115
Ontologies and the Ontologies and the Semantic Web II Semantic Web II Deborah L. McGuinness Deborah L. McGuinness Associate Director and Senior Associate Director and Senior Research Scientist Research Scientist Knowledge Systems Laboratory Knowledge Systems Laboratory Stanford University Stanford University Stanford, CA 94305 Stanford, CA 94305 650-723-9770 650-723-9770 [email protected] [email protected]

Ontologies and the Semantic Web II Deborah L. McGuinness Associate Director and Senior Research Scientist Knowledge Systems Laboratory Stanford University

Embed Size (px)

Citation preview

Page 1: Ontologies and the Semantic Web II Deborah L. McGuinness Associate Director and Senior Research Scientist Knowledge Systems Laboratory Stanford University

Ontologies and the Ontologies and the Semantic Web IISemantic Web II

Deborah L. McGuinnessDeborah L. McGuinnessAssociate Director and Senior Associate Director and Senior

Research ScientistResearch ScientistKnowledge Systems LaboratoryKnowledge Systems Laboratory

Stanford UniversityStanford UniversityStanford, CA 94305Stanford, CA 94305

650-723-9770650-723-9770 [email protected]@ksl.stanford.edu

Page 2: Ontologies and the Semantic Web II Deborah L. McGuinness Associate Director and Senior Research Scientist Knowledge Systems Laboratory Stanford University

August 9, 2004 2Deborah L. McGuinness

OutlineOutline Getting StartedGetting Started

Identify requirementsIdentify requirements Representation and reasoning examplesRepresentation and reasoning examples Environment compatibility (people, tools, languages)Environment compatibility (people, tools, languages)

Requirements dictate language and tool choicesRequirements dictate language and tool choices Ontology Development 101Ontology Development 101 Tools supporting applicationsTools supporting applications

EditorsEditors ReasonersReasoners Evolution EnvironmentsEvolution Environments

DiscussionDiscussion

This tutorial is based loosely on Ontology Development 101 (with Noy), OWL This tutorial is based loosely on Ontology Development 101 (with Noy), OWL Overview (with van Harmelen), and OWL Guide (with Smith and Welty), and Overview (with van Harmelen), and OWL Guide (with Smith and Welty), and How and When to Live with a Kl-ONE-like System (with Brachman, Borgida, How and When to Live with a Kl-ONE-like System (with Brachman, Borgida, and Resnick). Viewgraph input was also solicited from viewgraphs and/or and Resnick). Viewgraph input was also solicited from viewgraphs and/or interactions with Dean, Kendall, Noy, Pollock, & Stavert. interactions with Dean, Kendall, Noy, Pollock, & Stavert.

Page 3: Ontologies and the Semantic Web II Deborah L. McGuinness Associate Director and Senior Research Scientist Knowledge Systems Laboratory Stanford University

August 9, 2004 3Deborah L. McGuinness

What does your What does your application need?application need?

A controlled languageA controlled language A hierarchical structure (possibly for use A hierarchical structure (possibly for use

with query expansion)with query expansion)Red wines such as zinfandel, cabernet sauvignon, red Red wines such as zinfandel, cabernet sauvignon, red

burgundy, chianti, …burgundy, chianti, … Knowledge supporting structured queries –Knowledge supporting structured queries –

Find resumes of people who are academics, Find resumes of people who are academics, authors of semantic web languages, who have authors of semantic web languages, who have deployed applications, who worked in industry, deployed applications, who worked in industry, who consultwho consult

Precise dependable inferencePrecise dependable inference Computational tractabilityComputational tractability ExplainabilityExplainability

Page 4: Ontologies and the Semantic Web II Deborah L. McGuinness Associate Director and Senior Research Scientist Knowledge Systems Laboratory Stanford University

August 9, 2004 4Deborah L. McGuinness

More requirements?More requirements? Open world information gathering Open world information gathering

You may not know all of the information at any You may not know all of the information at any given time and more information may be given time and more information may be deduced or input along the waydeduced or input along the way

Consistency checking will be important – Consistency checking will be important – check the input as we put in someone’s check the input as we put in someone’s age (that it is not 200 for example)….age (that it is not 200 for example)….

Generate some use cases containing Generate some use cases containing information that is expected to be available information that is expected to be available and questions that need to be answerableand questions that need to be answerable

Page 5: Ontologies and the Semantic Web II Deborah L. McGuinness Associate Director and Senior Research Scientist Knowledge Systems Laboratory Stanford University

August 9, 2004 5Deborah L. McGuinness

While gathering While gathering requirements…requirements…

Identify user community(ies)Identify user community(ies) Language compatibilities Language compatibilities Environmental compatibilitiesEnvironmental compatibilities Legacy systemsLegacy systems

Page 6: Ontologies and the Semantic Web II Deborah L. McGuinness Associate Director and Senior Research Scientist Knowledge Systems Laboratory Stanford University

August 9, 2004 6Deborah L. McGuinness

OntologiesOntologies

““A Specification of a A Specification of a Conceptualization” – Gruber ’91Conceptualization” – Gruber ’91

Ontologies provide and explicit and Ontologies provide and explicit and declarative description of a domain:declarative description of a domain: concepts – descriptions of classesconcepts – descriptions of classes properties/attributes of conceptsproperties/attributes of concepts constraints on properties and attributesconstraints on properties and attributes individualsindividuals

Page 7: Ontologies and the Semantic Web II Deborah L. McGuinness Associate Director and Senior Research Scientist Knowledge Systems Laboratory Stanford University

August 9, 2004 7Deborah L. McGuinness

Benefits of OntologiesBenefits of Ontologies

Shared vocabulary (for humans and Shared vocabulary (for humans and agents)agents)

Shared common understanding of Shared common understanding of the structure of information the structure of information

Reuse of domain knowledgeReuse of domain knowledge to avoid “re-inventing the wheel”to avoid “re-inventing the wheel” to introduce standardsto introduce standards

Page 8: Ontologies and the Semantic Web II Deborah L. McGuinness Associate Director and Senior Research Scientist Knowledge Systems Laboratory Stanford University

August 9, 2004 8Deborah L. McGuinness

More BenefitsMore Benefits Assumptions become explicit enablingAssumptions become explicit enabling

Explaining assumptionsExplaining assumptions Changing assumptionsChanging assumptions Hypothetical reasoning (multiple scenarios)Hypothetical reasoning (multiple scenarios) Support for evolving systems where time and Support for evolving systems where time and

situations change necessitating re-evaluation of situations change necessitating re-evaluation of assumptionsassumptions

Support for interoperation with other Support for interoperation with other (potentially legacy) systems(potentially legacy) systems

Separation of types of knowledge:Separation of types of knowledge: Declarative domain knowledge vs procedural Declarative domain knowledge vs procedural

knowledgeknowledge Background (unchanging) knowledge from Background (unchanging) knowledge from

changing informationchanging information Authoritative vs. other….Authoritative vs. other….

Page 9: Ontologies and the Semantic Web II Deborah L. McGuinness Associate Director and Senior Research Scientist Knowledge Systems Laboratory Stanford University

August 9, 2004 9Deborah L. McGuinness

Ontology DevelopmentOntology Development

Define domain terms and inter-Define domain terms and inter-relationshipsrelationships Define concepts in the domain (classes)Define concepts in the domain (classes) Identify subclass/superclass relationships Identify subclass/superclass relationships

(thereby defining a class hierachy.(thereby defining a class hierachy. Identify attributes/properties/slots Identify attributes/properties/slots Restrict slot values Restrict slot values Define individuals Define individuals Define interrelationships between Define interrelationships between

individuals (filling in slots)individuals (filling in slots)

Page 10: Ontologies and the Semantic Web II Deborah L. McGuinness Associate Director and Senior Research Scientist Knowledge Systems Laboratory Stanford University

August 9, 2004 10Deborah L. McGuinness

Ontology Development Ontology Development ProcessProcess

Determine the scope of the domain informationDetermine the scope of the domain information Identify resources that may be appropriate to Identify resources that may be appropriate to

reuse/integratereuse/integrate Identify terms in the vocabularyIdentify terms in the vocabulary Define classes/conceptsDefine classes/concepts Define propertiesDefine properties Add restrictionsAdd restrictions Generate individualsGenerate individuals

Note that this process is iterativeNote that this process is iterative

Page 11: Ontologies and the Semantic Web II Deborah L. McGuinness Associate Director and Senior Research Scientist Knowledge Systems Laboratory Stanford University

August 9, 2004 11Deborah L. McGuinness

Determine Application Determine Application Domain and Use Case(s)Domain and Use Case(s)

Describe the domain for the ontologyDescribe the domain for the ontology Describe how the ontology will be usedDescribe how the ontology will be used Identify types example questions and Identify types example questions and

answers for the ontology-supported answers for the ontology-supported applicationapplication

Identify ontology users, owners, and Identify ontology users, owners, and maintainersmaintainers

Describe anticipated evolution path for the Describe anticipated evolution path for the ontology and the applicationontology and the application

Page 12: Ontologies and the Semantic Web II Deborah L. McGuinness Associate Director and Senior Research Scientist Knowledge Systems Laboratory Stanford University

August 9, 2004 12Deborah L. McGuinness

Example QuestionsExample Questions Which wine characteristics determine wine choice? Which wine characteristics determine wine choice? What is the color of a burgundy? What is the color of a burgundy? Is rose wine made in burgundy?Is rose wine made in burgundy? Describe a wine that matches a shellfish dishDescribe a wine that matches a shellfish dish Should I serve zinfandel with crab?Should I serve zinfandel with crab? What wines in my wine cellar are recommended What wines in my wine cellar are recommended

choices to be served with pasta with a spicy red choices to be served with pasta with a spicy red sauce?sauce?

What wines could I buy from wine.com that are What wines could I buy from wine.com that are recommended to be served with to be served with recommended to be served with to be served with Deborah’s specialty?Deborah’s specialty?

Page 13: Ontologies and the Semantic Web II Deborah L. McGuinness Associate Director and Senior Research Scientist Knowledge Systems Laboratory Stanford University

August 9, 2004 13Deborah L. McGuinness

ReuseReuse Why reuse other ontologies?Why reuse other ontologies?

to interoperate with other to interoperate with other ontologies/controlled vocabulariesontologies/controlled vocabularies

to leverage other people’s ontology to leverage other people’s ontology building workbuilding work

to use previously validated and/or to use previously validated and/or authoritative source ontologiesauthoritative source ontologies

to interact with applications and/or tools to interact with applications and/or tools that use other ontologiesthat use other ontologies

……..

Page 14: Ontologies and the Semantic Web II Deborah L. McGuinness Associate Director and Senior Research Scientist Knowledge Systems Laboratory Stanford University

August 9, 2004 14Deborah L. McGuinness

Reuse Starting PointsReuse Starting Points Ontology Libraries/RegistriesOntology Libraries/Registries

DAML ontology library DAML ontology library (www.daml.org/ontologies)(www.daml.org/ontologies) Ontolingua ontology library Ontolingua ontology library

(www.ksl.stanford.edu/software/ontolingua/)(www.ksl.stanford.edu/software/ontolingua/) SchemaWeb - SchemaWeb - http://www.schemaweb.info/ new http://www.schemaweb.info/ new

evolving collectionevolving collection Ontaria – W3C collection of Semantic Web Data Ontaria – W3C collection of Semantic Web Data

http://www.w3.org/2004/ontaria/ http://www.w3.org/2004/ontaria/ Upper ontologiesUpper ontologies

IEEE Standard Upper Ontology Working IEEE Standard Upper Ontology Working Group Group ((suo.ieee.org, suo.ieee.org, http://ontology.teknowledge.com/)http://ontology.teknowledge.com/)

Cyc Cyc (www.cyc.com)(www.cyc.com)

Page 15: Ontologies and the Semantic Web II Deborah L. McGuinness Associate Director and Senior Research Scientist Knowledge Systems Laboratory Stanford University

August 9, 2004 15Deborah L. McGuinness

Reuse Reuse

General ontologiesGeneral ontologies DMOZ DMOZ (www.dmoz.org)(www.dmoz.org) WordNet WordNet

((www.cogsci.princeton.edu/~wn/)www.cogsci.princeton.edu/~wn/) Domain-specific ontologiesDomain-specific ontologies

COGNACOGNA US Geography SurveyUS Geography Survey Geospatial Data StandardsGeospatial Data Standards National GeoSpatial Clearinghouse …National GeoSpatial Clearinghouse …

Page 16: Ontologies and the Semantic Web II Deborah L. McGuinness Associate Director and Senior Research Scientist Knowledge Systems Laboratory Stanford University

August 9, 2004 16Deborah L. McGuinness

Usually there are many Usually there are many starting pointsstarting points

Page 17: Ontologies and the Semantic Web II Deborah L. McGuinness Associate Director and Senior Research Scientist Knowledge Systems Laboratory Stanford University

August 9, 2004 17Deborah L. McGuinness

More…More…

Page 18: Ontologies and the Semantic Web II Deborah L. McGuinness Associate Director and Senior Research Scientist Knowledge Systems Laboratory Stanford University

August 9, 2004 18Deborah L. McGuinness

Identify TermsIdentify Terms

List the nouns in your domain – what classes of List the nouns in your domain – what classes of things will you talk about?things will you talk about?

List the verbs or relationships between things in List the verbs or relationships between things in your domainyour domain

List the attributes / properties of the termsList the attributes / properties of the terms Generate example descriptionsGenerate example descriptions

Living with Classic suggests a brainstorming Living with Classic suggests a brainstorming session just to collect the controlled vocabularysession just to collect the controlled vocabulary

CISCO’s CAT process does a similar thingCISCO’s CAT process does a similar thing

Page 19: Ontologies and the Semantic Web II Deborah L. McGuinness Associate Director and Senior Research Scientist Knowledge Systems Laboratory Stanford University

August 9, 2004 19Deborah L. McGuinness

Example Wine TermsExample Wine Terms

wine, grape, winery, location, wine, grape, winery, location,

wine color, wine body, wine flavor, wine color, wine body, wine flavor, sugar contentsugar content

white wine, red wine, Bordeaux winewhite wine, red wine, Bordeaux wine

food, seafood, fish, meat, vegetables, food, seafood, fish, meat, vegetables, cheesecheese

Page 20: Ontologies and the Semantic Web II Deborah L. McGuinness Associate Director and Senior Research Scientist Knowledge Systems Laboratory Stanford University

August 9, 2004 20Deborah L. McGuinness

Define Classes and the Class Define Classes and the Class HierarchyHierarchy

A class is a concept in the domainA class is a concept in the domain a class of winesa class of wines a class of wines from a particular region (burgundies)a class of wines from a particular region (burgundies)

A class is a collection of elements with some A class is a collection of elements with some similar propertiessimilar properties

A class contains necessary conditions for A class contains necessary conditions for membership (made from a wine grape, alcohol membership (made from a wine grape, alcohol content > xx percent)content > xx percent)

Instances of classesInstances of classes A particular bottle of wine in your wine cellarA particular bottle of wine in your wine cellar Marietta WineryMarietta Winery

Page 21: Ontologies and the Semantic Web II Deborah L. McGuinness Associate Director and Senior Research Scientist Knowledge Systems Laboratory Stanford University

August 9, 2004 21Deborah L. McGuinness

Class InheritanceClass Inheritance

Classes are organized into a subclass-superclass Classes are organized into a subclass-superclass (or generalization-specialization) hierarchy(or generalization-specialization) hierarchy

True subclass relationships are the basis of the True subclass relationships are the basis of the formal IS-A hierarchyformal IS-A hierarchy

Classes are “is-a” related if an instance of Classes are “is-a” related if an instance of the subclass is an instance of the the subclass is an instance of the superclasssuperclass

Classes may be viewed as setsClasses may be viewed as sets Subclasses of a class are comprised of a subset Subclasses of a class are comprised of a subset

of the supersetof the superset

Page 22: Ontologies and the Semantic Web II Deborah L. McGuinness Associate Director and Senior Research Scientist Knowledge Systems Laboratory Stanford University

August 9, 2004 22Deborah L. McGuinness

SubClass ExampleSubClass Example

RedWine is a subclass of WineRedWine is a subclass of WineEvery red wine is a wine or every instance of a Every red wine is a wine or every instance of a

red wine (like Marietta Old Vines Red) is an red wine (like Marietta Old Vines Red) is an instance of wineinstance of wine

NapaValleyWine is a subclass of NapaValleyWine is a subclass of CaliforniaWineCaliforniaWineEvery wine from napa valley is a wine from Every wine from napa valley is a wine from

californiacalifornia

Page 23: Ontologies and the Semantic Web II Deborah L. McGuinness Associate Director and Senior Research Scientist Knowledge Systems Laboratory Stanford University

August 9, 2004 23Deborah L. McGuinness

Levels in the Class Levels in the Class HierarchyHierarchy

Different modes of developmentDifferent modes of development top-down - define the most general top-down - define the most general

concepts first and then specialize themconcepts first and then specialize them bottom-up - define the most specific bottom-up - define the most specific

concepts and then organize them in concepts and then organize them in more general classesmore general classes

combination (typical – breadth at the combination (typical – breadth at the top level and depth along a few top level and depth along a few branches to test design)branches to test design)

Page 24: Ontologies and the Semantic Web II Deborah L. McGuinness Associate Director and Senior Research Scientist Knowledge Systems Laboratory Stanford University

August 9, 2004 24Deborah L. McGuinness

Wine Hierarchy (portion)Wine Hierarchy (portion)

Bottomlevel

Middlelevel

Toplevel

Taken from ontology development 101 from Protege

Page 25: Ontologies and the Semantic Web II Deborah L. McGuinness Associate Director and Senior Research Scientist Knowledge Systems Laboratory Stanford University

August 9, 2004 25Deborah L. McGuinness

Define Properties of Define Properties of ClassesClasses

Slots in a class definition describe Slots in a class definition describe attributes of members of a classattributes of members of a classeach wine will have color, sugar content, each wine will have color, sugar content,

flavor, body, etc.flavor, body, etc.

Page 26: Ontologies and the Semantic Web II Deborah L. McGuinness Associate Director and Senior Research Scientist Knowledge Systems Laboratory Stanford University

August 9, 2004 26Deborah L. McGuinness

SlotsSlots Types of propertiesTypes of properties

““intrinsic” properties: intrinsic” properties: flavorflavor and and colorcolor of wine of wine ““extrinsic” properties: extrinsic” properties: namename and and priceprice of wine of wine parts: parts: ingredientsingredients in a dish in a dish relations to other objects: relations to other objects: producerproducer of wine of wine

(winery)(winery) Data and object propertiesData and object properties

simple (datatype) contain primitive values simple (datatype) contain primitive values (strings, numbers)(strings, numbers)

complex properties: contain other objects complex properties: contain other objects (e.g., a winery instance)(e.g., a winery instance)

Page 27: Ontologies and the Semantic Web II Deborah L. McGuinness Associate Director and Senior Research Scientist Knowledge Systems Laboratory Stanford University

August 9, 2004 27Deborah L. McGuinness

Example Slots for the Example Slots for the class Wineclass Wine

Page 28: Ontologies and the Semantic Web II Deborah L. McGuinness Associate Director and Senior Research Scientist Knowledge Systems Laboratory Stanford University

August 9, 2004 28Deborah L. McGuinness

Slot and Class Slot and Class InheritanceInheritance

A subclass inherits all the slots from A subclass inherits all the slots from the superclassthe superclassIf a wine has a name and flavor, a red wine If a wine has a name and flavor, a red wine

also has a name and flavoralso has a name and flavor If a class has multiple super classes, it If a class has multiple super classes, it

inherits slots from all of theminherits slots from all of themPort is both a dessert wine and a red wine. Port is both a dessert wine and a red wine.

It inherits “sugar content: sweet” from It inherits “sugar content: sweet” from the dessert wine and “color:red” from red the dessert wine and “color:red” from red winewine

Page 29: Ontologies and the Semantic Web II Deborah L. McGuinness Associate Director and Senior Research Scientist Knowledge Systems Laboratory Stanford University

August 9, 2004 29Deborah L. McGuinness

Property ConstraintsProperty Constraints

Property constraints describe or Property constraints describe or limit the set of possible values for a limit the set of possible values for a slotslotthe name of a wine is a stringthe name of a wine is a string

the wine producer is an instance of the wine producer is an instance of WineryWinery

a winery has exactly one locationa winery has exactly one location

Page 30: Ontologies and the Semantic Web II Deborah L. McGuinness Associate Director and Senior Research Scientist Knowledge Systems Laboratory Stanford University

August 9, 2004 30Deborah L. McGuinness

Example: Wine Properties Example: Wine Properties and restrictionsand restrictions

Page 31: Ontologies and the Semantic Web II Deborah L. McGuinness Associate Director and Senior Research Scientist Knowledge Systems Laboratory Stanford University

August 9, 2004 31Deborah L. McGuinness

Slot restrictions: Slot restrictions: CardinalityCardinality

Slot cardinality – the number of values a slot Slot cardinality – the number of values a slot can or must havecan or must have CardinalityCardinality

Cardinality N means that the slot must have N valuesCardinality N means that the slot must have N values Minimum cardinalityMinimum cardinality

Minimum cardinality 1 means that the slot must have a Minimum cardinality 1 means that the slot must have a value (required)value (required)

Minimum cardinality 0 means that the slot value is Minimum cardinality 0 means that the slot value is optionaloptional

Maximum cardinalityMaximum cardinality Maximum cardinality 1 means that the slot can have at Maximum cardinality 1 means that the slot can have at

most one value (single-valued slot)most one value (single-valued slot) Maximum cardinality N means that the slot can have up to Maximum cardinality N means that the slot can have up to

N values. When N is greater than 1 it is a multiple-valued N values. When N is greater than 1 it is a multiple-valued slotslot

Page 32: Ontologies and the Semantic Web II Deborah L. McGuinness Associate Director and Senior Research Scientist Knowledge Systems Laboratory Stanford University

August 9, 2004 32Deborah L. McGuinness

Value TypeValue Type Slot value type – what values can the slot Slot value type – what values can the slot

havehave String: a string of characters (“ChString: a string of characters (“Chââteau Lafite”)teau Lafite”) Number: an integer or a float (15, 4.5)Number: an integer or a float (15, 4.5) Boolean: a true/false flagBoolean: a true/false flag Enumerated type: a list of allowed values (red, Enumerated type: a list of allowed values (red,

white, rose)white, rose) Filler: a single value. E.g., the color slot for a Filler: a single value. E.g., the color slot for a

redWine must be filled with the single value “red”redWine must be filled with the single value “red” Object type – a class defined in an ontology. E.g., Object type – a class defined in an ontology. E.g.,

Winery is the value restriction on the hasMaker Winery is the value restriction on the hasMaker slot on the class Wineslot on the class Wine

Page 33: Ontologies and the Semantic Web II Deborah L. McGuinness Associate Director and Senior Research Scientist Knowledge Systems Laboratory Stanford University

August 9, 2004 33Deborah L. McGuinness

Slot ExampleSlot Example

Page 34: Ontologies and the Semantic Web II Deborah L. McGuinness Associate Director and Senior Research Scientist Knowledge Systems Laboratory Stanford University

August 9, 2004 34Deborah L. McGuinness

Domain and Range of Domain and Range of SlotSlot

Domain of a slot – the class (or Domain of a slot – the class (or classes) that may have the slot e.g., classes) that may have the slot e.g., Wine is the domain of the slot Wine is the domain of the slot hasWineColorhasWineColor

Range of a slot – the class (or Range of a slot – the class (or classes) to which slot values belong classes) to which slot values belong e.g., everything that fills the e.g., everything that fills the hasWineColor slot is an instance of hasWineColor slot is an instance of the enumerated class the enumerated class {red,white,rose}{red,white,rose}

Page 35: Ontologies and the Semantic Web II Deborah L. McGuinness Associate Director and Senior Research Scientist Knowledge Systems Laboratory Stanford University

August 9, 2004 35Deborah L. McGuinness

Properties and Class Properties and Class InheritanceInheritance

A subclass A subclass inheritsinherits all the slots from the all the slots from the superclasssuperclass

A subclass can add constraints to A subclass can add constraints to “narrow” the list of allowed values“narrow” the list of allowed values Make the cardinality range smallerMake the cardinality range smaller Replace a class in the range with a subclassReplace a class in the range with a subclass

Wine

Frenchwine

Winery

Frenchwinery

is-a is-a

hasMaker

hasMaker

Page 36: Ontologies and the Semantic Web II Deborah L. McGuinness Associate Director and Senior Research Scientist Knowledge Systems Laboratory Stanford University

August 9, 2004 36Deborah L. McGuinness

Create InstancesCreate Instances

Create an instance of a classCreate an instance of a class The class becomes a The class becomes a parent of (or type of)parent of (or type of) the the

instanceinstance Any superclass of a class is an ancestor (or Any superclass of a class is an ancestor (or

type) of the instancetype) of the instance Assign slot values for the instance frameAssign slot values for the instance frame

Slot values should conform to the constraints Slot values should conform to the constraints such as range, value type, cardinality such as range, value type, cardinality restrictions, etc.restrictions, etc.

Page 37: Ontologies and the Semantic Web II Deborah L. McGuinness Associate Director and Senior Research Scientist Knowledge Systems Laboratory Stanford University

August 9, 2004 37Deborah L. McGuinness

Creating an Instance: Creating an Instance: ExampleExample

Page 38: Ontologies and the Semantic Web II Deborah L. McGuinness Associate Director and Senior Research Scientist Knowledge Systems Laboratory Stanford University

August 9, 2004 38Deborah L. McGuinness

Expanding the OntologyExpanding the Ontology

Breadth-orientedBreadth-oriented Identify all/most of the top level classes, Identify all/most of the top level classes,

properties needed at the top level, and properties needed at the top level, and constraints at the top level before deeperconstraints at the top level before deeper

Depth Oriented Depth Oriented Pick an important branch and go down it Pick an important branch and go down it

identifying specific subclasses, sub-sub-identifying specific subclasses, sub-sub-classes, etc. and the appropriate properties.classes, etc. and the appropriate properties.

Typical ontology design is a combination Typical ontology design is a combination of bothof both

Page 39: Ontologies and the Semantic Web II Deborah L. McGuinness Associate Director and Senior Research Scientist Knowledge Systems Laboratory Stanford University

August 9, 2004 39Deborah L. McGuinness

Defining Classes and a Class Defining Classes and a Class HierarchyHierarchy

Make sure the isa hierarchy is Make sure the isa hierarchy is formal –formal – i.e., is every instance of a subclass an i.e., is every instance of a subclass an

instance of the superclassinstance of the superclass There is no single best correct class There is no single best correct class

hierarchy but there are some rules hierarchy but there are some rules of thumbof thumb

Page 40: Ontologies and the Semantic Web II Deborah L. McGuinness Associate Director and Senior Research Scientist Knowledge Systems Laboratory Stanford University

August 9, 2004 40Deborah L. McGuinness

Class Hierarchy Class Hierarchy TransitivityTransitivity

The is-a The is-a relationship is relationship is transitivetransitive::B is a subclass of AB is a subclass of A

C is a subclass of BC is a subclass of B

C is a subclass of AC is a subclass of A

Page 41: Ontologies and the Semantic Web II Deborah L. McGuinness Associate Director and Senior Research Scientist Knowledge Systems Laboratory Stanford University

August 9, 2004 41Deborah L. McGuinness

Multiple InheritanceMultiple Inheritance

A class can have more than one A class can have more than one superclasssuperclass

The subclass inherits slots and The subclass inherits slots and restrictions from all the parentsrestrictions from all the parents

Different systems may resolve Different systems may resolve conflicts differentlyconflicts differently

Page 42: Ontologies and the Semantic Web II Deborah L. McGuinness Associate Director and Senior Research Scientist Knowledge Systems Laboratory Stanford University

August 9, 2004 42Deborah L. McGuinness

Avoiding Class CyclesAvoiding Class Cycles

Class cycles are rarely Class cycles are rarely desirabledesirable

Classes A, B, and C have Classes A, B, and C have equivalent sets of equivalent sets of instancesinstances By many definitions, A, B, By many definitions, A, B,

and C are thus equivalentand C are thus equivalent

Page 43: Ontologies and the Semantic Web II Deborah L. McGuinness Associate Director and Senior Research Scientist Knowledge Systems Laboratory Stanford University

August 9, 2004 43Deborah L. McGuinness

Disjoint ClassesDisjoint Classes Classes are Classes are disjointdisjoint if they cannot have common if they cannot have common

instancesinstances Disjoint classes cannot have any common Disjoint classes cannot have any common

subclassessubclasses either either E.g., if winery and wine are disjoint, then there E.g., if winery and wine are disjoint, then there

is no instance that is both a winery and a wine. is no instance that is both a winery and a wine. Similarly, there is no class that is both a Similarly, there is no class that is both a subclass of winery and simultaneously a subclass of winery and simultaneously a subclass of winesubclass of wine

Disjointness is often defined to help consistency Disjointness is often defined to help consistency checkingchecking

Page 44: Ontologies and the Semantic Web II Deborah L. McGuinness Associate Director and Senior Research Scientist Knowledge Systems Laboratory Stanford University

August 9, 2004 44Deborah L. McGuinness

Siblings in the Class Siblings in the Class HierarchyHierarchy All the siblings in the All the siblings in the

class hierarchy should class hierarchy should be at the same level of be at the same level of generalitygenerality

Compare to section and Compare to section and subsections in a booksubsections in a book

Page 45: Ontologies and the Semantic Web II Deborah L. McGuinness Associate Director and Senior Research Scientist Knowledge Systems Laboratory Stanford University

August 9, 2004 45Deborah L. McGuinness

Levels of hierarchyLevels of hierarchy

If a class has only one If a class has only one child, there may be a child, there may be a modeling problem. This is modeling problem. This is often a sign that the often a sign that the definition is incompletedefinition is incomplete

If the only Red Burgundy If the only Red Burgundy we have is Cwe have is Côôtes d’Or, why tes d’Or, why introduce the introduce the subhierarchy?subhierarchy?

Compare to bullets in a Compare to bullets in a bulleted listbulleted list

Page 46: Ontologies and the Semantic Web II Deborah L. McGuinness Associate Director and Senior Research Scientist Knowledge Systems Laboratory Stanford University

August 9, 2004 46Deborah L. McGuinness

Creating Levels and Creating Levels and SubclassesSubclasses

If a class has a large If a class has a large number of subclasses, it number of subclasses, it may be useful to define may be useful to define intermediate subclassesintermediate subclasses

E.g., in the domain of E.g., in the domain of wines, there are natural wines, there are natural groupings around wine groupings around wine colorcolor

However, if no natural However, if no natural classification exists, the classification exists, the long list may be more long list may be more naturalnatural

Page 47: Ontologies and the Semantic Web II Deborah L. McGuinness Associate Director and Senior Research Scientist Knowledge Systems Laboratory Stanford University

August 9, 2004 47Deborah L. McGuinness

Single and Plural Class Single and Plural Class NamesNames

A “wine” is not a A “wine” is not a kind-ofkind-of “wines”“wines”

A wine is an A wine is an instanceinstance of of the class Winesthe class Wines

Class names should be Class names should be eithereither all singularall singular all pluralall pluralMariettaOldVinesRed

Class

Instance

instance-of

Page 48: Ontologies and the Semantic Web II Deborah L. McGuinness Associate Director and Senior Research Scientist Knowledge Systems Laboratory Stanford University

August 9, 2004 48Deborah L. McGuinness

SynonymsSynonyms

Synonym namesSynonym names for the same for the same concept are not different classesconcept are not different classes

Many systems allow listing Many systems allow listing synonyms as part of the class synonyms as part of the class definitiondefinition

OWL allows defining necessary and OWL allows defining necessary and sufficiency condition definitions sufficiency condition definitions thereby allowing synonym thereby allowing synonym definitions to be “first class” termsdefinitions to be “first class” terms

Page 49: Ontologies and the Semantic Web II Deborah L. McGuinness Associate Director and Senior Research Scientist Knowledge Systems Laboratory Stanford University

August 9, 2004 49Deborah L. McGuinness

One example One example hierarchy hierarchy of winesof wines

Page 50: Ontologies and the Semantic Web II Deborah L. McGuinness Associate Director and Senior Research Scientist Knowledge Systems Laboratory Stanford University

August 9, 2004 50Deborah L. McGuinness

When to introduce a new When to introduce a new class?class?

Subclasses of a class usually haveSubclasses of a class usually have Additional Additional propertiesproperties Additional slot Additional slot restrictionsrestrictions Participate in different Participate in different relationshipsrelationships

Page 51: Ontologies and the Semantic Web II Deborah L. McGuinness Associate Director and Senior Research Scientist Knowledge Systems Laboratory Stanford University

August 9, 2004 51Deborah L. McGuinness

A new class or a property A new class or a property value?value?

Do concepts with different slot values become Do concepts with different slot values become restrictions for different slots?restrictions for different slots?

How important is the distinction for the domain?How important is the distinction for the domain? A class of an instance should not change oftenA class of an instance should not change often

Page 52: Ontologies and the Semantic Web II Deborah L. McGuinness Associate Director and Senior Research Scientist Knowledge Systems Laboratory Stanford University

August 9, 2004 52Deborah L. McGuinness

A Class Or An InstanceA Class Or An Instance

Individual instances are the most specific objects in an ontologyIndividual instances are the most specific objects in an ontology If concepts form a natural hierarchy, represent them as classesIf concepts form a natural hierarchy, represent them as classes If they will have instances below them, represent them as classesIf they will have instances below them, represent them as classes

Page 53: Ontologies and the Semantic Web II Deborah L. McGuinness Associate Director and Senior Research Scientist Knowledge Systems Laboratory Stanford University

Inverse SlotsInverse Slots hasMaker hasMaker

andand

hasProducerhasProducer

are are inverseinverse slotsslots

Page 54: Ontologies and the Semantic Web II Deborah L. McGuinness Associate Director and Senior Research Scientist Knowledge Systems Laboratory Stanford University

Inverse Slots (II)Inverse Slots (II)

Inverse slots contain Inverse slots contain redundant redundant informationinformation, but, but Allow acquisition of the information in either directionAllow acquisition of the information in either direction Enable additional verificationEnable additional verification Allow presentation of information in both directionsAllow presentation of information in both directions

The actual The actual implementationimplementation differs from differs from system to systemsystem to system Are both values stored?Are both values stored? When are the inverse values filled in?When are the inverse values filled in? What happens if we change the link to an inverse slot?What happens if we change the link to an inverse slot?

Page 55: Ontologies and the Semantic Web II Deborah L. McGuinness Associate Director and Senior Research Scientist Knowledge Systems Laboratory Stanford University

Limiting the ScopeLimiting the Scope

An ontology should not contain An ontology should not contain allall the possible information about the the possible information about the domaindomain No need to specialize or generalize No need to specialize or generalize

more than the application requiresmore than the application requires No need to include all possible No need to include all possible

properties of a classproperties of a class Only the most salient propertiesOnly the most salient properties Only the properties that the applications Only the properties that the applications

requirerequire

Page 56: Ontologies and the Semantic Web II Deborah L. McGuinness Associate Director and Senior Research Scientist Knowledge Systems Laboratory Stanford University

Limiting the Scope (II)Limiting the Scope (II) Ontology of wine, food, and their pairings Ontology of wine, food, and their pairings

probably will not include details not probably will not include details not related to wine and food pairings such as:related to wine and food pairings such as: Bottle size (half bottle, full bottle, magnum, …)Bottle size (half bottle, full bottle, magnum, …) Label colorLabel color Wine bottle color (green, amber, …)Wine bottle color (green, amber, …)

Page 57: Ontologies and the Semantic Web II Deborah L. McGuinness Associate Director and Senior Research Scientist Knowledge Systems Laboratory Stanford University

August 9, 2004 57Deborah L. McGuinness

Moving to InfrastructureMoving to Infrastructure

Page 58: Ontologies and the Semantic Web II Deborah L. McGuinness Associate Director and Senior Research Scientist Knowledge Systems Laboratory Stanford University

August 9, 2004 58Deborah L. McGuinness

Ontology SupportOntology Support

In order to use an ontology-based In order to use an ontology-based solution, you must have:solution, you must have:

A languageA language A way to encode information (editing A way to encode information (editing

environment)environment) A way to update (evolution environment)A way to update (evolution environment) A way to reason with the information A way to reason with the information

(reasoner)(reasoner) ……

Page 59: Ontologies and the Semantic Web II Deborah L. McGuinness Associate Director and Senior Research Scientist Knowledge Systems Laboratory Stanford University

August 9, 2004 59Deborah L. McGuinness

IssuesIssues Collaboration among distributed teamsCollaboration among distributed teams Interconnectivity with many systems/standardsInterconnectivity with many systems/standards Language compatibilityLanguage compatibility Analysis and diagnosisAnalysis and diagnosis ScaleScale VersioningVersioning SecuritySecurity Ease of useEase of use Diverse training levels / user supportDiverse training levels / user support Presentation stylePresentation style LifecycleLifecycle ExtensibilityExtensibility See Das, Wu, and McGuinness. ``Industrial Strength Ontology Management''. In Isabel See Das, Wu, and McGuinness. ``Industrial Strength Ontology Management''. In Isabel

Cruz, Stefan Decker, Jerome Euzenat, and Deborah L. McGuinness, eds. Cruz, Stefan Decker, Jerome Euzenat, and Deborah L. McGuinness, eds. The Emerging The Emerging

Semantic Web. Semantic Web. IOS Press, 2002.IOS Press, 2002.

Page 60: Ontologies and the Semantic Web II Deborah L. McGuinness Associate Director and Senior Research Scientist Knowledge Systems Laboratory Stanford University

August 9, 2004 60Deborah L. McGuinness

Ontology Tools SurveyOntology Tools Survey

Page 61: Ontologies and the Semantic Web II Deborah L. McGuinness Associate Director and Senior Research Scientist Knowledge Systems Laboratory Stanford University

August 9, 2004 61Deborah L. McGuinness

Editor SurveyEditor Survey

Page 62: Ontologies and the Semantic Web II Deborah L. McGuinness Associate Director and Senior Research Scientist Knowledge Systems Laboratory Stanford University

August 9, 2004 62Deborah L. McGuinness

SurveySurvey VersionVersion Release DateRelease Date SourceSource Modeling Modeling

Features/LimitationsFeatures/Limitations Base LanguageBase Language Web SupportWeb Support Import/Export FormatsImport/Export Formats Graph ViewGraph View Consistency ChecksConsistency Checks

Multi-User SupportMulti-User Support MergingMerging Lexical SupportLexical Support Information Information

ExtractionExtraction CommentsComments Info URLInfo URL ContactContact

Page 63: Ontologies and the Semantic Web II Deborah L. McGuinness Associate Director and Senior Research Scientist Knowledge Systems Laboratory Stanford University

August 9, 2004 63Deborah L. McGuinness

Tool EnhancementsTool EnhancementsFeatureFeature PercentPercentAbstraction for knowledge modelingAbstraction for knowledge modeling 18% 18% Visual/intuitive navigation of ontology Visual/intuitive navigation of ontology 13%13%Reasoning and problem solving facilities Reasoning and problem solving facilities 12%12%Ontology alignment and data resource integrationOntology alignment and data resource integration 12%12%Support of standard industry domain and core vocabulariesSupport of standard industry domain and core vocabularies

9%9%Natural language processingNatural language processing 7%7%Versioning controlVersioning control 7%7%Ontology language standardization Ontology language standardization 6%6%Built-ins (wizards) for best practice methods Built-ins (wizards) for best practice methods 6%6%

Information extraction facilities Information extraction facilities 4%4%Features to learn user's editing style and needs Features to learn user's editing style and needs 3%3%Collaborative development support Collaborative development support 1%1%Ontology support for contexts Ontology support for contexts 1%1%

Page 64: Ontologies and the Semantic Web II Deborah L. McGuinness Associate Director and Senior Research Scientist Knowledge Systems Laboratory Stanford University

August 9, 2004 64Deborah L. McGuinness

Tool CollectionsTool Collections

Page 65: Ontologies and the Semantic Web II Deborah L. McGuinness Associate Director and Senior Research Scientist Knowledge Systems Laboratory Stanford University

August 9, 2004 65Deborah L. McGuinness

A Few Example ToolsA Few Example Tools Sandpiper Sandpiper

www.sandsoft.comwww.sandsoft.com Network Inference - Network Inference -

www.networkinference.comwww.networkinference.com Chimaera Chimaera

-www.ksl.stanford.edu/software/chi-www.ksl.stanford.edu/software/chimaera/maera/

Inference Web Inference Web iw.stanford.eduiw.stanford.edu

ProtégéProtégéprotege.stanford.eduprotege.stanford.edu

Page 66: Ontologies and the Semantic Web II Deborah L. McGuinness Associate Director and Senior Research Scientist Knowledge Systems Laboratory Stanford University

August 9, 2004 66Deborah L. McGuinness

Sandpiper Knowledge Sandpiper Knowledge Engineering Engineering

framework (prototype)framework (prototype)

Question Answering Environment

Java Theorem Prover(JTP)

Hybrid Reasoning System

OWL-enabled Script(s) of Questions

OWL-QL

Query in OWL-QLAnswer in OWL-QL, including binding set with bound variables;Explanation (Proof) URI

Explanation URI

Authoring and Analysis Question Answering

Draft Ontologies, Metadata in OWL

Chimaera (KSL)/ Sandpiper Analysis Server

KSL Reusable Time OntologyUSGS GILS Metadata Management OntologyISO 11179-based Metadata Registry SupportUtility Ontologies (e.g., SI Units, US Customary Units, ISO 1000 and 31 compliant)OWL-S/SWSL based SWS SupportOther Domain-Specific Ontologies

UML Ontology Libraries

Validated, IntegratedOWL Ontologies and

Meta-knowledge

IBM Rational Rose® 2004

Visual Ontology Modeler™ (VOM)

Activity Log

Query & Results Logging

Activity Log Analysis

Web-based Explanations

Page 67: Ontologies and the Semantic Web II Deborah L. McGuinness Associate Director and Senior Research Scientist Knowledge Systems Laboratory Stanford University

August 9, 2004 67Deborah L. McGuinness

visual ontology modelervisual ontology modeler™ ™ (VOM) 1.x(VOM) 1.x

Page 68: Ontologies and the Semantic Web II Deborah L. McGuinness Associate Director and Senior Research Scientist Knowledge Systems Laboratory Stanford University

August 9, 2004 68Deborah L. McGuinness

Sandpiper Product PlansSandpiper Product PlansVisual Ontology ModelerVisual Ontology Modeler ™ (VOM v1.x)™ (VOM v1.x)

VOM v1.x customers and user community – ISX Corp, VOM v1.x customers and user community – ISX Corp, Fujitsu, Raytheon IIS, GE Global Research, IBM, MetLife, Fujitsu, Raytheon IIS, GE Global Research, IBM, MetLife, NIST, Stanford KSL, several smaller systems integratorsNIST, Stanford KSL, several smaller systems integrators

In use and/or under evaluation by many government In use and/or under evaluation by many government programs including: NIST, DARPA/XG, DARPA/DAML. EBO programs including: NIST, DARPA/XG, DARPA/DAML. EBO (Effects Based Operations), EU AgentCities programs, …(Effects Based Operations), EU AgentCities programs, …

Component-based ontology authoring in UML, an add-in to Component-based ontology authoring in UML, an add-in to IBM Rational Rose®; growing library of standards and IBM Rational Rose®; growing library of standards and utility ontologies, including ISO/IEC and US metadata utility ontologies, including ISO/IEC and US metadata standardsstandards

VOM v1.5: RDFS/OWL import/export, XMI import/export, VOM v1.5: RDFS/OWL import/export, XMI import/export, MOF-based integration with Adaptive repositoryMOF-based integration with Adaptive repository

Integrated Ontology Development Environment Integrated Ontology Development Environment

(planned)(planned) IBM Rational WSAD/Eclipse and Java based; integrated, IBM Rational WSAD/Eclipse and Java based; integrated,

DB2 deductive KB and ontology management & evolution DB2 deductive KB and ontology management & evolution environmentenvironment

Graphical analysis, alignment, merging and composition of Graphical analysis, alignment, merging and composition of ontology componentsontology components

OWL-S/SWSL support for semantic web servicesOWL-S/SWSL support for semantic web services Integrated, modular reasoning support for consistency Integrated, modular reasoning support for consistency

checking, alignment, composition; explanation and checking, alignment, composition; explanation and provenanceprovenance

Page 69: Ontologies and the Semantic Web II Deborah L. McGuinness Associate Director and Senior Research Scientist Knowledge Systems Laboratory Stanford University

August 9, 2004 69Deborah L. McGuinness

CONSTRUCTCONSTRUCTVisio-Based OWL ModelingVisio-Based OWL Modeling

Construct Visual OWL ModelingConstruct Visual OWL ModelingPlug-In to Microsoft Visio (supports Visio 2002 Plug-In to Microsoft Visio (supports Visio 2002 & 2003)& 2003)

Team-Based Ontology DevelopmentTeam-Based Ontology DevelopmentModularization of Large Ontologies for Distributed TeamsModularization of Large Ontologies for Distributed Teams

Consistency Checking for Linked OntologiesConsistency Checking for Linked Ontologies Change SynchronizationChange Synchronization

Standard Visio “tab” Based Palates SupportStandard Visio “tab” Based Palates Support Hierarchical or Thematic Linking of PagesHierarchical or Thematic Linking of Pages

Legacy System Integration (Ontology Mapping)Legacy System Integration (Ontology Mapping)Directly Map OWL to Legacy EnvironmentsDirectly Map OWL to Legacy Environments

Relational Sources and/or XML/Web Service WSDLsRelational Sources and/or XML/Web Service WSDLs

Standards Support (OWL export)Standards Support (OWL export)Automated, Real-Time Syntax Checking for All Linked Automated, Real-Time Syntax Checking for All Linked OntologyOntology

* All Rights Reserved by Network Inference Inc

Page 70: Ontologies and the Semantic Web II Deborah L. McGuinness Associate Director and Senior Research Scientist Knowledge Systems Laboratory Stanford University

August 9, 2004 70Deborah L. McGuinness

CONSTRUCTCONSTRUCTRapid Modeling, Visual EditingRapid Modeling, Visual Editing

Provides graphical and text environment for editing

Exports to OWL; Processed by Cerebra Server

* All Rights Reserved by Network Inference Inc

Page 71: Ontologies and the Semantic Web II Deborah L. McGuinness Associate Director and Senior Research Scientist Knowledge Systems Laboratory Stanford University

August 9, 2004 71Deborah L. McGuinness

CEREBRACEREBRAEnterprise Inferencing EngineEnterprise Inferencing Engine

Cerebra ServerCerebra ServerCommercial-grade Commercial-grade inference platforminference platform

Provides industry-standard Provides industry-standard query, deployment, and query, deployment, and management capabilitiesmanagement capabilitiesUser ManagementUser ManagementOntology ManagementOntology ManagementQuery ManagementQuery ManagementAccess ManagementAccess Management

Emphasis on scalability, Emphasis on scalability, load balancing, and load balancing, and robustnessrobustness

Standards based (OWL)Standards based (OWL)

* All Rights Reserved by Network Inference Inc

Page 72: Ontologies and the Semantic Web II Deborah L. McGuinness Associate Director and Senior Research Scientist Knowledge Systems Laboratory Stanford University

August 9, 2004 72Deborah L. McGuinness

Intensive Data Grids

Sensors & Live Data Feeds

Business Applications

Meta-Access Tier(lower)

Core Service Tier(upper) Cerebra OWL/RDF Server Search

Meta-Data Layer

Security

HeterogeneousData

TECHNOLOGY STACKTECHNOLOGY STACKWhere Network Inference Fits InWhere Network Inference Fits In

Data Warehouses “Webbed” Information

Enterprise Data (SAP, Siebel, etc)

SAP ORACLE BI TOOLS

Implement solutions today that add value above current approaches & capture a piece of the infrastructure

* All Rights Reserved by Network Inference Inc

Page 73: Ontologies and the Semantic Web II Deborah L. McGuinness Associate Director and Senior Research Scientist Knowledge Systems Laboratory Stanford University

August 9, 2004 73Deborah L. McGuinness

Example Inference Engine Example Inference Engine PointersPointers

Cerebra – Network Inference’s EngineCerebra – Network Inference’s Engine Jena – HP’s Semantic Web PlatformJena – HP’s Semantic Web Platformhttp://www.hpl.hp.com/semweb/jena.htmhttp://www.hpl.hp.com/semweb/jena.htm A few university options: A few university options:

FACT (Manchester)FACT (Manchester) Racer (http://www.sts.tu-harburg.de/~r.f.moeller/racer/ ) Racer (http://www.sts.tu-harburg.de/~r.f.moeller/racer/ )

suggested DIG (DL Interface) reasoner for Protégé suggested DIG (DL Interface) reasoner for Protégé http://protege.stanford.edu/plugins/owl/ http://protege.stanford.edu/plugins/owl/

JTP - Hybrid Reasoner – FOL plus special purpose reasonerJTP - Hybrid Reasoner – FOL plus special purpose reasoner SemWebCentralSemWebCentral

Largely open source collection including emerging toolsLargely open source collection including emerging tools With an evolving workflows section helping guide choice With an evolving workflows section helping guide choice

http://semwebcentral.org/index.jsp?page=workflows http://semwebcentral.org/index.jsp?page=workflows SWeDe – SW Development Environment – plugins SWeDe – SW Development Environment – plugins

for Eclipse (BBN) http://owl-for Eclipse (BBN) http://owl-eclipse.projects.semwebcentral.org/ eclipse.projects.semwebcentral.org/

Page 74: Ontologies and the Semantic Web II Deborah L. McGuinness Associate Director and Senior Research Scientist Knowledge Systems Laboratory Stanford University

August 9, 2004 74Deborah L. McGuinness

Evolution EnvironmentsEvolution Environments

Multiple Ontology Support Multiple Ontology Support MergingMerging Knowledge Base review and suggestionsKnowledge Base review and suggestions

DiagnosticsDiagnostics Provable problems, possible problems, …Provable problems, possible problems, …

Updates, Source Code Control, Updates, Source Code Control, Versioning,Versioning,

Collaborative ComponentsCollaborative Components

Page 75: Ontologies and the Semantic Web II Deborah L. McGuinness Associate Director and Senior Research Scientist Knowledge Systems Laboratory Stanford University

August 9, 2004 75Deborah L. McGuinness

Merging (Chimaera)Merging (Chimaera)

Page 76: Ontologies and the Semantic Web II Deborah L. McGuinness Associate Director and Senior Research Scientist Knowledge Systems Laboratory Stanford University

August 9, 2004 76Deborah L. McGuinness

StrategiesStrategies Lexical AnalysisLexical Analysis

Term name similarityTerm name similarity Term containment (sportsCar and car)Term containment (sportsCar and car)

Semantic AnalysisSemantic Analysis Structural similarity – the same parents, the Structural similarity – the same parents, the

same slots with the same value restrictions, ….same slots with the same value restrictions, …. Meta information exploitationMeta information exploitation

Expanded merging issuesExpanded merging issues If multiple ontologies were used and one If multiple ontologies were used and one

contained a disjoint decomposition and the contained a disjoint decomposition and the contained those terms along with others at the contained those terms along with others at the same level, they may be candidates for the same level, they may be candidates for the merged disjoint decompositionmerged disjoint decomposition

Page 77: Ontologies and the Semantic Web II Deborah L. McGuinness Associate Director and Senior Research Scientist Knowledge Systems Laboratory Stanford University
Page 78: Ontologies and the Semantic Web II Deborah L. McGuinness Associate Director and Senior Research Scientist Knowledge Systems Laboratory Stanford University

August 9, 2004 78Deborah L. McGuinness

Inference Web Inference Web **Framework for Framework for explainingexplaining question answering tasks by question answering tasks by

abstracting, storing, exchanging, combining, annotating, abstracting, storing, exchanging, combining, annotating, filtering, segmenting, comparing, and rendering proofs filtering, segmenting, comparing, and rendering proofs and proof fragments provided by question answerersand proof fragments provided by question answerers

IW’s Proof Markup Language (PML)IW’s Proof Markup Language (PML) is an is an interlingua for proof interchangeinterlingua for proof interchange

IWBaseIWBase is a distributed repository of meta- is a distributed repository of meta-information related to proofs and their explanationsinformation related to proofs and their explanations

IW BrowserIW Browser is an IW tool for displaying PML is an IW tool for displaying PML documents containing proofs and explanations documents containing proofs and explanations (possibly from multiple inference engines)(possibly from multiple inference engines)

IW ExplainerIW Explainer is an IW tool for abstracting proofs is an IW tool for abstracting proofs into more understandable formatsinto more understandable formats

*Work with Pinheiro da Silva*Work with Pinheiro da Silva

Page 79: Ontologies and the Semantic Web II Deborah L. McGuinness Associate Director and Senior Research Scientist Knowledge Systems Laboratory Stanford University

August 9, 2004 79Deborah L. McGuinness

Registry InformationRegistry Information

IWBase has core and domain-specific repositories of IWBase has core and domain-specific repositories of meta-data useful for disclosing knowledge meta-data useful for disclosing knowledge provenance and reasoning information such as provenance and reasoning information such as descriptions of descriptions of Question answering systems (Inference Engines, Question answering systems (Inference Engines,

Extractors, …) along with their supported Extractors, …) along with their supported inference rulesinference rules

Information sources such as organizations, Information sources such as organizations, publications and ontologiespublications and ontologies

Representation languages along with their Representation languages along with their axiomsaxioms

Page 80: Ontologies and the Semantic Web II Deborah L. McGuinness Associate Director and Senior Research Scientist Knowledge Systems Laboratory Stanford University

August 9, 2004 80Deborah L. McGuinness

Page 81: Ontologies and the Semantic Web II Deborah L. McGuinness Associate Director and Senior Research Scientist Knowledge Systems Laboratory Stanford University

August 9, 2004 81Deborah L. McGuinness

Page 82: Ontologies and the Semantic Web II Deborah L. McGuinness Associate Director and Senior Research Scientist Knowledge Systems Laboratory Stanford University

August 9, 2004 82Deborah L. McGuinness

ExplainerExplainer

Presents Presents QueryQuery AnswerAnswer Abstraction of Justification (PML information)Abstraction of Justification (PML information) Limited Meta InformationLimited Meta Information Suggests Drill down options (also provides Suggests Drill down options (also provides

feedback options)feedback options)

Page 83: Ontologies and the Semantic Web II Deborah L. McGuinness Associate Director and Senior Research Scientist Knowledge Systems Laboratory Stanford University

August 9, 2004 83Deborah L. McGuinness

UIMA ExplanationUIMA Explanation

Page 84: Ontologies and the Semantic Web II Deborah L. McGuinness Associate Director and Senior Research Scientist Knowledge Systems Laboratory Stanford University

August 9, 2004 84Deborah L. McGuinness

Follow-up : MetadataFollow-up : Metadata

Page 85: Ontologies and the Semantic Web II Deborah L. McGuinness Associate Director and Senior Research Scientist Knowledge Systems Laboratory Stanford University

August 9, 2004 85Deborah L. McGuinness

Follow-up: AssumptionsFollow-up: Assumptions

Page 86: Ontologies and the Semantic Web II Deborah L. McGuinness Associate Director and Senior Research Scientist Knowledge Systems Laboratory Stanford University

August 9, 2004 86Deborah L. McGuinness

Follow-up: PML AbstractionFollow-up: PML Abstraction(Techies only)(Techies only)

Page 87: Ontologies and the Semantic Web II Deborah L. McGuinness Associate Director and Senior Research Scientist Knowledge Systems Laboratory Stanford University

August 9, 2004 87Deborah L. McGuinness

Observations on Explaining Observations on Explaining Extracted Entities (Techies)Extracted Entities (Techies)

Sentences in English

Sentences in annotated English

Sentences in logical format, i.e., KIF

Page 88: Ontologies and the Semantic Web II Deborah L. McGuinness Associate Director and Senior Research Scientist Knowledge Systems Laboratory Stanford University

August 9, 2004 88Deborah L. McGuinness

Further Observations on Further Observations on Explaining Extracted Explaining Extracted

EntitiesEntitiesSource: fbi_01.txt

Source Usage: span from 01 to 78

This extractor decided that Person_fbi-01.txt_46

is a Person and not Occupation

Same conclusion from multiple extractors

conflicting conclusion from one extractor

Page 89: Ontologies and the Semantic Web II Deborah L. McGuinness Associate Director and Senior Research Scientist Knowledge Systems Laboratory Stanford University

August 9, 2004 89Deborah L. McGuinness

StatusStatus

Inference Web infrastructure (PML, browser, explainer, Inference Web infrastructure (PML, browser, explainer, registry) being used in government programs such as PAL registry) being used in government programs such as PAL and NIMD, commercial research labs – IBM, and NIMD, commercial research labs – IBM, BBoeing, oeing, SRI, Universities – USC, U MD, …SRI, Universities – USC, U MD, …

Integration and registration process underway with Integration and registration process underway with extraction communityextraction community

Useful now for helping decide if information is Useful now for helping decide if information is trustworthy, comes from authoritative sources, consistent, trustworthy, comes from authoritative sources, consistent, reliablereliable

Benefits from more meta data and more information Benefits from more meta data and more information population but is useful in an population but is useful in an incremental natureincremental nature

Page 90: Ontologies and the Semantic Web II Deborah L. McGuinness Associate Director and Senior Research Scientist Knowledge Systems Laboratory Stanford University

August 9, 2004 90Deborah L. McGuinness

OWL Lite Features OWL Lite Features RDF Schema FeaturesRDF Schema Features

Class, rdfs:subClassOf , Individual Class, rdfs:subClassOf , Individual rdf:Property, rdfs:subPropertyOf rdf:Property, rdfs:subPropertyOf rdfs:domain , rdfs:range rdfs:domain , rdfs:range

Equality and InequalityEquality and Inequality sameClassAs , samePropertyAs , sameIndividualAs sameClassAs , samePropertyAs , sameIndividualAs differentIndividualFrom differentIndividualFrom

Restricted CardinalityRestricted Cardinality minCardinality, maxCardinality (restricted to 0 or 1) minCardinality, maxCardinality (restricted to 0 or 1) cardinality (restricted to 0 or 1) cardinality (restricted to 0 or 1)

Property CharacteristicsProperty Characteristics inverseOf , TransitiveProperty , SymmetricProperty inverseOf , TransitiveProperty , SymmetricProperty FunctionalProperty(unique) , InverseFunctionalPropertyFunctionalProperty(unique) , InverseFunctionalProperty allValuesFrom, someValuesFrom (universal and existential allValuesFrom, someValuesFrom (universal and existential

local range restrictions)local range restrictions) DatatypesDatatypes

Following the decisions of RDF Core. Following the decisions of RDF Core. Header Information Header Information

imports , Dublin Core Metadata , versionInfoimports , Dublin Core Metadata , versionInfo

Page 91: Ontologies and the Semantic Web II Deborah L. McGuinness Associate Director and Senior Research Scientist Knowledge Systems Laboratory Stanford University

August 9, 2004 91Deborah L. McGuinness

OWL FeaturesOWL Features Class AxiomsClass Axioms

oneOfoneOf (enumerated classes) (enumerated classes) disjointWithdisjointWith sameClassAssameClassAs applied to class expressions applied to class expressions rdfs:subClassOfrdfs:subClassOf applied to class expressions applied to class expressions

Boolean Combinations of Class Expressions Boolean Combinations of Class Expressions unionOfunionOf intersectionOfintersectionOf complementOfcomplementOf

Arbitrary Cardinality Arbitrary Cardinality minCardinalityminCardinality maxCardinalitymaxCardinality cardinalitycardinality

Filler InformationFiller Information hasValuehasValue Descriptions can include specific value Descriptions can include specific value

information information

Page 92: Ontologies and the Semantic Web II Deborah L. McGuinness Associate Director and Senior Research Scientist Knowledge Systems Laboratory Stanford University

August 9, 2004 92Deborah L. McGuinness

OWL Lite and OWLOWL Lite and OWL

Feature Synopsis:Feature Synopsis:

http://www.w3.org/TR/owl-features/http://www.w3.org/TR/owl-features/ Guide:Guide:

http://www.w3.org/TR/owl-features/http://www.w3.org/TR/owl-features/ Reference Description:Reference Description:

http://www.w3.org/TR/owl-ref/http://www.w3.org/TR/owl-ref/

Page 93: Ontologies and the Semantic Web II Deborah L. McGuinness Associate Director and Senior Research Scientist Knowledge Systems Laboratory Stanford University

August 9, 2004 93Deborah L. McGuinness

ValidatorsValidators

For RDF: For RDF: http://www.w3.org/RDF/Validator/ http://www.w3.org/RDF/Validator/

For OWL: For OWL: http://owl.bbn.com/validator/ http://owl.bbn.com/validator/ http://phoebus.cs.man.ac.uk:9999/http://phoebus.cs.man.ac.uk:9999/

OWL/Validator OWL/Validator http://www.mindswap.org/2003/pellet/http://www.mindswap.org/2003/pellet/

demo.shtmldemo.shtml

Page 94: Ontologies and the Semantic Web II Deborah L. McGuinness Associate Director and Senior Research Scientist Knowledge Systems Laboratory Stanford University

August 9, 2004 94Deborah L. McGuinness

A few direction setting A few direction setting programsprograms

DARPA Personal Assistant that Learns (PAL)DARPA Personal Assistant that Learns (PAL) Enable computer systems that can reason, learn from Enable computer systems that can reason, learn from

experience, be told what to do, explain what they are experience, be told what to do, explain what they are doing, reflect on their experience, and respond robustly doing, reflect on their experience, and respond robustly to surprise. to surprise.

DARPA Rapid Knowledge Formation (RKF)DARPA Rapid Knowledge Formation (RKF) Goal: allow distributed teams of subject matter experts Goal: allow distributed teams of subject matter experts

to quickly and easily build, maintain, and use knowledge to quickly and easily build, maintain, and use knowledge bases without need for specialized training. bases without need for specialized training.

DARPA High Performance Knowledge Base (HPKB)DARPA High Performance Knowledge Base (HPKB) Goal: advance the technology of how computers acquire, Goal: advance the technology of how computers acquire,

represent and manipulate knowledgerepresent and manipulate knowledgeARDA’s Novel Intelligence for Massive Data (NIMD)ARDA’s Novel Intelligence for Massive Data (NIMD)

Goal – Avoid strategic surprise by helping analysts be Goal – Avoid strategic surprise by helping analysts be more effective (focus attention on critical information more effective (focus attention on critical information and help analyze/prune/refine/explain/reuse/…)and help analyze/prune/refine/explain/reuse/…)

ARDA’s Advanced Question & Answering for Intelligence ARDA’s Advanced Question & Answering for Intelligence (AQUAINT)(AQUAINT) Goal – Advance QA against structured and unstructured Goal – Advance QA against structured and unstructured

infoinfoConsulting including search, ecommerce, configuration, …Consulting including search, ecommerce, configuration, …

Page 95: Ontologies and the Semantic Web II Deborah L. McGuinness Associate Director and Senior Research Scientist Knowledge Systems Laboratory Stanford University

August 9, 2004 95Deborah L. McGuinness

Many Program BAAs Using Many Program BAAs Using ResultsResults

NGA BAANGA BAA DARPA XG RadioDARPA XG Radio DARPA CoordinatorsDARPA Coordinators DARPA advanced soldier sensor DARPA advanced soldier sensor

information system and technology information system and technology (ASSIST)(ASSIST)

DARPA Situation Aware Protocols in DARPA Situation Aware Protocols in Edge Network Technologies (SAPIENT)Edge Network Technologies (SAPIENT)

……..

Page 96: Ontologies and the Semantic Web II Deborah L. McGuinness Associate Director and Senior Research Scientist Knowledge Systems Laboratory Stanford University

August 9, 2004 96Deborah L. McGuinness

DiscussionDiscussion Choose a language (maybe OWL)Choose a language (maybe OWL) Find an editing environment (text editor, Find an editing environment (text editor,

protégé, construct, vom, …)protégé, construct, vom, …) Generate some markup Generate some markup ValidateValidate Use it for…Use it for…

Search (tap, findUR, …)Search (tap, findUR, …) Consistency checkingConsistency checking Policy checking, enforcement, classification,Policy checking, enforcement, classification, ConfigurationConfiguration Analysis of all types (including NIMD style)Analysis of all types (including NIMD style) AssistantAssistant InteroperabilityInteroperability ……..

Page 97: Ontologies and the Semantic Web II Deborah L. McGuinness Associate Director and Senior Research Scientist Knowledge Systems Laboratory Stanford University

August 9, 2004 97Deborah L. McGuinness

PointersPointersSelected Papers:- McGuinness. Ontologies come of age, 2003- Das, Wei, McGuinness, Industrial Strength Ontology Evolution Environments, 2002.- Kendall, Dutra, McGuinness. Towards a Commercial Strength Ontology Development Environment, 2002.- McGuinness Description Logics Emerge from Ivory Towers, 2001.- McGuinness. Ontologies and Online Commerce, 2001.- McGuinness. Conceptual Modeling for Distributed Ontology Environments, 2000.- McGuinness, Fikes, Rice, Wilder. An Environment for Merging and Testing Large Ontologies, 2000.- Brachman, Borgida, McGuinness, Patel-Schneider. Knowledge Representation meets Reality, 1999.- McGuinness. Ontological Issues for Knowledge-Enhanced Search, 1998.- McGuinness and Wright. Conceptual Modeling for Configuration, 1998.

Selected Tutorials:-Smith, Welty, McGuinness. OWL Web Ontology Language Guide, 2003.-Noy, McGuinness. Ontology Development 101: A Guide to Creating your First Ontology. 2001.- Brachman, McGuinness, Resnick, Borgida. How and When to Use a KL-ONE-like System, 1991.

Languages, Environments, Software:- OWL - http://www.w3.org/TR/owl-features/ , http://www.w3.org/TR/owl-guide/- DAML+OIL: http://www.daml.org/- Inference Web - http://www.ksl.stanford.edu/software/iw/ - Chimaera - http://www.ksl.stanford.edu/software/chimaera/ - FindUR - http://www.research.att.com/people/~dlm/findur/ - TAP – http://tap.stanford.edu/- DQL - http://www.ksl.stanford.edu/projects/dql/

Page 98: Ontologies and the Semantic Web II Deborah L. McGuinness Associate Director and Senior Research Scientist Knowledge Systems Laboratory Stanford University

August 9, 2004 98Deborah L. McGuinness

ExtrasExtras

Page 99: Ontologies and the Semantic Web II Deborah L. McGuinness Associate Director and Senior Research Scientist Knowledge Systems Laboratory Stanford University

August 9, 2004 99Deborah L. McGuinness

General Nature of General Nature of DescriptionsDescriptions

a WINE

a LIQUIDa POTABLE

grape: chardonnay, ... [>= 1]sugar-content: dry, sweet, off-drycolor: red, white, roseprice: a PRICEwinery: a WINERY

grape dictates color (modulo skin)harvest time and sugar are related

general categories

structured components

interconnectionsbetween parts

Page 100: Ontologies and the Semantic Web II Deborah L. McGuinness Associate Director and Senior Research Scientist Knowledge Systems Laboratory Stanford University

August 9, 2004 100Deborah L. McGuinness

General Nature of General Nature of DescriptionsDescriptions

a WINE

a LIQUIDa POTABLE

grape: chardonnay, ... [>= 1]sugar-content: dry, sweet, off-drycolor: red, white, roseprice: a PRICEwinery: a WINERY

grape dictates color (modulo skin)harvest time and sugar are related

general categories

structured components

interconnectionsbetween parts

number/card restrictions

valuerestrictions

class

superclass

Roles/properties

Page 101: Ontologies and the Semantic Web II Deborah L. McGuinness Associate Director and Senior Research Scientist Knowledge Systems Laboratory Stanford University

August 9, 2004 101Deborah L. McGuinness

Some uses of OntologiesSome uses of OntologiesSimple ontologies (taxonomies) provide:Simple ontologies (taxonomies) provide: Controlled shared vocabulary (search engines, Controlled shared vocabulary (search engines,

authors, users, databases, programs/agents all speak authors, users, databases, programs/agents all speak same language)same language)

Site Organization and Navigation SupportSite Organization and Navigation Support Expectation setting (left side of many web pages)Expectation setting (left side of many web pages) ““Umbrella” Upper Level Structures (for extension)Umbrella” Upper Level Structures (for extension) Browsing support (tagged structures such as Browsing support (tagged structures such as

Yahoo!)Yahoo!) Search support (query expansion approaches such Search support (query expansion approaches such

as FindUR, e-Cyc)as FindUR, e-Cyc) Sense disambiguationSense disambiguation

Page 102: Ontologies and the Semantic Web II Deborah L. McGuinness Associate Director and Senior Research Scientist Knowledge Systems Laboratory Stanford University

August 9, 2004 102Deborah L. McGuinness

KSL Wine AgentKSL Wine AgentSemantic Web IntegrationSemantic Web Integration

Wine Agent receives a meal description and retrieves a selection of matching wines available on the Web, using an ensemble of emerging standards and tools:

• DAML+OIL / OWL for representing a domain ontology of foods, wines, their properties, and relationships between them• JTP theorem prover for deriving appropriate pairings• DQL for querying a knowledge base consisting of the above• Inference Web for explaining and validating the response• [Web Services for interfacing with vendors]• Utilities for conducting and caching the above transactions

Page 103: Ontologies and the Semantic Web II Deborah L. McGuinness Associate Director and Senior Research Scientist Knowledge Systems Laboratory Stanford University

August 9, 2004 103Deborah L. McGuinness

Page 104: Ontologies and the Semantic Web II Deborah L. McGuinness Associate Director and Senior Research Scientist Knowledge Systems Laboratory Stanford University

August 9, 2004 104Deborah L. McGuinness

<rdfs:Class rdf:ID="BLAND-FISH-COURSE"> <daml:intersectionOf rdf:parseType="daml:collection"> <rdfs:Class rdf:about="#MEAL-COURSE"/> <daml:Restriction> <daml:onProperty rdf:resource="#FOOD"/> <daml:toClass rdf:resource="#BLAND-FISH"/> </daml:Restriction> </daml:intersectionOf> <rdfs:subClassOf rdf:resource="#DRINK-HAS-DELICATE-FLAVOR-

RESTRICTION"/> </rdfs:Class> <rdfs:Class rdf:ID="BLAND-FISH"> <rdfs:subClassOf rdf:resource="#FISH"/> <daml:disjointWith rdf:resource="#NON-BLAND-FISH"/> </rdfs:Class> <rdf:Description rdf:ID="FLOUNDER"> <rdf:type rdf:resource="#BLAND-FISH"/> </rdf:Description> <rdfs:Class rdf:ID="CHARDONNAY"> <rdfs:subClassOf rdf:resource="#WHITE-COLOR-RESTRICTION"/> <rdfs:subClassOf rdf:resource="#MEDIUM-OR-FULL-BODY-

RESTRICTION"/> <rdfs:subClassOf rdf:resource="#MODERATE-OR-STRONG-FLAVOR-

RESTRICTION"/> […] </rdfs:Class> <rdf:Description rdf:ID="BANCROFT-CHARDONNAY"> <rdf:type rdf:resource="#CHARDONNAY"/> <REGION rdf:resource="#NAPA"/> <MAKER rdf:resource="#BANCROFT"/> <SUGAR rdf:resource="#DRY"/> […] </rdf:Description>

Page 105: Ontologies and the Semantic Web II Deborah L. McGuinness Associate Director and Senior Research Scientist Knowledge Systems Laboratory Stanford University

August 9, 2004 105Deborah L. McGuinness

ProcessingProcessing Given a description of a meal,Given a description of a meal,

Use DQL to state a premise (the meal) and query the Use DQL to state a premise (the meal) and query the knowledge base for a suggestion for a wine knowledge base for a suggestion for a wine description or set of instancesdescription or set of instances

Use JTP to deduce answers (and proofs)Use JTP to deduce answers (and proofs) Use Inference Web to explain results (descriptions, Use Inference Web to explain results (descriptions,

instances, provenance, reasoning engines, etc.)instances, provenance, reasoning engines, etc.) Access relevant web sites (wine.com, …) to access Access relevant web sites (wine.com, …) to access

current informationcurrent information Use DAML-S for markup and protocol*Use DAML-S for markup and protocol*

http://www.ksl.stanford.edu/projects/wine/explanation.html

Page 106: Ontologies and the Semantic Web II Deborah L. McGuinness Associate Director and Senior Research Scientist Knowledge Systems Laboratory Stanford University

August 9, 2004 106Deborah L. McGuinness

Page 107: Ontologies and the Semantic Web II Deborah L. McGuinness Associate Director and Senior Research Scientist Knowledge Systems Laboratory Stanford University

August 9, 2004 107Deborah L. McGuinness

Page 108: Ontologies and the Semantic Web II Deborah L. McGuinness Associate Director and Senior Research Scientist Knowledge Systems Laboratory Stanford University

August 9, 2004 108Deborah L. McGuinness

Querying multiple online Querying multiple online sourcessources

Page 109: Ontologies and the Semantic Web II Deborah L. McGuinness Associate Director and Senior Research Scientist Knowledge Systems Laboratory Stanford University

August 9, 2004 109Deborah L. McGuinness

A Few Observations about A Few Observations about OntologiesOntologies Simple ontologies can be built by non-expertsSimple ontologies can be built by non-experts

Verity’s Topic Editor, Collaborative Topic Builder, GFP, Chimaera, Protégé, OIL-Verity’s Topic Editor, Collaborative Topic Builder, GFP, Chimaera, Protégé, OIL-ED, etc.ED, etc.

Ontologies can be semi-automatically generatedOntologies can be semi-automatically generated from crawls of site such as yahoo!, amazon, excite, etc.from crawls of site such as yahoo!, amazon, excite, etc. Semi-structured sites can provide starting pointsSemi-structured sites can provide starting points

Ontologies are exploding (business pull instead of technology push)Ontologies are exploding (business pull instead of technology push) e-commerce - MySimon, Amazon, Yahoo! Shopping, VerticalNet, …e-commerce - MySimon, Amazon, Yahoo! Shopping, VerticalNet, … Controlled vocabularies (for the web) abound - SIC codes, UMLS, UNSPSC, Controlled vocabularies (for the web) abound - SIC codes, UMLS, UNSPSC,

Open Directory (DMOZ), Rosetta Net, SUMOOpen Directory (DMOZ), Rosetta Net, SUMO Business interest expanding – ontology directors, business ontologies are Business interest expanding – ontology directors, business ontologies are

becoming more complicated (roles, value restrictions, …), VC firms interested,becoming more complicated (roles, value restrictions, …), VC firms interested, Markup Languages growing XML, RDF, DAML, RuleML, xxMLMarkup Languages growing XML, RDF, DAML, RuleML, xxML ““Real” ontologies are becoming more central to applicationsReal” ontologies are becoming more central to applications Search companies moving towards them – Yahoo, recently GoogleSearch companies moving towards them – Yahoo, recently Google

Page 110: Ontologies and the Semantic Web II Deborah L. McGuinness Associate Director and Senior Research Scientist Knowledge Systems Laboratory Stanford University

August 9, 2004 110Deborah L. McGuinness

Page 111: Ontologies and the Semantic Web II Deborah L. McGuinness Associate Director and Senior Research Scientist Knowledge Systems Laboratory Stanford University

August 9, 2004 111Deborah L. McGuinness

Page 112: Ontologies and the Semantic Web II Deborah L. McGuinness Associate Director and Senior Research Scientist Knowledge Systems Laboratory Stanford University

August 9, 2004 112Deborah L. McGuinness

Implications and Needs Implications and Needs for Ontology-enhanced for Ontology-enhanced

applicationsapplications Ontology Language Syntax and Semantics Ontology Language Syntax and Semantics

(DAML+OIL, OWL)(DAML+OIL, OWL) Upper Level/Core ontologies for reuse (Cyc, SUMO, Upper Level/Core ontologies for reuse (Cyc, SUMO,

CNS coalition, DAML-S…)CNS coalition, DAML-S…) Environments for Creation of Ontologies (Protégé, Environments for Creation of Ontologies (Protégé,

Sandpiper, Construct, OilEd, …)Sandpiper, Construct, OilEd, …) Environments for Maintenance of Ontologies Environments for Maintenance of Ontologies

(Chimaera, OntoBuilder, …)(Chimaera, OntoBuilder, …) Reasoning Environments (Cerebra, Fact, JTP, Snark, Reasoning Environments (Cerebra, Fact, JTP, Snark,

…)…) Training (Conceptual Modeling, reasoning usage, Training (Conceptual Modeling, reasoning usage,

tutorials – OWL Guide, Ontologies 101, OWL Tutorial, tutorials – OWL Guide, Ontologies 101, OWL Tutorial, …)…)

Page 113: Ontologies and the Semantic Web II Deborah L. McGuinness Associate Director and Senior Research Scientist Knowledge Systems Laboratory Stanford University

August 9, 2004 113Deborah L. McGuinness

DAML/OWL DAML/OWL Language Language

Web Languages

RDF/SXML

DAML-ONT

Formal FoundationsDescription Logics

FACT, CLASSIC, DLP, …

Frame Systems

DAML+OILOWL

OIL

•Extends vocabulary of XML and RDF/S•Rich ontology representation language•Language features chosen for efficient implementations

Page 114: Ontologies and the Semantic Web II Deborah L. McGuinness Associate Director and Senior Research Scientist Knowledge Systems Laboratory Stanford University

August 9, 2004 114Deborah L. McGuinness

IssuesIssues Collaboration among distributed teamsCollaboration among distributed teams Interconnectivity with many systems/standardsInterconnectivity with many systems/standards Analysis and diagnosisAnalysis and diagnosis ScaleScale VersioningVersioning SecuritySecurity Ease of useEase of use Diverse training levels / user supportDiverse training levels / user support Presentation stylePresentation style LifecycleLifecycle ExtensibilityExtensibility

Page 115: Ontologies and the Semantic Web II Deborah L. McGuinness Associate Director and Senior Research Scientist Knowledge Systems Laboratory Stanford University

August 9, 2004 115Deborah L. McGuinness

PointersPointersSelected Papers:- McGuinness. Ontologies come of age, 2003- Das, Wei, McGuinness, Industrial Strength Ontology Evolution Environments, 2002.- Kendall, Dutra, McGuinness. Towards a Commercial Strength Ontology Development Environment, 2002.- McGuinness Description Logics Emerge from Ivory Towers, 2001.- McGuinness. Ontologies and Online Commerce, 2001.- McGuinness. Conceptual Modeling for Distributed Ontology Environments, 2000.- McGuinness, Fikes, Rice, Wilder. An Environment for Merging and Testing Large Ontologies, 2000.- Brachman, Borgida, McGuinness, Patel-Schneider. Knowledge Representation meets Reality, 1999.- McGuinness. Ontological Issues for Knowledge-Enhanced Search, 1998.- McGuinness and Wright. Conceptual Modeling for Configuration, 1998.

Selected Tutorials:-Smith, Welty, McGuinness. OWL Web Ontology Language Guide, 2003.-Noy, McGuinness. Ontology Development 101: A Guide to Creating your First Ontology. 2001.- Brachman, McGuinness, Resnick, Borgida. How and When to Use a KL-ONE-like System, 1991.

Languages, Environments, Software:- OWL - http://www.w3.org/TR/owl-features/ , http://www.w3.org/TR/owl-guide/- DAML+OIL: http://www.daml.org/- Inference Web - http://www.ksl.stanford.edu/software/iw/ - Chimaera - http://www.ksl.stanford.edu/software/chimaera/ - FindUR - http://www.research.att.com/people/~dlm/findur/ - TAP – http://tap.stanford.edu/- DQL - http://www.ksl.stanford.edu/projects/dql/