluation

Docum

entation

Project management: controlling, planing, quality assurance etc.


Ontology engineering activitiesManagement SupportDevelopment oriented


Methodologies for building ontologies

n Validated guidelines on how the ontology building processshould be structured

n Well-organized process instead of „ontology developmentdriven by inspiration and intuition only“

n Methodologies do not support all of the aforementionedactivities

n They implicitly assume a particular development paradigmfor the ontology engineering process

n Some of them also provide supporting methods and toolsn There are many different methodologies for building

ontologies

57


How to build an ontology?Natalya F. Noy and Deborah L. McGuinness. ̀ `Ontology Development 101: A Guide to Creating Your First Ontology''. Stanford Knowledge SystemsLaboratory Technical Report KSL-01-05 and Stanford Medical Informatics Technical Report SMI-2001-0880, March 2001.

n Step 1: Determine the domain and scope of the ontologyn What is the domain that the ontology will cover?n For what we are going to use the ontology?n For what types of questions the information in the ontology should

provide answers?n Who will use and maintain the ontology?

n Competency Questionsn A set of queries which place demands on the underlying ontologyn Ontology must be able to represent the questions using its

terminology and the answers based on the axiomsn Ideally, in a staged manner, where consequent questions require

the input from the preceeding onesn A rationale for each competency question should be given

58


How to build an ontology? (cont’)Natalya F. Noy and Deborah L. McGuinness. ̀ `Ontology Development 101: A Guide to Creating Your First Ontology''. Stanford Knowledge SystemsLaboratory Technical Report KSL-01-05 and Stanford Medical Informatics Technical Report SMI-2001-0880, March 2001.

n Step 2: Consider reusing existing ontologiesn Reuse ensures interoperability and reduces costsn Ontology libraries and tools for customization are required for this stepn Sub-steps

n Discover potential reuse candidatesn Evaluate their usabilityn Customize ontologies to be reusedn Integrate and merge to the target ontology

n Step 3: Enumerate important terms in the ontologyn What are the terms we would like to talk about?n What properties do those terms have?n What would we like to say about those terms?

59



n Step 4: Define classes and class hierarchyn From the list created in Step 3, select the terms that describe objects

having independent existence rather than terms that describe theseobjectsn These terms will be classes in the ontology

n Organize the classes into a hierarchical taxonomy by asking if by being aninstance of one class, the object will necessarily (i.e., by definition) be aninstance of some other classn If a class A is a superclass of class B, then every instance of B is also

an instance of An Classes as unary predicates—questions that have one argument. For

example, “Is this object a wine?”n Later: binary predicates (or slots)—questions that have two

arguments. For example, “Is the flavor of this object strong?” “What isthe flavor of this object?”

60



n Step 5: Define attributes and relationshipsn Step 4 selected classes from the list of terms we created in Step 3

n Most of the remaining terms are likely to be properties of theseclasses

n For each property in the list, we must determine which class itdescribes

n Types of propertiesn “intrinsic” propertiesn “extrinsic” propertiesn parts, if the object is structured (physical or abstract)n relationships to other individuals

n Properties are inherited and should be attached to the most general classin the hierarchy

61


MagicDraw – Models with code generation sets


UML models

63


Quick guided tour of the language

nGeneral mechanismsn Packagesn Classes

n UML profilesn Stereotypesn Tagged valuesn Notesn Constraints

n Basic diagramsn Class diagramsn Object diagramsn Use case diagramsn Sequence diagramsn Collaboration diagramsn Statechart diagramsn Activity diagramsn Component diagramsn Deployment diagrams

64


UML Information Modeling

n Ref also ISO 19103 Standard for Conceptual Modeling

n The following material is for reference …..

65


Packages

nModelsstructuring

66


Class diagrams

n Class diagrams represent a set of classes, interfacescollaborations, and their relationships.

n They are the most frequent diagrams in the objectssystems modeling.

n They offer the static conceptual view of a system.n Class diagrams, (including active classes), offer the static

process view of a system.

67

Zebra<<interface>>

Zebu


Classes

n A class is an abstract descriptionof set of objects

n A class can be seen as afactorization of commonelements to a set of objects

n A class describes the definitiondomain of a set of objects

n Description conceptually split intwo partsn A class specification which defines

the definition domain and theproperties of the class instances(datatype)

n A realization defines how aspecification is realized

n Classes featuresn Attributesn Operationsn Réceptionsn Relationsn Multiplicityn Persistencen Component

68


Objects

Can represents

n One instance

n One type, interface

n One class

Ola :Person

Person

<<Interface>>Person

Objekt modellering


Object and classes- notation

Person

navn : stringpersonnr. : integeradresse : stringgasje : moneystillingstittel : string

endre-stilling()endre-adresse()

Example - object class

Objekt modellering


Objektklasser -notasjon

Klassenavn

attributtnavn-1 : datatype-1 = defaultverdi-1attributtnavn-2 : datatype-2 = defaultverdi-2

operasjonsnavn-1 (argumentliste-1) : resultattype-1operasjonsnavn-2 (argumentliste-2) : resultattype-2

. . . .

. . . .

Objekt modellering


GM_Primitive

+ boundary() : Set(GM_Primitive)

(from GeomPrimitive) GM_Complex(from GeomComplex)

GM_Object+ mbRegion : GM_Object+ representativePoint : DirectPosition

+ SRS() : SpatialReferenceSystem+ transform(SRS : SpatialReferenceSystem) : GM_Object+ equals(object : GM_Object) : Boolean+ distance(object : GM_Object) : Distance+ dimension() : Integer+ dimension(point : DirectPosition) : Integer+ coordinateDimension() : Integer+ envelope() : Envelop+ boundary() : Set(GM_Object)+ buffer(radius : Distance) : GM_Object+ convexHull() : GM_Object+ centroid() : DirectPosition+ representativePoint() : GM_Point+ isInComplexes() : Set(GM_Complex)+ isPartOf(geomCplx : GM_Complex) : Boolean+ universe() : GM_Complex

SpatialReferenceSystem(from DirectPositioningSchema)

1

+srs

1

Class diagram


Class attributes

ClassName

/ /* derived attribute

+ /* public visibility# /* protected visibility- /* private visibilityunderline /* class scope


[visibility] name [multiplicity] [:type] [= initial value] [{property-string}]

+ origin [0..1] : Point = (0,0) {frozen}

defined properties: changeable, addOnly, frozen (const)

Class attributes


Attributes and data types

n A data type specifies a legal value domain andthe operations on values of that domain

n Four categoriesn a) Basic data types (integer, real, string, …)n b) Collection data types (from OCL)n c) Enumerated data types (user-definable finite sets)n d) Model types


Operations

An operation is a specification of• a transformation, or• a query

A method is a procedure thatimplements an operation.


Operations

[visibility] name [(parameter-list)] [:return-type] [{property-string}]

[(parameter-list element)] ::=[direction] name : type [= default-value]

[direction] ::= in | out | inout

+ set ( in name : Name, in place : String = ‘Oslo’): Boolean {concurrency=sequential}

defined properties: isQuery, concurrency, ...


Class diagrams

78

parameter

parameterized class

a class another classassociation name

associationclass nameattributesoperations

a class another classassociation name

/ derived association

role r1 role 2

class name

attributesattr_name : type = initial value

operationsop_name (arg_list):type

class name

- private attribute+ public attribute# protected attribute

- private operation+ public operation# protected operation


Properties visibility

n Public +Accessible to all object within the

systemn Protected #

Accessible to instances ofimplementing class and itssubclasses

n Private -Accessible to instances of the

implementing classn Underlined

Class variable/operation

79

+public_operation()#protected_operation()-private_operation()

+public_attribute#protected_attribute

-private_attribute

Thing


Relations between classes

n Associationn Aggregationn CompositionnGeneralizationn Dependency

n An associationexpresses a bi-directional semanticconnection betweenclasses

n An association is anabstraction of the linksexisting between theinstances objects ofthe associated classes

80


Optional/Conditional

n In UML all attributes and operations are perdefault mandatory.

n Optional data values for attributes can beshown through multiplicity [0..1].

n Conditional should relate to a note withconstraint expressed in text/OCL (ISO19103)


Relationships

Association (the “glue”)A semantic relationship between elements(involves connections among their instances)

Generalization (inheritance)A relationship between an elementand the sub-elements that may be substitutedfor it

DependencyThe use of one element by another

Refinement/realisationA shift in levels of abstraction


Purpose: To show relationships between model entitiesTo define multi-way constraints

AssociationNameRoleNameRoleName

1..*2

Multiplicity ofan Association

Dependency (Client-Server) Composition (Strong Aggregation)

Generalization Aggregation

Link AttributeAssociation Notationused to “anchor” at noteto a model entity

Different relationships in UML Diagrams


Generalisation and Association

WomanMarried with

0..1 0..1

Person Bookloans

Loancustomer

Loan-entity

0..1 0..*

Man

{not siblings}

Generalisation

Superclass

Subclasses (inherits attributes and operations)

wifehusband

Arrow can be used toshow direction ofknowledge

Association


Example

85


Associations naming

n Indicating the reading direction

86

University StudentWelcomes

University StudentStudies in


Roles naming

n A role describes an association end

87

University Person+Student

+Employer +Teacher


Roles multiplicity

88

1 One only0..1 Zero or oneM .. N From M to N (natural numbers)* Several0 .. * Zero to several1 .. * One to several


Roles visibility

n Publicn Protectedn Private

89

A

B

+Public role

*

-Private role

*

C

*

#Protected role

*


Navigability

n given an unlabelled associationbetween two classes, it ispossible to navigate from objectsof one kind to objects of the otherkind.

n by default an association is bi-directional.

n a navigability indication suggeststhat an object at one end isdirectly and easily reachable froman object at the other end.

n putting it into practice, forinstance a source objectmemorizes a direct reference tothe target objects.

n Given a user, it is desirable to haveaccess to his / her passwords

n Given a password, it is not desirableto have access to the correspondinguser

90


Example

91


Aggregation

n Antisymmetric bi-directional semantic connectionsn An association expressing a stronger coupling between

classesn Relationship representation

n master and slaven whole and partsn composed and composing

92


Composition

n Physical compositionmodelingn Maximum multiplicity of

1 on the aggregate siden Automatic propagation

of the destruction

93

Aggregate Part

1 *


Classes hierarchy

nManaging thecomplexityn Class tree of increasing

abstraction

94

Sub-class

Super-class

More generalclass

More specificclass


Multippel arving - eksempel

Ansatt

Selger Sjef

DirektørSalgssjef

{overlapping}

Objekt modellering


Multiplicity constraints

Multiplicity shows how many instances of a class that canbe related to one instance of the class at the other end ofthe association

Class

Class

Class

Class

Class

0..*

1

0..1Class*

1..*

3,5,10


Aggregation

Loose (weak) part-of relationship

Phonebook Subscription

0..* *

Open diamond

Multiplicity > 1 possibleExists even if a phone bookgoes away

listedIn subscriptions

associationrole names


Composition(“strong aggregation”)

Strong part-of relationship

File Recordcontains1 *

Filled diamond

Multiplicity max 1 Existence dependent


Associations

Fig. 3-31, UML Notation Guide

Person

Manages

JobCompany

boss

worker

employeeemployer1..*

*

*

0..1

Job

Account

Person

Corporation

{Xor}

salary


Association Ends

Fig. 3-32, UML Notation Guide

Polygon PointContains

{ordered}

3..*1

GraphicsBundle

colortexturedensity

1

1

-bundle

+points


Association class

Are used instead of link-attributes ifnThe association are related to other objectsnOperations are attached to the association

Person BookLoans

Loan dateDelivery data

Loan

CheckDate()Expired loans

1 1..*

*


Constraints on relationships

Portfolio

BankAccount

{secure}

Persont

Corporation

{or}

Department

Person

member manager

{subset}

* *

1..* 1


Geometry Package(from Logical View)+ Accuracy+ TopologicalGeometry+ RepresentationalGeometry+ TemporalGeometry+ Geometry+ Vector{dimension}+ SpatialVector

Packages in UML

n This is a grouping of modelelements and diagrams.

n Package dependencies usuallysummarize dependencies amongthe contents of the packages.

n Packages can contain otherPackages.

n Packages can show theClass/Entities found in a givenPackage.


Containment and dependency

Foundation

ModelManagement

BehavioralElements

CoreExtensionMechanisms

DataTypes


Commentsandconstraints

<<requirement>>Shall conform to ….

Notes:

Common Mechanisms

n Adornments: notesn Extensibility mechanisms: stereotypes,

tagged values, constraints


Stereotypes

n Used to define derivative modeling conceptsbased on existing generic modelingconcepts

n Defined by:n base (meta-)class = UML meta-class or

stereotypen constraintsn required tags (0..*)

n often used for modeling pseudo-attributesn icon

n A model element can have at most onestereotype


Tagged values - properties

Name (tag) separator (=) value (of the tag)

properties on an element - relevant for code generation orconfiguration management (Can be applied to all UML elements)

Server{processors = 3}

Billing{version=3.2status=checkedOutby=ajb}


OCL - Object Constraint Language

First order predicate logic, for Boolean expressions - included in UML 1.1

Can be used for:invariants,value restrictions,pre- and postconditions

Personself.age > 0

Married people are of age >= 18self.wife->notEmpty implies self.wife.age >= 18 andself.husband->notEmpty implies self.husband.age >= 18 and

Expressions with:and, or, not, implies, exists, for all,Collections (select, reject, collect, iterate)


UML profiles

109


UML profiles

n They allow us to adapt the UML language to the needs ofthe analysts or the application domain

n Allows designers to model using application domainconcepts.

n There are three extension mechanisms:n Stereotypesn Tagged valuesn Constraints


UML for a specific purpose (benefits)= a specific profilen Profiles exist for :

n technical targets (C++, RDB, Java, CORBA, XML, Real Time, etc.)n specific disciplines (Analysis, Design, etc.)n requirement analysisn tests (test for Java, test for EJB, test for C++, etc.)n metrics, quality checking, and profiles for managing configuration

management rules.n From early requirement down to final tests, UML profiles

can drive each specific discipline involved.

111


Stereotype

n Extends the vocabulary of UML with new construction elementsderived from existing UML but specific to a problem domain

n Can have associated tagged values and constraintsn Possibility of assigning an icon for a better graphical representation

DB Partners


Examples of other stereotypes


Tagged value

n Is a property associated to a model elementn Used to store information about the element

n Management information, documentation, coding parameters, ...n Generally, the tools store this information but it is not shown in the

diagrams


Constraints

n Any semantic relationship between model elementsn Expressed in OCL (Object Constraint Language) or in

natural languagen {constraint}, inv, pre-, post-condition

115


Profile: Example (Usage of UMLextensions)n Below is a user model example, with extensions. That example

supports extensions.n We will see how they are defined at the profile definition level.

116

<<persistent>>Customer

{storageMode = hash}

<<identifier>> number : int

Stereotype

Tagged Value

ClassAttribute

PROFILE

UML

EditCheckGenerate SQL…

Operation


Extending Metaclasses

n A profile consists primarily of stereotypesn restricted metaclasses that can only be used to extend existing

metaclasses

117

«profile» EJB

«metaclass»Component

«stereotype»Bean

{required}

«stereotype»Session

«stereotype»Entity

state: StateKind

«enumeration»StateKind

statelessstateful


Applying a Profile

118

«profile» EJB

«metaclass»Component

«stereotype»Bean

{required}

«stereotype»Session

«stereotype»Entity

state: StateKind

«enumeration»StateKind

statelessstateful

UserModel

«apply»

«session»Customer

name: String

state=stateless

Applying a profile makes the profile extension available to themodel packages to which it is applied.


Implementation neutral! Not implementation specific!

ISO 19103 – Conceptual modeling with UML- a UML profile for Information modeling

n Basic data typesn Stereotypesn Namingn Documentation of modelsn Information modelling guidelines


Basic Types

n Primitiven Decimal, Integer, Number, Real, Vector, Character,

CharacterString, Date, Time DateTime, Binaryn Sign, Boolean, Logical, Probability

n Collectionsn Set<T>, Bag<T>, Sequence<T>, Dictionary<K,T>

n Measuresn Angle, Area, Distance, length, Scale, MTime, Velocity, Volume,

UnitOfMeasuren Records, Namespace


Number

Number<<BasicType>>

Integer<<BasicType>>

Real<<BasicType>>


Sign+ positive+ negative

<<Enumeration>>

MD_Restrictions1+ copyright = 1+ patent = 2+ patentPending = 3+ trademark = 4+ licence = 5+ intellectualPropertyRights = 6+ restricted = 7+ otherRestric tions = 8

<<CodeL ist>>

1 copyr ight2 patent3 patentPending4 trademark5 licence6 intel lectualProperty Rights7 res tricted8 otherRestr ict ions

MD_Restrictions2<<CodeList>>

DirectPosition+ coordinate : Sequence<Num ber>/+ dim ension : Int eger

<<DataType>>

GM_Posit ion+ di rect : DirectPosition+ indirect : GM_PointRef

<<Union>>

User defined data types

n <<DataType>>

n <<Enumeration>>

n <<CodeList>>

n <<Union>>


Class Stereotypes

n NONE

n <<Interface>>

n <<Type>>

n <<Abstract>>

n <<MetaClass>>

GM_GenericCurve

+ startPoint() : DirectPosition+ endPoint() : DirectPosition+ param(s : Distance) : DirectPosition+ tangent(s : Distance) : Vector+ ...()

<<Int erface>>

GM_LineString+ controlPoint : GM_PointArray

+ GM_LineString(points[2..*] : GM_Position) : GM_LineString+ asGM_LineSegment() : Sequence<GM_LineSegment>

GM_Curve<<Type>>

GM_CurveSegment+ interpoilation : GM_CurveInterpolation = "linear"+ numDerivatesAtStart[0,1] : Integer = 0+ numDerivatesAtEnd[0,1] : Integer = 0+ numDerivateInterior[0,1] : Integer = 0

+ samplePoint() : GM_PointArray+ boundary() : GM_CurveBoundary+ reverse() : GM_CurveSegment

<<Abstract>>

0..1 1..*

+curve

0..1

+segment

1..*Segmentation


GM_CurveSegmentMD_CitationGM_FeatureAttribute

Class (package, association) names

n All classes must have unique names.n Names shall start with upper case letter.n Should not have a name that is based on its

external usage, since this may limit reuse.n Should not contain spaces

n Separate words in a class should be concatenated, e.g.“XnnnYmmmm”

n Class names should start with bialpha prefixes foreach standard part.


computePartialDerivatescompute Partial Derivatescompute_partial_derivatives

Attribute, operation and role names

n Shall start with lower case lettern Concatenate words shall begin with capital lettern Do not repeat class names in attribute namesn Keep names technical, meaningful and short, if possible


GM_Curve<<Type>>

GM_CurveSegment<<Abstract>>

1..*0..1

+segment

1..*

+curve

0..1

Segmentation

Associations

n Multiplicity shall be defined for both associations endsn All associations ends (roles) representing the direction

of a relationship must be named or else the associationitself must be named

n The role name must be unique within the context of aclass and its supertypes.

n The direction of an association must be specified


NB! Font size > 8pt

GM _Curve<<T ype>>

Documentation of modelsn Diagrams

n Package dependency diagramsn Class diagramsn Class context diagrams

n Textn Class

n General description, semantics,supertypes/subtypes

n Attributes, Associationsn Operations

n preconditions, input/output parameters, return value,post conditions, exceptions, constraints

n Constraints

inf5120 – model-based system development · telecom and informatics 3 inf5120 - lecture plan -...

Documents