Top Banner
1 1 An Overview and An Overview and Underview of the Underview of the Semantic Web Semantic Web Tim Finin Tim Finin University of Maryland University of Maryland Baltimore County Baltimore County Semantic Web for Science Workshop Newark NJ, October 2002 http://umbc.edu/~finin/talks/swsw02/ recommend tell register tell register UMBC UMBC an Honors University in Maryland an Honors University in Maryland 2 2 Overview Overview The Problem: building intelligent The Problem: building intelligent information systems information systems The Semantic web as part of the The Semantic web as part of the solution solution Background on the semantic web Background on the semantic web Comments and Conclusion Comments and Conclusion s s UMBC UMBC an Honors University in Maryland an Honors University in Maryland 3 3 The problem The problem I’ve been engaged in research aimed at I’ve been engaged in research aimed at developing intelligent information systems for developing intelligent information systems for thirty years. thirty years. The problem is hard, progress is slow, but the The problem is hard, progress is slow, but the incremental results are worth it. incremental results are worth it. It’s a task for many generations to come. It’s a task for many generations to come. Today’s environment is very different than that in Today’s environment is very different than that in 1972. 1972. UMBC UMBC an Honors University in Maryland an Honors University in Maryland 4 4 They way we were… They way we were… AN IBM 360 circa 1972 AN IBM 360 circa 1972
18

The Semantic web as part of thefinin/talks/swsw02/rutgers.pdf · 2002. 10. 24. · Next: overview of Semantic Web UMBC an Honors University in Maryland 11 W3C’s Semantic Web Goals

Aug 23, 2020

Download

Documents

dariahiddleston
Welcome message from author
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
Page 1: The Semantic web as part of thefinin/talks/swsw02/rutgers.pdf · 2002. 10. 24. · Next: overview of Semantic Web UMBC an Honors University in Maryland 11 W3C’s Semantic Web Goals

NSF

IDM

Wor

ksho

p M

ay 6

, 200

2Ti

m F

inin

, mai

lto:fi

nin@

umbc

.edu 1

11

An

Ove

rvie

w a

nd

An

Ove

rvie

w a

nd

Un

derv

iew

of

the

Un

derv

iew

of

the

Sem

anti

c W

ebSe

man

tic

Web

Tim

Fin

inTi

m F

inin

Un

iver

sity

of

Mar

ylan

d U

niv

ersi

ty o

f M

aryl

and

Bal

tim

ore

Cou

nty

Bal

tim

ore

Cou

nty

Sem

antic

Web

for

Sci

ence

Wor

ksho

pN

ewar

k N

J, O

ctob

er 2

002

http

://u

mbc

.edu

/~fin

in/t

alks

/sw

sw02

/

reco

mm

end

tell

regi

ster

tell

regi

ster

UM

BC

UM

BC

an H

onor

s U

nive

rsity

in M

aryl

and

an H

onor

s U

nive

rsity

in M

aryl

and

22

Ove

rvie

wO

verv

iew

The

Prob

lem

: bu

ildin

g in

telli

gent

Th

e Pr

oble

m:

build

ing

inte

llige

nt

info

rmat

ion

syst

ems

info

rmat

ion

syst

ems

The

Sem

antic

web

as

part

of

the

The

Sem

antic

web

as

part

of

the

solu

tion

solu

tion

Back

grou

nd o

n th

e se

man

tic w

ebBa

ckgr

ound

on

the

sem

antic

web

Com

men

ts a

nd C

oncl

usio

nCo

mm

ents

and

Con

clus

ion ss

UM

BC

UM

BC

an H

onor

s U

nive

rsity

in M

aryl

and

an H

onor

s U

nive

rsity

in M

aryl

and

33

The

prob

lem

The

prob

lem

I’ve

been

eng

aged

in r

esea

rch

aim

ed a

t I’v

e be

en e

ngag

ed in

res

earc

h ai

med

at

deve

lopi

ng in

telli

gent

info

rmat

ion

syst

ems

for

deve

lopi

ng in

telli

gent

info

rmat

ion

syst

ems

for

thirt

y ye

ars.

thirt

y ye

ars.

The

prob

lem

is h

ard,

pro

gres

s is

slo

w, b

ut t

he

The

prob

lem

is h

ard,

pro

gres

s is

slo

w, b

ut t

he

incr

emen

tal r

esul

ts a

re w

orth

it.

incr

emen

tal r

esul

ts a

re w

orth

it.

It’s

a t

ask

for

man

y ge

nera

tions

to

com

e.It

’s a

tas

k fo

r m

any

gene

ratio

ns t

o co

me.

Toda

y’s

envi

ronm

ent

is v

ery

diff

eren

t th

an t

hat

in

Toda

y’s

envi

ronm

ent

is v

ery

diff

eren

t th

an t

hat

in

1972

.19

72.

UM

BC

UM

BC

an H

onor

s U

nive

rsity

in M

aryl

and

an H

onor

s U

nive

rsity

in M

aryl

and

44

They

way

we

wer

e…Th

ey w

ay w

e w

ere…

AN I

BM 3

60 c

irca

1972

AN I

BM 3

60 c

irca

1972

Page 2: The Semantic web as part of thefinin/talks/swsw02/rutgers.pdf · 2002. 10. 24. · Next: overview of Semantic Web UMBC an Honors University in Maryland 11 W3C’s Semantic Web Goals

NSF

IDM

Wor

ksho

p M

ay 6

, 200

2Ti

m F

inin

, mai

lto:fi

nin@

umbc

.edu 2

UM

BC

UM

BC

an H

onor

s U

nive

rsity

in M

aryl

and

an H

onor

s U

nive

rsity

in M

aryl

and

55

They

way

we

will

be…

They

way

we

will

be…

UM

BC

UM

BC

an H

onor

s U

nive

rsity

in M

aryl

and

an H

onor

s U

nive

rsity

in M

aryl

and

66

Wh

at’s

new

?W

hat

’s n

ew?

Inte

rnet

.In

tern

et.

Virt

ually

of

the

com

pute

rs in

the

wor

ld

Virt

ually

of

the

com

pute

rs in

the

wor

ld

have

bee

n co

nnec

ted.

have

bee

n co

nnec

ted.

Scal

e.Sc

ale.

Ever

y da

y m

any

mor

e co

mpu

ting

and

Ever

y da

y m

any

mor

e co

mpu

ting

and

com

mun

icat

ion

devi

ces

are

join

ing.

com

mun

icat

ion

devi

ces

are

join

ing.

Pow

er.

Pow

er.

Raw

com

putin

g po

wer

con

tinue

s to

clim

b.Raw

com

putin

g po

wer

con

tinue

s to

clim

b.W

irel

ess.

Wir

eles

s.N

ew t

echn

olog

ies

(GSM

, 802

.11,

N

ew t

echn

olog

ies

(GSM

, 802

.11,

Bl

ueto

oth,

UW

B?, IR

, etc

) ar

e cr

eatin

g a

perv

asiv

e,

Blue

toot

h, U

WB?

, IR

, etc

) ar

e cr

eatin

g a

perv

asiv

e,

ubiq

uito

us c

ompu

ting

envi

ronm

ent

ubiq

uito

us c

ompu

ting

envi

ronm

ent

Web

.W

eb.

Anyo

ne c

an p

ublis

h co

nten

t an

d pr

ovid

e An

yone

can

pub

lish

cont

ent

and

prov

ide

serv

ices

, pow

erfu

l sea

rch

engi

nes

supp

ort

disc

over

y,

serv

ices

, pow

erfu

l sea

rch

engi

nes

supp

ort

disc

over

y,

evol

ving

sta

ndar

ds e

nhan

ce in

tero

pera

bilit

yev

olvi

ng s

tand

ards

enh

ance

inte

rope

rabi

lity

UM

BC

UM

BC

an H

onor

s U

nive

rsity

in M

aryl

and

an H

onor

s U

nive

rsity

in M

aryl

and

77

The

way

we

will

be…

The

way

we

will

be…

Peo

ple,

age

nts

, dev

ices

, & s

ervi

ces

nee

d to

Peo

ple,

age

nts

, dev

ices

, & s

ervi

ces

nee

d to

Find

oth

ers

in t

heir

envi

ronm

ent

Find

oth

ers

in t

heir

envi

ronm

ent

Des

crib

e th

e se

rvic

es t

hey

offe

r an

d se

ekD

escr

ibe

the

serv

ices

the

y of

fer

and

seek

Exch

ange

API

sEx

chan

ge A

PIs

Neg

otia

te f

or s

ervi

ces,

per

mis

sion

s, p

rivac

y, p

aym

ent,

Neg

otia

te f

or s

ervi

ces,

per

mis

sion

s, p

rivac

y, p

aym

ent,

Rea

son

abou

t se

rvic

es t

o cr

eate

com

posi

te s

ervi

ces

Rea

son

abou

t se

rvic

es t

o cr

eate

com

posi

te s

ervi

ces

Coor

dina

te a

nd c

oope

rate

as

need

edCo

ordi

nate

and

coo

pera

te a

s ne

eded

Sens

e th

eir

cont

ext

and

the

activ

ities

of

hum

ans

Sens

e th

eir

cont

ext

and

the

activ

ities

of

hum

ans

Dea

l with

new

ent

ities

nev

er b

efor

e en

coun

tere

dD

eal w

ith n

ew e

ntiti

es n

ever

bef

ore

enco

unte

red

An

d to

do

this

dyn

amic

ally

An

d to

do

this

dyn

amic

ally

UM

BC

UM

BC

an H

onor

s U

nive

rsity

in M

aryl

and

an H

onor

s U

nive

rsity

in M

aryl

and

88

Info

rmat

ion

an

d D

ata

Man

agem

ent

Ch

alle

nge

sIn

form

atio

n a

nd

Dat

a M

anag

emen

t C

hal

len

ges

The

envi

ronm

ent

mak

es n

ew d

eman

ds a

nd o

ffer

s Th

e en

viro

nmen

t m

akes

new

dem

ands

and

off

ers

new

cha

lleng

es, e

noug

h to

kee

p al

l of

us b

usy,

e.g

.:ne

w c

halle

nges

, eno

ugh

to k

eep

all o

f us

bus

y, e

.g.:

Very

ope

n en

viro

nmen

tsVe

ry o

pen

envi

ronm

ents

Larg

e an

d di

vers

e co

mm

unity

of

serv

ice

and

Larg

e an

d di

vers

e co

mm

unity

of

serv

ice

and

cont

ent

prov

ider

sco

nten

t pr

ovid

ers

Lots

of

rela

tive

auto

nom

yLo

ts o

f re

lativ

e au

tono

my

Dyn

amic

ad

hoc

netw

orks

Dyn

amic

ad

hoc

netw

orks

Syst

ems

with

wid

ely

vary

ing

reso

urce

s Sy

stem

s w

ith w

idel

y va

ryin

g re

sour

ces

----ba

ndw

idth

, con

nect

ivity

, cpu

, mem

ory,

dis

k,

band

wid

th, c

onne

ctiv

ity, c

pu, m

emor

y, d

isk,

po

wer

, sof

twar

e, k

now

ledg

e, in

telli

genc

e, e

tc.

pow

er, s

oftw

are,

kno

wle

dge,

inte

llige

nce,

etc

.

Page 3: The Semantic web as part of thefinin/talks/swsw02/rutgers.pdf · 2002. 10. 24. · Next: overview of Semantic Web UMBC an Honors University in Maryland 11 W3C’s Semantic Web Goals

NSF

IDM

Wor

ksho

p M

ay 6

, 200

2Ti

m F

inin

, mai

lto:fi

nin@

umbc

.edu 3

UM

BC

UM

BC

an H

onor

s U

nive

rsity

in M

aryl

and

an H

onor

s U

nive

rsity

in M

aryl

and

99

Res

earc

h to

pics

Res

earc

h to

pics

Con

cept

s ad

dres

sin

g th

ese

chal

len

ges

incl

ude

:C

once

pts

addr

essi

ng

thes

e ch

alle

nge

s in

clu

de:

Mul

tiage

nt s

yste

ms

Mul

tiage

nt s

yste

ms

Info

rmat

ion

and

know

ledg

e sh

arin

g th

roug

h co

mm

on

Info

rmat

ion

and

know

ledg

e sh

arin

g th

roug

h co

mm

on

repr

esen

tatio

n la

ngua

ges,

ont

olog

ies

and

prot

ocol

sre

pres

enta

tion

lang

uage

s, o

ntol

ogie

s an

d pr

otoc

ols

Auto

mat

ic s

ervi

ce d

escr

iptio

n, d

isco

very

, com

posi

tion

Auto

mat

ic s

ervi

ce d

escr

iptio

n, d

isco

very

, com

posi

tion

Neg

otia

tion

for

serv

ices

and

info

rmat

ion

Neg

otia

tion

for

serv

ices

and

info

rmat

ion

Trus

t ba

sed

mod

els

for

auth

oriz

atio

n, c

redi

bilit

y an

d se

curit

yTr

ust

base

d m

odel

s fo

r au

thor

izat

ion,

cre

dibi

lity

and

secu

rity

Soci

al a

nd n

orm

gov

erne

d be

havi

or

Soci

al a

nd n

orm

gov

erne

d be

havi

or

Del

egat

ion

and

degr

ees

of a

uton

omy

Del

egat

ion

and

degr

ees

of a

uton

omy

Coor

dina

tion

and

team

wor

k m

odel

sCo

ordi

natio

n an

d te

amw

ork

mod

els

UM

BC

UM

BC

an H

onor

s U

nive

rsity

in M

aryl

and

an H

onor

s U

nive

rsity

in M

aryl

and

1010

Sem

anti

c W

ebSe

man

tic

Web

I’ll a

rgue

tha

t th

e se

man

tic w

eb p

rovi

des

a go

od

I’ll a

rgue

tha

t th

e se

man

tic w

eb p

rovi

des

a go

od

appr

oach

, lan

guag

e an

d to

ols

to s

uppo

rt t

he

appr

oach

, lan

guag

e an

d to

ols

to s

uppo

rt t

he

deve

lopm

ent

of in

telli

gent

info

rmat

ion

syst

ems

in

deve

lopm

ent

of in

telli

gent

info

rmat

ion

syst

ems

in

this

env

ironm

ent.

this

env

ironm

ent.

This

isn’

t ob

viou

s, s

ince

the

SW

see

ms

grou

nded

Th

is is

n’t

obvi

ous,

sin

ce t

he S

W s

eem

s gr

ound

ed

in t

he “

trad

ition

al”

hype

rtex

t on

the

wire

d w

eb.

in t

he “

trad

ition

al”

hype

rtex

t on

the

wire

d w

eb.

But,

the

prin

cipl

es w

hich

driv

e it

are

the

right

Bu

t, t

he p

rinci

ples

whi

ch d

rive

it ar

e th

e rig

ht

ones

for

age

nts

as w

ell a

s pe

rvas

ive

com

putin

g.on

es f

or a

gent

s as

wel

l as

perv

asiv

e co

mpu

ting.

And,

by

grou

ndin

g ag

ents

in w

eb t

echn

olog

y,

And,

by

grou

ndin

g ag

ents

in w

eb t

echn

olog

y,

they

may

mak

e it

out

of t

he la

b.th

ey m

ay m

ake

it ou

t of

the

lab.

Nex

t: o

verv

iew

of

Sem

antic

Web

Nex

t: o

verv

iew

of

Sem

antic

Web

UM

BC

UM

BC

an H

onor

s U

nive

rsity

in M

aryl

and

an H

onor

s U

nive

rsity

in M

aryl

and

1111

W3

C’s

Sem

anti

c W

eb G

oals

W3

C’s

Sem

anti

c W

eb G

oals

Focu

s on

mac

hin

e co

nsu

mpt

ion

:Fo

cus

on m

ach

ine

con

sum

ptio

n:

"The

Sem

antic

Web

is a

n ex

tens

ion

of th

e cu

rren

t web

in

whi

ch in

form

atio

n is

giv

en w

ell-d

efin

ed m

eani

ng, b

ette

r en

ablin

g co

mpu

ters

and

peo

ple

to w

ork

in c

oope

ratio

n."

----Be

rner

sBe

rner

s --Le

e, H

endl

er a

nd L

assi

la, T

he S

eman

tic W

eb,

Lee,

Hen

dler

and

Las

sila

, The

Sem

antic

Web

, Sc

ient

ific

Amer

ican

, 200

1Sc

ient

ific

Amer

ican

, 200

1

Wh

erea

s th

e W

eb h

as m

ade

peop

le s

mar

ter,

W

her

eas

the

Web

has

mad

e pe

ople

sm

arte

r,

the

SW w

ill m

ake

mac

hin

es s

mar

ter.

the

SW w

ill m

ake

mac

hin

es s

mar

ter.

The

curr

ent

Web

Th

e cu

rren

t W

eb s

tore

sst

ores

thin

gs w

her

eas

the

thin

gs w

her

eas

the

SW e

nab

les

agen

ts w

hic

h

SW e

nab

les

agen

ts w

hic

h d

odoth

ings

.th

ings

.

UM

BC

UM

BC

an H

onor

s U

nive

rsity

in M

aryl

and

an H

onor

s U

nive

rsity

in M

aryl

and

1212

Ori

gin

s of

th

e Se

man

tic

Web

Ori

gin

s of

th

e Se

man

tic

Web

http

://w

ww

.w3.

org/

His

tory

/198

9/pr

opos

al.h

tml

Cap

sule

his

tory

Cap

sule

his

tory

Tim

Ber

ners

-Lee

pro

pose

d W

WW

as

a W

eb o

f re

latio

nshi

ps a

mon

g na

med

ob

ject

s (8

9)G

uha

desi

gned

MCF

(~

94)

XML+

MCF

=>

RD

F (~

96)

RDF+

OO

=>

RD

FS (

~99

)RD

FS+

KR=

>D

AML+

OIL

(00

)W

3C’s

SW

act

ivity

(01

)W

3C’s

OW

L (0

2?)

Page 4: The Semantic web as part of thefinin/talks/swsw02/rutgers.pdf · 2002. 10. 24. · Next: overview of Semantic Web UMBC an Honors University in Maryland 11 W3C’s Semantic Web Goals

NSF

IDM

Wor

ksho

p M

ay 6

, 200

2Ti

m F

inin

, mai

lto:fi

nin@

umbc

.edu 4

UM

BC

UM

BC

an H

onor

s U

nive

rsity

in M

aryl

and

an H

onor

s U

nive

rsity

in M

aryl

and

1313

Sem

anti

c W

eb

Sem

anti

c W

eb d

oes

does

wh

at?

wh

at?

Con

cept

Con

cept

-- bas

ed s

earc

hba

sed

sear

ch≠≠

keyw

ord

keyw

ord --

base

d se

arch

base

d se

arch

Sem

anti

c n

avig

atio

nSe

man

tic

nav

igat

ion

≠≠lin

klin

k --ba

sed

navi

gatio

nba

sed

navi

gatio

nP

erso

nal

izat

ion

Per

son

aliz

atio

n≠≠

one

size

fits

all

one

size

fits

all

Qu

ery

answ

erin

gQ

uer

y an

swer

ing

≠≠do

cum

ent

retr

ieva

ldo

cum

ent

retr

ieva

lSe

rvic

esSe

rvic

es≠≠

CGI

calls

, but

ser

vice

CGI

calls

, but

ser

vice

-- des

crip

tion

lang

uage

s,

desc

riptio

n la

ngua

ges,

ne

gotia

tion,

ne

gotia

tion,

se

rvic

e co

mpo

sitio

n,

etc

serv

ice

com

posi

tion,

et

c U

MB

CU

MB

Can

Hon

ors

Uni

vers

ity in

Mar

ylan

dan

Hon

ors

Uni

vers

ity in

Mar

ylan

d1414

Wh

y is

thi

s h

ard?

Wh

y is

thi

s h

ard?

afte

r Fr

ank

van

Har

mel

en a

nd J

im H

endl

er

UM

BC

UM

BC

an H

onor

s U

nive

rsity

in M

aryl

and

an H

onor

s U

nive

rsity

in M

aryl

and

1515

Wh

at a

web

pag

e lo

oks

like

to a

mac

hin

e…

And

und

erst

andi

ng

natu

ral l

angu

age

is

easi

er th

an im

ages

!

“Webscraping”

is

mos

tly d

one

by h

and

craf

ted

rule

s or

rule

s ge

nera

ted

by

supe

rvis

ed le

arni

ng

Eith

er w

ay, t

he ru

les

can

brea

k w

hen

the

page

str

uctu

re c

hang

es.

afte

r Fr

ank

van

Har

mel

en a

nd J

im H

endl

erU

MB

CU

MB

Can

Hon

ors

Uni

vers

ity in

Mar

ylan

dan

Hon

ors

Uni

vers

ity in

Mar

ylan

d1616

OK

, so

HTM

L is

not

hel

pfu

lO

K, s

o H

TML

is n

ot h

elpf

ulC

ould

we

tell

the

mac

hine

wha

t the

di

ffere

nt p

arts

of t

he

text

repr

esen

t?

afte

r Fr

ank

van

Har

mel

en a

nd J

im H

endl

er

title

time

spea

ker

loca

tion

abst

ract

bios

ketc

hho

st

Page 5: The Semantic web as part of thefinin/talks/swsw02/rutgers.pdf · 2002. 10. 24. · Next: overview of Semantic Web UMBC an Honors University in Maryland 11 W3C’s Semantic Web Goals

NSF

IDM

Wor

ksho

p M

ay 6

, 200

2Ti

m F

inin

, mai

lto:fi

nin@

umbc

.edu 5

UM

BC

UM

BC

an H

onor

s U

nive

rsity

in M

aryl

and

an H

onor

s U

nive

rsity

in M

aryl

and

1717

XM

L to

th

e re

scu

e?X

ML

to t

he

resc

ue?

XML

fans

pro

pose

cr

eatin

g a

XML

tag

set

to u

se fo

r eac

h ap

plic

atio

n.

For t

alks

, we

can

choo

se <

title

>,

<spe

aker

>, e

tc.

afte

r Fr

ank

van

Har

mel

en a

nd J

im H

endl

er

<tit

le>

<tim

e><

spea

ker>

<lo

catio

n>

<ab

stra

ct>

<bi

oske

tch>

<ho

st>

</h

ost>

</b

iosk

etch

>

</a

bstr

act>

</lo

catio

n>

</t

ime>

</s

peak

er>

</t

itle>

UM

BC

UM

BC

an H

onor

s U

nive

rsity

in M

aryl

and

an H

onor

s U

nive

rsity

in M

aryl

and

1818

XM

L X

ML ≠≠

mac

hin

e ac

cess

ible

mea

nin

gm

ach

ine

acce

ssib

le m

ean

ing B

ut,

to

you

r m

ach

ine,

the

ta

gs s

till

look

lik

e th

is…

.

The

tag

nam

es

carr

y n

o m

ean

ing.

XM

L D

TDs

and

Sch

emas

hav

e lit

tle

or n

o se

man

tics

.

afte

r Fr

ank

van

Har

mel

en a

nd J

im H

endl

er

<τιτλε>

<τιµε>

<σπεακερ>

<λοχατιον>

<αβσ

τραχτ>

<βιοσκ

ετχη

>

<ηοσ

τ></ηο

στ></βιοσ

κετχη>

</αβσ

τραχτ>

</λοχα

τιον>

</τιµε>

</σπ

εακερ>

</τιτλε>

UM

BC

UM

BC

an H

onor

s U

nive

rsity

in M

aryl

and

an H

onor

s U

nive

rsity

in M

aryl

and

1919

<?x

ml v

ersi

on=

"1.0

" en

codi

ng=

"utf

-8"?

><

xs:s

chem

a xm

lns:

xs=

"htt

p://

ww

w.w

3.or

g/20

01/X

MLS

chem

a">

<xs

:ele

men

t na

me=

"boo

k">

<xs

:com

plex

Type

><

xs:s

eque

nce>

<xs

:ele

men

t na

me=

"titl

e" t

ype=

"xs:

strin

g"/>

<xs

:ele

men

t na

me=

"aut

hor"

typ

e="x

s:st

ring"

/><

xs:e

lem

ent

nam

e="c

hara

cter

" m

inO

ccur

s="0

" m

axO

ccur

s="u

nbou

nded

"><

xs:c

ompl

exTy

pe>

<xs

:seq

uenc

e><

xs:e

lem

ent

nam

e="n

ame"

typ

e="x

s:st

ring"

/><

xs:e

lem

ent

nam

e="f

riend

-of"

typ

e="x

s:st

ring

" m

inO

ccur

s="0

"m

axO

ccur

s="u

nbou

nded

"/>

<xs

:ele

men

t na

me=

"sin

ce"

type

="x

s:da

te"/

><

xs:e

lem

ent

nam

e="q

ualif

icat

ion"

typ

e="x

s:st

ring"

/><

/xs:

sequ

ence

><

/xs:

com

plex

Type

><

/xs:

elem

ent>

</x

s:se

quen

ce>

<xs

:att

ribut

e na

me=

"isb

n" t

ype=

"xs:

strin

g"/>

</x

s:co

mpl

exTy

pe>

</x

s:el

emen

t><

/xs:

sche

ma>

XM

L Sc

hem

a h

elps

XM

L Sc

hem

a h

elps

XML

Sche

mas

pro

vide

a

XML

Sche

mas

pro

vide

a

sim

ple

mec

hani

sm t

o de

fine

sim

ple

mec

hani

sm t

o de

fine

shar

ed v

ocab

ular

ies.

shar

ed v

ocab

ular

ies.

<τιτλε>

<τιµε>

<σπεακερ>

<λοχατιον>

<αβσ

τραχτ>

<βιοσκετχη>

<ηοσ

τ></ηο

στ>

</βιοσ

κετχη>

</αβσ

τραχτ>

</λοχα

τιον>

</τιµε>

</σπεακερ>

</τιτλε>

<τιτλε>

<τιµε>

<σπεακερ>

<λοχατιον>

<αβσ

τραχτ>

<βιοσκετχη>

<ηοσ

τ></ηο

στ>

</βιοσ

κετχη>

</αβσ

τραχτ>

</λοχα

τιον>

</τιµε>

</σπεακερ>

</τιτλε>

XML

Sche

ma

file

afte

r Fr

ank

van

Har

mel

en a

nd J

im H

endl

erU

MB

CU

MB

Can

Hon

ors

Uni

vers

ity in

Mar

ylan

dan

Hon

ors

Uni

vers

ity in

Mar

ylan

d2020

Bu

t th

ere

are

man

y sc

hem

asB

ut

ther

e ar

e m

any

sch

emas

<τιτλε>

<τιµε>

<σπεακερ>

<λοχατιον>

<αβσ

τραχτ>

<βιοσκετχη>

<ηοσ

τ></ηο

στ>

</βιοσ

κετχη>

</αβσ

τραχτ>

</λοχα

τιον>

</τιµε>

</σπεακερ>

</τιτλε>

<title>

<tim

e><spe

aker

><lo

cation

><abs

trac

t>

<biosk

etch

>

<hos

t></ho

st>

</bios

ketch>

</ab

stract>

</loca

tion

></time>

</sp

eake

r></title>

<?x

ml v

ersi

on=

"1.0

" en

codi

ng=

"utf

-8"?

><

xs:s

chem

a xm

lns:

xs=

"htt

p://

ww

w.w

3.or

g/20

01/X

MLS

chem

a">

<xs

:ele

men

t na

me=

"boo

k">

<xs

:com

plex

Type

><

xs:s

eque

nce>

<xs

:ele

men

t na

me=

"titl

e" t

ype=

"xs:

strin

g"/>

<xs

:ele

men

t na

me=

"aut

hor"

typ

e="x

s:st

ring"

/><

xs:e

lem

ent

nam

e="c

hara

cter

" m

inO

ccur

s="0

" m

axO

ccur

s="u

nbou

nded

"><

xs:c

ompl

exTy

pe>

<xs

:seq

uenc

e><

xs:e

lem

ent

nam

e="n

ame"

typ

e="x

s:st

ring"

/><

xs:e

lem

ent

nam

e="f

riend

-of"

typ

e="x

s:st

ring

" m

inO

ccur

s="0

"m

axO

ccur

s="u

nbou

nded

"/>

<xs

:ele

men

t na

me=

"sin

ce"

type

="x

s:da

te"/

><

xs:e

lem

ent

nam

e="q

ualif

icat

ion"

typ

e="x

s:st

ring"

/><

/xs:

sequ

ence

><

/xs:

com

plex

Type

><

/xs:

elem

ent>

</x

s:se

quen

ce>

<xs

:att

ribut

e na

me=

"isb

n" t

ype=

"xs:

strin

g"/>

</x

s:co

mpl

exTy

pe>

</x

s:el

emen

t><

/xs:

sche

ma>XM

L Sc

hem

a fil

e 1

<?x

ml v

ersi

on=

"1.0

" en

codi

ng=

"utf

-8"?

><

xs:s

chem

a xm

lns:

xs=

"htt

p://

ww

w.w

3.or

g/20

01/X

MLS

chem

a">

<xs

:ele

men

t na

me=

"boo

k">

<xs

:com

plex

Type

><

xs:s

eque

nce>

<xs

:ele

men

t na

me=

"titl

e" t

ype=

"xs:

strin

g"/>

<xs

:ele

men

t na

me=

"aut

hor"

typ

e="x

s:st

ring"

/><

xs:e

lem

ent

nam

e="c

hara

cter

" m

inO

ccur

s="0

" m

axO

ccur

s="u

nbou

nded

"><

xs:c

ompl

exTy

pe>

<xs

:seq

uenc

e><

xs:e

lem

ent

nam

e="n

ame"

typ

e="x

s:st

ring"

/><

xs:e

lem

ent

nam

e="f

riend

-of"

typ

e="x

s:st

ring

" m

inO

ccur

s="0

"m

axO

ccur

s="u

nbou

nded

"/>

<xs

:ele

men

t na

me=

"sin

ce"

type

="x

s:da

te"/

><

xs:e

lem

ent

nam

e="q

ualif

icat

ion"

typ

e="x

s:st

ring"

/><

/xs:

sequ

ence

><

/xs:

com

plex

Type

><

/xs:

elem

ent>

</x

s:se

quen

ce>

<xs

:att

ribut

e na

me=

"isb

n" t

ype=

"xs:

strin

g"/>

</x

s:co

mpl

exTy

pe>

</x

s:el

emen

t><

/xs:

sche

ma>XM

L Sc

hem

a fil

e 42

afte

r Fr

ank

van

Har

mel

en a

nd J

im H

endl

er

Page 6: The Semantic web as part of thefinin/talks/swsw02/rutgers.pdf · 2002. 10. 24. · Next: overview of Semantic Web UMBC an Honors University in Maryland 11 W3C’s Semantic Web Goals

NSF

IDM

Wor

ksho

p M

ay 6

, 200

2Ti

m F

inin

, mai

lto:fi

nin@

umbc

.edu 6

UM

BC

UM

BC

an H

onor

s U

nive

rsity

in M

aryl

and

an H

onor

s U

nive

rsity

in M

aryl

and

2121

Ther

e’s

no

way

to

rela

te s

chem

aTh

ere’

s n

o w

ay t

o re

late

sch

ema

<τιτλε>

<τιµε>

<σπεακερ>

<λοχατιον>

<αβσ

τραχτ>

<βιοσκετχη>

<ηοσ

τ></ηο

στ>

</βιοσ

κετχη>

</αβσ

τραχτ>

</λοχα

τιον>

</τιµε>

</σπεακερ>

</τιτλε>

<title>

<tim

e><spe

aker

><lo

cation

><abs

trac

t>

<biosk

etch

>

<hos

t></ho

st>

</bios

ketch>

</ab

stract>

</loca

tion

></time>

</sp

eake

r></title>

<?x

ml v

ersi

on=

"1.0

" en

codi

ng=

"utf

-8"?

><

xs:s

chem

a xm

lns:

xs=

"htt

p://

ww

w.w

3.or

g/20

01/X

MLS

chem

a">

<xs

:ele

men

t na

me=

"boo

k">

<xs

:com

plex

Type

><

xs:s

eque

nce>

<xs

:ele

men

t na

me=

"titl

e" t

ype=

"xs:

strin

g"/>

<xs

:ele

men

t na

me=

"aut

hor"

typ

e="x

s:st

ring"

/><

xs:e

lem

ent

nam

e="c

hara

cter

" m

inO

ccur

s="0

" m

axO

ccur

s="u

nbou

nded

"><

xs:c

ompl

exTy

pe>

<xs

:seq

uenc

e><

xs:e

lem

ent

nam

e="n

ame"

typ

e="x

s:st

ring"

/><

xs:e

lem

ent

nam

e="f

riend

-of"

typ

e="x

s:st

ring

" m

inO

ccur

s="0

"m

axO

ccur

s="u

nbou

nded

"/>

<xs

:ele

men

t na

me=

"sin

ce"

type

="x

s:da

te"/

><

xs:e

lem

ent

nam

e="q

ualif

icat

ion"

typ

e="x

s:st

ring"

/><

/xs:

sequ

ence

><

/xs:

com

plex

Type

><

/xs:

elem

ent>

</x

s:se

quen

ce>

<xs

:att

ribut

e na

me=

"isb

n" t

ype=

"xs:

strin

g"/>

</x

s:co

mpl

exTy

pe>

</x

s:el

emen

t><

/xs:

sche

ma>XM

L Sc

hem

a fil

e 1

<?x

ml v

ersi

on=

"1.0

" en

codi

ng=

"utf

-8"?

><

xs:s

chem

a xm

lns:

xs=

"htt

p://

ww

w.w

3.or

g/20

01/X

MLS

chem

a">

<xs

:ele

men

t na

me=

"boo

k">

<xs

:com

plex

Type

><

xs:s

eque

nce>

<xs

:ele

men

t na

me=

"titl

e" t

ype=

"xs:

strin

g"/>

<xs

:ele

men

t na

me=

"aut

hor"

typ

e="x

s:st

ring"

/><

xs:e

lem

ent

nam

e="c

hara

cter

" m

inO

ccur

s="0

" m

axO

ccur

s="u

nbou

nded

"><

xs:c

ompl

exTy

pe>

<xs

:seq

uenc

e><

xs:e

lem

ent

nam

e="n

ame"

typ

e="x

s:st

ring"

/><

xs:e

lem

ent

nam

e="f

riend

-of"

typ

e="x

s:st

ring

" m

inO

ccur

s="0

"m

axO

ccur

s="u

nbou

nded

"/>

<xs

:ele

men

t na

me=

"sin

ce"

type

="x

s:da

te"/

><

xs:e

lem

ent

nam

e="q

ualif

icat

ion"

typ

e="x

s:st

ring"

/><

/xs:

sequ

ence

><

/xs:

com

plex

Type

><

/xs:

elem

ent>

</x

s:se

quen

ce>

<xs

:att

ribut

e na

me=

"isb

n" t

ype=

"xs:

strin

g"/>

</x

s:co

mpl

exTy

pe>

</x

s:el

emen

t><

/xs:

sche

ma>XM

L Sc

hem

a fil

e 42

Eith

er m

anu

ally

or

auto

mat

ical

ly

Eith

er m

anu

ally

or

auto

mat

ical

ly -

---X

ML

XM

L Sc

hem

a is

ver

y w

eak

on s

eman

tics

Sch

ema

is v

ery

wea

k on

sem

anti

csU

MB

CU

MB

Can

Hon

ors

Uni

vers

ity in

Mar

ylan

dan

Hon

ors

Uni

vers

ity in

Mar

ylan

d2222

On

tolo

gies

can

hel

p …

On

tolo

gies

can

hel

p …

An

on

tolo

gy d

efin

es t

he

term

s u

sed

to d

escr

ibe

An

on

tolo

gy d

efin

es t

he

term

s u

sed

to d

escr

ibe

and

repr

esen

t an

are

a of

kn

owle

dge.

an

d re

pres

ent

an a

rea

of k

now

ledg

e.

Ont

olog

ies

are

used

by

peop

le, d

atab

ases

, and

ap

plic

atio

ns t

hat

need

to

shar

e do

mai

n in

form

atio

n (a

do

mai

n is

just

a s

peci

fic s

ubje

ct a

rea

or a

rea

of

know

ledg

e, li

ke m

edic

ine,

too

l man

ufac

turin

g, r

eal

esta

te, a

utom

obile

rep

air,

fin

anci

al m

anag

emen

t, e

tc.)

. O

ntol

ogie

s in

clud

e co

mpu

ter-

usab

le d

efin

ition

s of

bas

ic

conc

epts

in t

he d

omai

n an

d th

e re

latio

nshi

ps a

mon

g th

em ..

.Th

ey e

nco

de k

now

ledg

e in

a d

omai

n a

nd

also

Th

ey e

nco

de k

now

ledg

e in

a d

omai

n a

nd

also

kn

owle

dge

that

spa

ns

dom

ain

s.

know

ledg

e th

at s

pan

s do

mai

ns.

In

thi

s w

ay, t

hey

mak

e th

at k

now

ledg

e re

usab

le.

Wor

king

Dra

ft, W

eb O

ntol

ogy

Wor

king

Gro

up.

UM

BC

UM

BC

an H

onor

s U

nive

rsity

in M

aryl

and

an H

onor

s U

nive

rsity

in M

aryl

and

2323

On

tolo

gies

can

hel

p …

On

tolo

gies

can

hel

p …

Cat

alog

/ID

Gen

eral

Logi

cal

cons

train

ts

Term

s/gl

ossa

ry

Thes

auri

“nar

row

erte

rm”

rela

tion

Form

alis

-aFr

ames

(pro

perti

es)

Info

rmal

is-a

Form

alin

stan

ceV

alue

R

estri

ctio

n

Dis

join

tnes

s, In

vers

e,pa

rt of

Afte

r D

ebor

ah L

. McG

uinn

ess

(Sta

nfor

d)Af

ter

Deb

orah

L. M

cGui

nnes

s (S

tanf

ord)

Sim

ple

Taxo

nom

ies

Expr

essi

veO

nto

logi

es

Wor

dnet

CY

CR

DF

DA

ML

OO

DB

Sch

ema

RD

FSIE

EE

SU

OO

WL

UM

LS

UM

BC

UM

BC

an H

onor

s U

nive

rsity

in M

aryl

and

an H

onor

s U

nive

rsity

in M

aryl

and

2424

An

On

tolo

gy le

vel i

s n

eede

dA

n O

nto

logy

leve

l is

nee

ded

<?x

ml v

ersi

on=

"1.0

" en

codi

ng=

"utf

-8"?

><

xs:s

chem

a xm

lns:

xs=

"htt

p://

ww

w.w

3.or

g/20

01/X

MLS

chem

a">

<xs

:ele

men

t na

me=

"boo

k">

<xs

:com

plex

Type

><

xs:s

eque

nce>

<xs

:ele

men

t na

me=

"titl

e" t

ype=

"xs:

strin

g"/>

<xs

:ele

men

t na

me=

"aut

hor"

typ

e="x

s:st

ring"

/><

xs:e

lem

ent

nam

e="c

hara

cter

" m

inO

ccur

s="0

" m

axO

ccur

s="u

nbou

nded

"><

xs:c

ompl

exTy

pe>

<xs

:seq

uenc

e><

xs:e

lem

ent

nam

e="n

ame"

typ

e="x

s:st

ring"

/><

xs:e

lem

ent

nam

e="f

riend

-of"

typ

e="x

s:st

ring

" m

inO

ccur

s="0

"m

axO

ccur

s="u

nbou

nded

"/>

<xs

:ele

men

t na

me=

"sin

ce"

type

="x

s:da

te"/

><

xs:e

lem

ent

nam

e="q

ualif

icat

ion"

typ

e="x

s:st

ring"

/><

/xs:

sequ

ence

><

/xs:

com

plex

Type

><

/xs:

elem

ent>

</x

s:se

quen

ce>

<xs

:att

ribut

e na

me=

"isb

n" t

ype=

"xs:

strin

g"/>

</x

s:co

mpl

exTy

pe>

</x

s:el

emen

t><

/xs:

sche

ma>XM

L O

ntol

ogy

1

<?x

ml v

ersi

on=

"1.0

" en

codi

ng=

"utf

-8"?

><

xs:s

chem

a xm

lns:

xs=

"htt

p://

ww

w.w

3.or

g/20

01/X

MLS

chem

a">

<xs

:ele

men

t na

me=

"boo

k">

<xs

:com

plex

Type

><

xs:s

eque

nce>

<xs

:ele

men

t na

me=

"titl

e" t

ype=

"xs:

strin

g"/>

<xs

:ele

men

t na

me=

"aut

hor"

typ

e="x

s:st

ring"

/><

xs:e

lem

ent

nam

e="c

hara

cter

" m

inO

ccur

s="0

" m

axO

ccur

s="u

nbou

nded

"><

xs:c

ompl

exTy

pe>

<xs

:seq

uenc

e><

xs:e

lem

ent

nam

e="n

ame"

typ

e="x

s:st

ring"

/><

xs:e

lem

ent

nam

e="f

riend

-of"

typ

e="x

s:st

ring

" m

inO

ccur

s="0

"m

axO

ccur

s="u

nbou

nded

"/>

<xs

:ele

men

t na

me=

"sin

ce"

type

="x

s:da

te"/

><

xs:e

lem

ent

nam

e="q

ualif

icat

ion"

typ

e="x

s:st

ring"

/><

/xs:

sequ

ence

><

/xs:

com

plex

Type

><

/xs:

elem

ent>

</x

s:se

quen

ce>

<xs

:att

ribut

e na

me=

"isb

n" t

ype=

"xs:

strin

g"/>

</x

s:co

mpl

exTy

pe>

</x

s:el

emen

t><

/xs:

sche

ma>XM

L O

ntol

ogy

42

We

nee

d a

way

to

defi

ne

onto

logi

es in

XM

LW

e n

eed

a w

ay t

o de

fin

e on

tolo

gies

in X

ML

So w

e ca

n re

late

the

mSo

we

can

rela

te t

hem

So m

achi

nes

can

unde

rsta

nd (

to s

ome

degr

ee)

thei

r m

eani

ngSo

mac

hine

s ca

n un

ders

tand

(to

som

e de

gree

) th

eir

mea

ning

<?x

ml v

ersi

on=

"1.0

" en

codi

ng=

"utf

-8"?

><

xs:s

chem

a xm

lns:

xs=

"htt

p://

ww

w.w

3.or

g/20

01/X

MLS

chem

a">

<xs

:ele

men

t na

me=

"boo

k">

<xs

:com

plex

Type

><

xs:s

eque

nce>

<xs

:ele

men

t na

me=

"titl

e" t

ype=

"xs:

strin

g"/>

<xs

:ele

men

t na

me=

"aut

hor"

typ

e="x

s:st

ring"

/><

xs:e

lem

ent

nam

e="c

hara

cter

" m

inO

ccur

s="0

" m

axO

ccur

s="u

nbou

nded

"><

xs:c

ompl

exTy

pe>

<xs

:seq

uenc

e><

xs:e

lem

ent

nam

e="n

ame"

typ

e="x

s:st

ring"

/><

xs:e

lem

ent

nam

e="f

riend

-of"

typ

e="x

s:st

ring

" m

inO

ccur

s="0

"m

axO

ccur

s="u

nbou

nded

"/>

<xs

:ele

men

t na

me=

"sin

ce"

type

="x

s:da

te"/

><

xs:e

lem

ent

nam

e="q

ualif

icat

ion"

typ

e="x

s:st

ring"

/><

/xs:

sequ

ence

><

/xs:

com

plex

Type

><

/xs:

elem

ent>

</x

s:se

quen

ce>

<xs

:att

ribut

e na

me=

"isb

n" t

ype=

"xs:

strin

g"/>

</x

s:co

mpl

exTy

pe>

</x

s:el

emen

t><

/xs:

sche

ma>XM

L O

ntol

ogy

256

impo

rts

impo

rts

= <>

On

tolo

gies

add

•St

ruct

ure

•Co

nstr

aint

s•

Map

ping

s•

Shar

ing

<?x

ml v

ersi

on=

"1.0

" en

codi

ng=

"utf

-8"?

><

xs:s

chem

a xm

lns:

xs=

"htt

p://

ww

w.w

3.or

g/20

01/X

MLS

chem

a">

<xs

:ele

men

t na

me=

"boo

k">

<xs

:com

plex

Type

><

xs:s

eque

nce>

<xs

:ele

men

t na

me=

"titl

e"

type

="x

s:st

ring"

/><

xs:e

lem

ent

nam

e="a

utho

r"

type

="x

s:st

ring"

/><

xs:e

lem

ent

nam

e="c

hara

cter

" m

inO

ccur

s="0

" m

axO

ccur

s="u

nbou

nded

"><

xs:c

ompl

exTy

pe>

<xs

:seq

uenc

e><

xs:e

lem

ent

nam

e="n

ame"

ty

pe=

"xs:

strin

g"/>

<xs

:ele

men

t na

me=

"frie

nd-o

f"

XML

Ont

olog

y51

2

refe

renc

es

Page 7: The Semantic web as part of thefinin/talks/swsw02/rutgers.pdf · 2002. 10. 24. · Next: overview of Semantic Web UMBC an Honors University in Maryland 11 W3C’s Semantic Web Goals

NSF

IDM

Wor

ksho

p M

ay 6

, 200

2Ti

m F

inin

, mai

lto:fi

nin@

umbc

.edu 7

UM

BC

UM

BC

an H

onor

s U

nive

rsity

in M

aryl

and

an H

onor

s U

nive

rsity

in M

aryl

and

2525

On

tolo

gies

var

y…O

nto

logi

es v

ary…

Ont

olog

ies

vary

gre

atly

in t

heir

Ont

olog

ies

vary

gre

atly

in t

heir

Scop

eSc

ope

Com

plex

ityCo

mpl

exity

Leve

l of

deta

ilLe

vel o

f de

tail

Kind

of

know

ledg

e en

code

dKi

nd o

f kn

owle

dge

enco

ded

……

Two

exam

ples

…Tw

o ex

ampl

es…

UM

BC

UM

BC

an H

onor

s U

nive

rsity

in M

aryl

and

an H

onor

s U

nive

rsity

in M

aryl

and

2626

Du

blin

Cor

e D

ubl

in C

ore -

---A

Sim

ple

On

tolo

gyA

Sim

ple

On

tolo

gy

Dev

elop

ed b

y an

OCL

C sp

onso

red

Dev

elop

ed b

y an

OCL

C sp

onso

red

wor

ksho

p in

Dub

lin ~

95 a

s a

wor

ksho

p in

Dub

lin ~

95 a

s a

stan

dard

for

met

adat

a fo

r di

gita

l st

anda

rd f

or m

etad

ata

for

digi

tal

libra

ry r

esou

rces

on

web

libra

ry r

esou

rces

on

web

Cons

ists

of

15 c

ore

attr

ibut

es

Cons

ists

of

15 c

ore

attr

ibut

es

http

://d

ublin

core

.org

/ht

tp:/

/dub

linco

re.o

rg/

Neu

tral

on

how

DC

shou

ld b

e N

eutr

al o

n ho

w D

C sh

ould

be

repr

esen

ted

repr

esen

ted

HTM

L fo

und

to b

e in

adeq

uate

for

H

TML

foun

d to

be

inad

equa

te f

or

repr

esen

ting

com

plex

ities

of

repr

esen

ting

com

plex

ities

of

stru

ctur

ed u

se o

f D

Cst

ruct

ured

use

of

DC

Avai

labl

e as

an

RD

F sc

hem

a.Av

aila

ble

as a

n RD

F sc

hem

a.

15

DC

ele

men

ts1

5 D

C e

lem

ents

Con

ten

t el

emen

tsCo

vera

geCo

vera

geD

escr

iptio

nD

escr

iptio

nRel

atio

nRel

atio

nSo

urce

Sour

ceSu

bjec

tSu

bjec

tTi

tleTi

tleTy

pe

Type

In

telle

ctu

al P

rope

rty

Cont

ribut

orCo

ntrib

utor

Crea

tor

Crea

tor

Publ

ishe

rPu

blis

her

Rig

htRig

htIn

stan

tiat

ion

D

ate

Dat

eFo

rmat

Form

atId

entif

ier

Iden

tifie

rLa

ngua

geLa

ngua

ge

UM

BC

UM

BC

an H

onor

s U

nive

rsity

in M

aryl

and

an H

onor

s U

nive

rsity

in M

aryl

and

2727

Cyc

C

yc ––

a co

mpl

ex o

nto

logy

a co

mpl

ex o

nto

logy

Cyc

is a

larg

e, g

ener

al p

urpo

se o

ntol

ogy

with

Cy

c is

a la

rge,

gen

eral

pur

pose

ont

olog

y w

ith

asso

ciat

ed r

easo

ning

too

ls d

evel

oped

ove

r th

e as

soci

ated

rea

soni

ng t

ools

dev

elop

ed o

ver

the

past

~20

yea

rs b

y M

CC a

nd n

ow C

ycor

ppa

st ~

20 y

ears

by

MCC

and

now

Cyc

orp

Cyc

KB h

as >

100

k te

rms.

Cyc

KB h

as >

100

k te

rms.

Term

s ar

e ax

iom

atiz

ed b

y >

1M

han

dcra

fted

ass

ertio

nsTe

rms

are

axio

mat

ized

by

> 1

M h

andc

raft

ed a

sser

tions

Cyc

infe

renc

e en

gine

has

> 5

00 h

euris

tic le

vel m

odul

esCy

c in

fere

nce

engi

ne h

as >

500

heu

ristic

leve

l mod

ules

Goa

l is

to e

ncod

e kn

owle

dge

for

“com

mon

sen

se

Goa

l is

to e

ncod

e kn

owle

dge

for

“com

mon

sen

se

reas

onin

g” n

eede

d by

app

licat

ions

(e.

g., N

LP)

reas

onin

g” n

eede

d by

app

licat

ions

(e.

g., N

LP)

Avai

labl

e fr

ee in

lim

ited

form

fro

m

Avai

labl

e fr

ee in

lim

ited

form

fro

m

http

://o

penc

yc.o

rg/

http

://o

penc

yc.o

rg/

UM

BC

UM

BC

an H

onor

s U

nive

rsity

in M

aryl

and

an H

onor

s U

nive

rsity

in M

aryl

and

2828

Toda

y an

d to

mor

row

Toda

y an

d to

mor

row

We

are

in a

goo

d po

sitio

n to

use

sim

ple

We

are

in a

goo

d po

sitio

n to

use

sim

ple

onto

logi

es li

ke D

C to

day

onto

logi

es li

ke D

C to

day

This

is h

appe

ning

(e.

g., A

dobe

’s X

MP)

This

is h

appe

ning

(e.

g., A

dobe

’s X

MP)

We

hope

to

be a

ble

to m

ake

effe

ctiv

e us

e W

e ho

pe t

o be

abl

e to

mak

e ef

fect

ive

use

onto

logi

es li

ke C

yc in

the

com

ing

deca

deon

tolo

gies

like

Cyc

in t

he c

omin

g de

cade

Ther

e ar

e sk

eptic

s…Th

ere

are

skep

tics…

It’s

a g

reat

res

earc

h to

pic…

It’s

a g

reat

res

earc

h to

pic…

Page 8: The Semantic web as part of thefinin/talks/swsw02/rutgers.pdf · 2002. 10. 24. · Next: overview of Semantic Web UMBC an Honors University in Maryland 11 W3C’s Semantic Web Goals

NSF

IDM

Wor

ksho

p M

ay 6

, 200

2Ti

m F

inin

, mai

lto:fi

nin@

umbc

.edu 8

UM

BC

UM

BC

an H

onor

s U

nive

rsity

in M

aryl

and

an H

onor

s U

nive

rsity

in M

aryl

and

2929

TB

L’s

sem

antic

web

vis

ion

“The

Sem

antic

Web

w

ill g

loba

lize

KR,

just

as

the

WW

W

glob

aliz

e hy

pert

ext”

--Ti

m B

erne

rs-L

ee

you

are

her

e

UM

BC

UM

BC

an H

onor

s U

nive

rsity

in M

aryl

and

an H

onor

s U

nive

rsity

in M

aryl

and

3030

Sem

anti

c w

eb la

ngu

ages

tod

aySe

man

tic

web

lan

guag

es t

oday

Toda

y th

ere

are,

IM

HO

, tw

o se

man

tic w

eb

Toda

y th

ere

are,

IM

HO

, tw

o se

man

tic w

eb

lang

uage

sla

ngua

ges

DA

ML+

OIL

DA

ML+

OIL

––D

arpa

Age

nt M

arku

p La

ngua

geD

arpa

Age

nt M

arku

p La

ngua

geht

tp:/

/ww

w.d

aml.o

rg/

http

://w

ww

.dam

l.org

/R

DF

RD

F ––

Res

ourc

e D

escr

iptio

n Fr

amew

ork

Res

ourc

e D

escr

iptio

n Fr

amew

ork

http

://w

ww

.w3.

org/

RD

F/ht

tp:/

/ww

w.w

3.or

g/RD

F/an

d on

e un

der

deve

lopm

ent

by t

he W

3Can

d on

e un

der

deve

lopm

ent

by t

he W

3CO

WL

OW

L––

Ont

olog

y W

eb L

angu

age

Ont

olog

y W

eb L

angu

age

http

://w

ww

.w3.

org/

2001

/sw

/ht

tp:/

/ww

w.w

3.or

g/20

01/s

w/

Topi

c m

aps

Topi

c m

aps

(( htt

p://

topi

cmap

s.or

ght

tp:/

/top

icm

aps.

org /

) ar

e an

othe

r /)

are

ano

ther

br

eed

bree

dw

ith m

ore

to c

ome…

.w

ith m

ore

to c

ome…

.

UM

BC

UM

BC

an H

onor

s U

nive

rsity

in M

aryl

and

an H

onor

s U

nive

rsity

in M

aryl

and

3131

Topi

c M

aps

Topi

c M

aps

A To

pic

Map

is a

col

lect

ion

of t

opic

s an

d A

Topi

c M

ap is

a c

olle

ctio

n of

top

ics

and

(sem

antic

ally

mea

ning

ful)

rela

tions

hips

bet

wee

n (s

eman

tical

ly m

eani

ngfu

l) re

latio

nshi

ps b

etw

een

thes

e to

pics

thes

e to

pics

Topi

c M

aps

link

thes

e to

pics

with

ext

erna

l To

pic

Map

s lin

k th

ese

topi

cs w

ith e

xter

nal

refe

renc

es, s

uch

as r

esou

rces

beh

ind

URIs

refe

renc

es, s

uch

as r

esou

rces

beh

ind

URIs

Topi

c M

aps

are

a “s

uper

impo

sed

sem

antic

laye

r”To

pic

Map

s ar

e a

“sup

erim

pose

d se

man

tic la

yer”

conn

ectio

n be

twee

n to

pics

and

res

ourc

es a

re U

RLs

conn

ectio

n be

twee

n to

pics

and

res

ourc

es a

re U

RLs

Topi

c M

aps

capt

ure

real

Topi

c M

aps

capt

ure

real

-- wor

ld s

ubje

cts/

obje

cts

but

wor

ld s

ubje

cts/

obje

cts

but

also

con

cept

sal

so c

once

pts

thes

e ar

e de

fined

not

abs

olut

e bu

t re

lativ

e to

eac

h ot

her

thes

e ar

e de

fined

not

abs

olut

e bu

t re

lativ

e to

eac

h ot

her

UM

BC

UM

BC

an H

onor

s U

nive

rsity

in M

aryl

and

an H

onor

s U

nive

rsity

in M

aryl

and

3232

RD

F is

th

e fi

rst

SW la

ngu

age

RD

F is

th

e fi

rst

SW la

ngu

age

<rdf

:RDF

…….

.><…

.><…

.></

rdf:

RDF>

XML

Enco

ding

Gra

ph

stm

t(doc

Inst

, rdf

_typ

e, D

ocum

ent)

stm

t(per

sonI

nst,

rdf_

type

, Per

son)

stm

t(inr

oom

Inst

, rdf

_typ

e, In

Roo

m)

stm

t(per

sonI

nst,

hold

ing,

doc

Inst

)st

mt(i

nroo

mIn

st, p

erso

n, p

erso

nIns

t)

Trip

les

RD

FD

ata

Mod

elG

ood

for

Mac

hine

Proc

essi

ng

Goo

d Fo

r H

uman

View

ing

Goo

d Fo

r R

easo

ning

Page 9: The Semantic web as part of thefinin/talks/swsw02/rutgers.pdf · 2002. 10. 24. · Next: overview of Semantic Web UMBC an Honors University in Maryland 11 W3C’s Semantic Web Goals

NSF

IDM

Wor

ksho

p M

ay 6

, 200

2Ti

m F

inin

, mai

lto:fi

nin@

umbc

.edu 9

UM

BC

UM

BC

an H

onor

s U

nive

rsity

in M

aryl

and

an H

onor

s U

nive

rsity

in M

aryl

and

3333

Sim

ple

RD

F Ex

ampl

eSi

mpl

e R

DF

Exam

ple

http

://um

bc.e

du/~

finin

/talk

s/id

m02

/“I

ntel

ligen

t Inf

orm

atio

n Sy

stem

son

the

Web

and

in th

e A

ethe

r”

http

://um

bc.e

du/

dc:T

itle

dc:C

reat

or

bib:

Aff

“Tim

Fin

in”

“fin

in@

umbc

.edu

bib:

nam

ebi

b:em

ail

UM

BC

UM

BC

an H

onor

s U

nive

rsity

in M

aryl

and

an H

onor

s U

nive

rsity

in M

aryl

and

3434

XM

L en

codi

ng

for

RD

FX

ML

enco

din

g fo

r R

DF

<rdf

:RD

F xm

lns:

rdf=

"http

://w

ww

.w3.

org/

1999

/02/

22-r

df-s

ynta

x-ns

#"xm

lns:

dc="

http

://pu

rl.or

g/dc

/ele

men

ts/1

.1/"

xmln

s:bi

b="h

ttp://

dam

l.um

bc.e

du/o

ntol

ogie

s/bi

b/">

<des

crip

tion

abou

t="h

ttp://

umbc

.edu

/~fin

in/ta

lks/

idm

02/"

><d

c:tit

le> I

ntel

ligen

t Inf

orm

atio

n Sy

stem

s on

the

Web

and

in th

e A

ethe

r</d

c:Ti

tle>

<dc:

crea

tor>

<des

crip

tion>

<bib

:Nam

e>Ti

m F

inin

</bi

b:N

ame>

<bib

:Em

ail>

finin

@um

bc.e

du</

bib:

Emai

l><b

ib:A

ff re

sour

ce="

http

://um

bc.e

du/"

/></

desc

riptio

n>

</dc

:Cre

ator

></

desc

riptio

n></

rdf:R

DF>

“ian

@go

o.or

g”

UM

BC

UM

BC

an H

onor

s U

nive

rsity

in M

aryl

and

an H

onor

s U

nive

rsity

in M

aryl

and

3535

N t

ripl

e re

pres

enta

tion

N t

ripl

e re

pres

enta

tion

RD

F ca

n be

enc

oded

as

a se

t of

RD

F ca

n be

enc

oded

as

a se

t of

trip

les

trip

les .

. <<

subj

ect

subj

ect >

<>

<pr

edic

ate

pred

icat

e > <

> <

obje

ctob

ject

> .

> .

<ht

tp:/

/um

bc.e

du/~

finin

/tal

ks/id

m02

/>

<ht

tp:/

/pur

l.org

/dc/

elem

ents

/1.1

/Titl

e>

"Int

ellig

ent

Info

rmat

ion

Syst

ems

on t

he W

eb a

nd in

the

Aet

her"

._:

j109

49 <

http

://d

aml.u

mbc

.edu

/ont

olog

ies/

bib/

Nam

e> "

Tim

Fin

in"

._:

j109

49 <

http

://d

aml.u

mbc

.edu

/ont

olog

ies/

bib/

Emai

l> "

finin

@um

bc.e

du"

._:

j109

49 <

http

://d

aml.u

mbc

.edu

/ont

olog

ies/

bib/

Aff>

<ht

tp:/

/um

bc.e

du/>

._:

j109

49 <

http

://w

ww

.w3.

org/

1999

/02/

22-r

df-s

ynta

x-ns

#ty

pe>

<

Des

crip

tion>

.<

http

://u

mbc

.edu

/~fin

in/t

alks

/idm

02/>

<

http

://p

url.o

rg/d

c/el

emen

ts/1

.1/C

reat

or>

_:j

1094

9 .

<ht

tp:/

/um

bc.e

du/~

finin

/tal

ks/id

m02

/> <

http

://w

ww

.w3.

org/

1999

/02/

22-r

df-

synt

ax-n

s#ty

pe>

<D

escr

iptio

n> .

Not

e th

e ge

nsym

for

the

ano

nym

ous

node

(N

ote

the

gens

ym f

or t

he a

nony

mou

s no

de (

_:j1

0949

))

UM

BC

UM

BC

an H

onor

s U

nive

rsity

in M

aryl

and

an H

onor

s U

nive

rsity

in M

aryl

and

3636

Trip

le N

otes

Trip

le N

otes

RD

F tr

iple

s h

ave

one

of t

wo

form

s:R

DF

trip

les

hav

e on

e of

tw

o fo

rms:

<U

RI>

<U

RI>

<U

RI>

<U

RI>

<U

RI>

<U

RI>

<U

RI>

<U

RI>

<qu

oted

str

ing>

<U

RI>

<U

RI>

<qu

oted

str

ing>

Trip

les

are

also

eas

ily m

appe

d in

to lo

gic

Trip

les

are

also

eas

ily m

appe

d in

to lo

gic

<su

bjec

t> <

pred

icat

e> <

obje

ct>

<su

bjec

t> <

pred

icat

e> <

obje

ct>

<pr

edic

ate>

(<su

bjec

t>,<

obje

ct>

)<

pred

icat

e>(<

subj

ect>

,<ob

ject

>)

With

typ

e(<

S>,<

O>

) be

com

ing

<O

>(<

S>)

With

typ

e(<

S>,<

O>

) be

com

ing

<O

>(<

S>)

Exam

ple:

Exam

ple:

subc

lass

(man

,per

son)

subc

lass

(man

,per

son)

sex(

man

,mal

e)se

x(m

an,m

ale)

dom

ain(

sex,

anim

al)

dom

ain(

sex,

anim

al)

man

(ada

m)

man

(ada

m)

age(

adam

,100

)ag

e(ad

am,1

00)

Trip

les

are

easi

ly s

tore

d an

d m

anag

ed in

a D

BM

STr

iple

s ar

e ea

sily

sto

red

and

man

aged

in a

DB

MS

; N

ote:

we’

re n

ot

; sh

owin

g th

e ac

tual

;

URI

s fo

r cl

arity

Page 10: The Semantic web as part of thefinin/talks/swsw02/rutgers.pdf · 2002. 10. 24. · Next: overview of Semantic Web UMBC an Honors University in Maryland 11 W3C’s Semantic Web Goals

NSF

IDM

Wor

ksho

p M

ay 6

, 200

2Ti

m F

inin

, mai

lto:fi

nin@

umbc

.edu 10

UM

BC

UM

BC

an H

onor

s U

nive

rsity

in M

aryl

and

an H

onor

s U

nive

rsity

in M

aryl

and

3737

N3

not

atio

n f

or R

DF

N3

not

atio

n f

or R

DF

N3

is a

com

pact

not

atio

n fo

r tr

iple

s w

hich

is e

asie

r N

3 is

a c

ompa

ct n

otat

ion

for

trip

les

whi

ch is

eas

ier

for

peop

le t

o re

ad a

nd e

dit

for

peop

le t

o re

ad a

nd e

dit

Exam

ple:

Exam

ple:

@pr

efix

log:

@

pref

ix lo

g:

<ht

tp:/

/ww

w.w

3.or

g/20

00/1

0/sw

ap/lo

g#>

.<

http

://w

ww

.w3.

org/

2000

/10/

swap

/log#

> .

:Per

son

a rd

fs:C

lass

.:P

erso

n a

rdfs

:Cla

ss.

:Wom

an a

rdf

s:Cl

ass;

rdf

s:su

bCla

ssO

f :P

erso

n .

:Wom

an a

rdf

s:Cl

ass;

rdf

s:su

bCla

ssO

f :P

erso

n .

:Eve

a :

Wom

an;

age

“100

”.:E

ve a

:W

oman

; ag

e “1

00”.

:sis

ter

a rd

f:Pr

oper

ty.

:sis

ter

a rd

f:Pr

oper

ty.

:sis

ter

rdfs

:dom

ain

:Per

son;

rdf

s:ra

nge

:Wom

an.

:sis

ter

rdfs

:dom

ain

:Per

son;

rdf

s:ra

nge

:Wom

an.

UM

BC

UM

BC

an H

onor

s U

nive

rsity

in M

aryl

and

an H

onor

s U

nive

rsity

in M

aryl

and

3838

RD

F Sc

hem

a (R

DFS

)R

DF

Sch

ema

(RD

FS)

RD

F Sc

hem

a ad

ds

RD

F Sc

hem

a ad

ds

taxo

nom

ies

for

taxo

nom

ies

for

clas

ses

& p

rope

rtie

scl

asse

s &

pro

pert

ies

subC

lass

and

sub

Prop

erty

subC

lass

and

sub

Prop

erty

and

som

e m

etad

ata.

and

som

e m

etad

ata.

dom

ain

and

rang

edo

mai

n an

d ra

nge

cons

trai

nts

on p

rope

rtie

sco

nstr

aint

s on

pro

pert

ies

Seve

ral w

idel

y u

sed

Seve

ral w

idel

y u

sed

KB

too

ls c

an im

port

KB

too

ls c

an im

port

and

expo

rt in

RD

FSan

d ex

port

in R

DFS

Stan

ford

Pro

tégé

KB

edi

tor

Stan

ford

Pro

tégé

KB

edi

tor

•Ja

va, o

pen

sour

ced

•ex

tens

ible

, lot

s of

plu

g-in

s•

prov

ides

rea

soni

ng &

ser

ver

capa

bilit

ies

UM

BC

UM

BC

an H

onor

s U

nive

rsity

in M

aryl

and

an H

onor

s U

nive

rsity

in M

aryl

and

3939

New

and

Im

prov

ed!

100%

Bet

ter

than

XM

L!!

New

and

N

ew a

nd

Impr

oved

!Im

prov

ed!

100%

Bet

ter

100%

Bet

ter

than

XM

L!!

than

XM

L!!

RD

FS s

upp

orts

sim

ple

infe

ren

ces

RD

FS s

upp

orts

sim

ple

infe

ren

ces

An R

DF

onto

logy

plu

s so

me

RD

FAn

RD

F on

tolo

gy p

lus

som

e RD

Fst

atem

ents

may

impl

y ad

ditio

nal R

DF

stat

emen

ts.

stat

emen

ts m

ay im

ply

addi

tiona

l RD

F st

atem

ents

.Th

is is

not

tru

e of

XM

L.Th

is is

not

tru

e of

XM

L.Ex

ampl

e:Ex

ampl

e:do

mai

n(pa

rent

,per

son)

dom

ain(

pare

nt,p

erso

n)ra

nge(

pare

nt,p

erso

n)ra

nge(

pare

nt,p

erso

n)su

bpro

pert

y(m

othe

r,pa

rent

)su

bpro

pert

y(m

othe

r,pa

rent

)ra

nge(

mot

her,

wom

an)

rang

e(m

othe

r,w

oman

)m

othe

r(ev

e,ca

in)

mot

her(

eve,

cain

)

This

is

This

is p

art

of t

he

data

mod

elpa

rt o

f th

e da

ta m

odel

and

not

of t

he

and

not

of t

he

acce

ssin

g/pr

oces

sing

cod

eac

cess

ing/

proc

essi

ng c

ode

Impl

ies:

subc

lass

(wom

an,p

erso

n)pa

rent

(eve

,cai

n)pe

rson

(eve

)pe

rson

(cai

n)w

oman

(eve

)

Impl

ies:

Impl

ies:

subc

lass

(wom

an,p

erso

n)pa

rent

(eve

,cai

n)pe

rson

(eve

)pe

rson

(cai

n)w

oman

(eve

)

ontologyinstance

UM

BC

UM

BC

an H

onor

s U

nive

rsity

in M

aryl

and

an H

onor

s U

nive

rsity

in M

aryl

and

4040

RD

F is

bei

ng

alre

ady

in u

seR

DF

is b

ein

g al

read

y in

use

RD

F h

as a

sol

id s

peci

fica

tion

RD

F h

as a

sol

id s

peci

fica

tion

See

the

RD

F m

odel

the

ory

spec

-ht

tp:/

/ww

w.w

3.or

g/TR

/rdf

-mt/

RD

F is

bei

ng

use

d in

a n

um

ber

of W

3C

spe

cifi

cati

ons

RD

F is

bei

ng

use

d in

a n

um

ber

of W

3C

spe

cifi

cati

ons

CC/P

P (C

ompo

site

Cap

abili

ties/

Pref

eren

ce P

rofil

es)

http

://w

ww

.w3.

org/

Mob

ile/C

CPP/

P3P

(Pla

tfor

m f

or P

rivac

y Pr

efer

ence

s Pr

ojec

t)

http

://w

ww

.w3.

org/

P3P/

An

d in

oth

er w

eb s

tan

dard

sA

nd

in o

ther

web

sta

nda

rds

RSS

1.0

(RD

F Si

te S

umm

ary)

RD

F ca

lend

ar (

~ iC

alen

dar

in R

DF)

An

d in

oth

er s

yste

ms

An

d in

oth

er s

yste

ms

Net

scap

e’s

Moz

illa

web

bro

wse

rO

pen

dire

ctor

y (h

ttp:

//dm

oz.o

rg/)

Adob

e pr

oduc

ts v

ia X

MP

(eXt

ensi

ble

Met

adat

a Pl

atfo

rm)

Page 11: The Semantic web as part of thefinin/talks/swsw02/rutgers.pdf · 2002. 10. 24. · Next: overview of Semantic Web UMBC an Honors University in Maryland 11 W3C’s Semantic Web Goals

NSF

IDM

Wor

ksho

p M

ay 6

, 200

2Ti

m F

inin

, mai

lto:fi

nin@

umbc

.edu 11

UM

BC

UM

BC

an H

onor

s U

nive

rsity

in M

aryl

and

an H

onor

s U

nive

rsity

in M

aryl

and

4141

RD

F is

not

en

ough

, bu

t is

a g

ood

fou

nda

tion

RD

F is

not

en

ough

, bu

t is

a g

ood

fou

nda

tion

RD

F la

cks

expr

essi

ve a

dequ

acy

for

man

y ta

sks

RD

F la

cks

expr

essi

ve a

dequ

acy

for

man

y ta

sks

Onl

y ra

nge/

dom

ain

cons

trai

nts

(on

prop

ertie

s)O

nly

rang

e/do

mai

n co

nstr

aint

s (o

n pr

oper

ties)

No

prop

ertie

s of

pro

pert

ies

(tra

nsiti

ve, i

nver

se e

tc.)

No

prop

ertie

s of

pro

pert

ies

(tra

nsiti

ve, i

nver

se e

tc.)

No

equi

vale

nce,

dis

join

tnes

s, c

over

ings

, etc

.N

o eq

uiva

lenc

e, d

isjo

intn

ess,

cov

erin

gs, e

tc.

No

nece

ssar

y an

d su

ffic

ient

con

ditio

nsN

o ne

cess

ary

and

suff

icie

nt c

ondi

tions

No

rule

s, a

xiom

s, lo

gica

l con

stra

ints

No

rule

s, a

xiom

s, lo

gica

l con

stra

ints

DAM

L+O

IL e

xten

ds R

DF

DAM

L+O

IL e

xten

ds R

DF

Laye

ring

mak

es p

artia

l kno

wle

dge

avai

labl

e to

La

yerin

g m

akes

par

tial k

now

ledg

e av

aila

ble

to

appl

icat

ions

whi

ch o

nly

unde

rsta

nd R

DF

appl

icat

ions

whi

ch o

nly

unde

rsta

nd R

DF

NB:

Bui

ldin

g on

RD

F ha

s so

me

NB:

Bui

ldin

g on

RD

F ha

s so

me

disa

dvan

tage

sdi

sadv

anta

ges

XML

RD

F

DAM

L+O

IL

UM

BC

UM

BC

an H

onor

s U

nive

rsity

in M

aryl

and

an H

onor

s U

nive

rsity

in M

aryl

and

4242

We’

re g

oin

g do

wn

a f

amili

ar r

oad

We’

re g

oin

g do

wn

a f

amili

ar r

oad

KR

tre

nds

KR

tre

nds

5555-- 6

5: a

rbitr

ary

data

65

: ar

bitr

ary

data

st

ruct

ures

stru

ctur

es6565

-- 75:

sem

antic

net

wor

ks75

: se

man

tic n

etw

orks

7575-- 8

5: s

impl

e fr

ame

85:

sim

ple

fram

e sy

stem

ssy

stem

s8585

-- 95:

des

crip

tion

logi

cs95

: de

scrip

tion

logi

cs9595

-- ??:

logi

c?, r

ules

???

: lo

gic?

, rul

es?

Web

tre

nds

Web

tre

nds

9595-- 9

7:

97:

XML

XML

as a

rbitr

ary

as a

rbitr

ary

stru

ctur

esst

ruct

ures

9797-- 9

8:

98:

RD

FRD

F9898

-- 99:

99

: RD

FSRD

FS(s

chem

a) a

s (s

chem

a) a

s a

fram

ea

fram

e --lik

e sy

stem

like

syst

em0000

-- 01:

01

: D

AML+

OIL

DAM

L+O

IL0202

-- ??:

??

: O

WL,

O

WL,

…??

...

On

ly m

uch

fas

ter!

UM

BC

UM

BC

an H

onor

s U

nive

rsity

in M

aryl

and

an H

onor

s U

nive

rsity

in M

aryl

and

4343

DA

ML+

OIL

as

a Se

man

tic

Web

Lan

guag

eD

AM

L+O

IL a

s a

Sem

anti

c W

eb L

angu

age

DAM

L =

Dar

pa A

gent

Mar

kup

Lang

uage

DAM

L =

Dar

pa A

gent

Mar

kup

Lang

uage

DAR

PA p

rogr

am w

ith 1

7 pr

ojec

ts &

an

inte

grat

or

DAR

PA p

rogr

am w

ith 1

7 pr

ojec

ts &

an

inte

grat

or

deve

lopi

ng la

ngua

ge s

pec,

too

ls, a

pplic

atio

ns f

or S

W.

deve

lopi

ng la

ngua

ge s

pec,

too

ls, a

pplic

atio

ns f

or S

W.

OIL

= O

ntol

ogy

Infe

renc

e La

yer

OIL

= O

ntol

ogy

Infe

renc

e La

yer

An E

U e

ffor

t ai

med

at

deve

lopi

ng a

laye

red

appr

oach

to

An E

U e

ffor

t ai

med

at

deve

lopi

ng a

laye

red

appr

oach

to

repr

esen

ting

know

ledg

e on

the

web

.re

pres

entin

g kn

owle

dge

on t

he w

eb.

Proc

ess

Proc

ess

Join

t Co

mm

ittee

: U

S D

AML

and

EU S

eman

tic W

eb

Join

t Co

mm

ittee

: U

S D

AML

and

EU S

eman

tic W

eb

Tech

nolo

gies

par

ticip

ants

Te

chno

logi

es p

artic

ipan

ts

DAM

L+O

IL s

pecs

rel

ease

d 01

/01

& 0

3/01

DAM

L+O

IL s

pecs

rel

ease

d 01

/01

& 0

3/01

See

See

http

://w

ww

.dam

l.org

/ht

tp:/

/ww

w.d

aml.o

rg/

Incl

udes

mod

el t

heor

etic

and

axi

omat

ic

Incl

udes

mod

el t

heor

etic

and

axi

omat

ic

sem

antic

sse

man

tics

DA

ML+

OIL

UM

BC

UM

BC

an H

onor

s U

nive

rsity

in M

aryl

and

an H

onor

s U

nive

rsity

in M

aryl

and

4444

A S

impl

e D

AM

L Ex

ampl

eA

Sim

ple

DA

ML

Exam

ple

<rd

fs:C

lass

abo

ut=

"#An

imal

"/>

<rd

fs:C

lass

abo

ut=

"#An

imal

"/>

<rd

fs:C

lass

abo

ut=

"#Pl

ant"

><

rdfs

:Cla

ss a

bout

="#

Plan

t">

<da

ml:d

isjo

intF

rom

<da

ml:d

isjo

intF

rom

reso

urce

="#

Anim

al"/

>re

sour

ce=

"#An

imal

"/>

</

</ r

dfs:

Clas

srd

fs:C

lass

>>N

ote

the

mix

ture

of

RD

F (

Not

e th

e m

ixtu

re o

f RD

F ( p

lant

& a

nim

al a

re

plan

t & a

nim

al a

re

clas

ses)

cl

asse

s) a

nd D

AML

and

DAM

L(p

lant

& a

nim

al a

re d

isjo

int)

(pla

nt &

ani

mal

are

dis

join

t)If

you

r ce

ll ph

one

only

doe

s RD

F, it

stil

l If

you

r ce

ll ph

one

only

doe

s RD

F, it

stil

l un

ders

tand

s so

me

of th

is…

unde

rsta

nds

som

e of

this

Page 12: The Semantic web as part of thefinin/talks/swsw02/rutgers.pdf · 2002. 10. 24. · Next: overview of Semantic Web UMBC an Honors University in Maryland 11 W3C’s Semantic Web Goals

NSF

IDM

Wor

ksho

p M

ay 6

, 200

2Ti

m F

inin

, mai

lto:fi

nin@

umbc

.edu 12

UM

BC

UM

BC

an H

onor

s U

nive

rsity

in M

aryl

and

an H

onor

s U

nive

rsity

in M

aryl

and

4545

DA

ML+

OIL

D

AM

L+O

IL ↔↔

RD

FR

DF

DAM

L+O

IL o

ntol

ogy

is a

set

of

RD

F st

atem

ents

DAM

L+O

IL o

ntol

ogy

is a

set

of

RD

F st

atem

ents

DAM

L+O

IL d

efin

es s

eman

tics

for

cert

ain

stat

emen

tsD

AML+

OIL

def

ines

sem

antic

s fo

r ce

rtai

n st

atem

ents

Doe

s D

oes

NO

TN

OT

rest

rict

wha

t ca

n be

sai

dre

stric

t w

hat

can

be s

aid

Ont

olog

y ca

n in

clud

e ar

bitr

ary

RD

FO

ntol

ogy

can

incl

ude

arbi

trar

y RD

F

But

no s

eman

tics

for

non

But

no s

eman

tics

for

non --

DAM

L+O

IL s

tate

men

tsD

AML+

OIL

sta

tem

ents

Adds

cap

abili

ties

com

mon

to

desc

riptio

n lo

gics

:Ad

ds c

apab

ilitie

s co

mm

on t

o de

scrip

tion

logi

cs:

card

inal

ity c

onst

rain

ts, d

efin

ed c

lass

es (

=>

cla

ssifi

catio

n),

card

inal

ity c

onst

rain

ts, d

efin

ed c

lass

es (

=>

cla

ssifi

catio

n),

equi

vale

nce,

loca

l res

tric

tions

, dis

join

t cl

asse

s, e

tc.

equi

vale

nce,

loca

l res

tric

tions

, dis

join

t cl

asse

s, e

tc.

Mor

e su

ppor

t fo

r on

tolo

gies

Mor

e su

ppor

t fo

r on

tolo

gies

Ont

olog

y im

port

s on

tolo

gyO

ntol

ogy

impo

rts

onto

logy

But

not

(yet

) va

riabl

es, q

uant

ifica

tion,

and

Bu

t no

t (y

et)

varia

bles

, qua

ntifi

catio

n, a

nd

gene

ral r

ules

gene

ral r

ules

UM

BC

UM

BC

an H

onor

s U

nive

rsity

in M

aryl

and

an H

onor

s U

nive

rsity

in M

aryl

and

4646

DA

ML

in O

ne

Slid

eD

AM

L in

On

e Sl

ide

<rdf

:RD

F xm

lns:

rdf =

"http

://w

3.or

g/22

<rdf

:RD

F xm

lns:

rdf =

"http

://w

3.or

g/22 --

rdfrdf --syn

tax

synta

x --ns#

"ns#

"xm

lns:

rdfs

="ht

tp://

w3.

org/

rdf

xmln

s:rd

fs="

http

://w

3.or

g/rd

f --sch

ema#

"sch

ema#

"xm

lns:

dam

l="h

ttp://

dam

l.org

/dam

l+oi

l#“>

xmln

s:da

ml=

"http

://da

ml.o

rg/d

aml+

oil#

“><d

aml:O

ntol

ogy

rdf:a

bout

="">

<dam

l:Ont

olog

y rd

f:abo

ut="

"><d

aml:i

mpo

rts rd

f:res

ourc

e="h

ttp://

dam

l.org

/dam

l+oi

l"/>

<dam

l:im

ports

rdf:r

esou

rce=

"http

://da

ml.o

rg/d

aml+

oil"

/></

dam

l:Ont

olog

y></

dam

l:Ont

olog

y><r

dfs:

Cla

ss rd

f:ID

="Pe

rson

"><r

dfs:

Cla

ss rd

f:ID

="Pe

rson

"><r

dfs:

subC

lass

Of r

df:re

sour

ce="

#Ani

mal

"/>

<rdf

s:su

bCla

ssO

f rdf

:reso

urce

="#A

nim

al"/

><r

dfs:

subC

lass

Of>

<rdf

s:su

bCla

ssO

f><d

aml:R

estri

ctio

n><d

aml:R

estri

ctio

n><d

aml:o

nPro

perty

rdf:r

esou

rce=

"#ha

sPar

ent"

/><d

aml:o

nPro

perty

rdf:r

esou

rce=

"#ha

sPar

ent"

/><d

aml:t

oCla

ss rd

f:res

ourc

e="#

Pers

on"/

><d

aml:t

oCla

ss rd

f:res

ourc

e="#

Pers

on"/

></

dam

l:Res

trict

ion>

</da

ml:R

estri

ctio

n></

rdfs

:sub

Cla

ssO

f></

rdfs

:sub

Cla

ssO

f><r

dfs:

subC

lass

Of>

<rdf

s:su

bCla

ssO

f><d

aml:R

estri

ctio

n da

ml:c

ardi

nalit

y="1

"><d

aml:R

estri

ctio

n da

ml:c

ardi

nalit

y="1

"><d

aml:o

nPro

perty

rdf:r

esou

rce=

"#ha

sFat

her"

/><d

aml:o

nPro

perty

rdf:r

esou

rce=

"#ha

sFat

her"

/></

dam

l:Res

trict

ion>

</rd

fs:s

ubC

lass

Of>

</rd

fs:C

lass

></

dam

l:Res

trict

ion>

</rd

fs:s

ubC

lass

Of>

</rd

fs:C

lass

><P

erso

n rd

f:abo

ut=“

http

://um

bc.e

du/~

finin

/">

<Per

son

rdf:a

bout

=“ht

tp://

umbc

.edu

/~fin

in/"

><r

dfs:

com

men

t>Fi

nin

is a

per

son.

</rd

fs:c

omm

ent>

<rdf

s:co

mm

ent>

Fini

n is

a p

erso

n.</

rdfs

:com

men

t></

Pers

on>

</Pe

rson

>

DA

ML

is b

uilt

on to

p of

X

ML

and

RD

F

It ca

n be

use

d to

add

m

etad

ata

abou

t any

thin

g w

hich

has

a U

RI.

ever

ythi

ng h

as U

RI

DA

ML

is ~

= a

fram

e ba

sed

know

ledg

e re

pres

enta

tion

lang

uage

It al

low

s the

def

initi

on,

shar

ing,

com

posi

tion

and

use

of o

ntol

ogie

s

UR

Is a

re a

W3C

stan

dard

ge

nera

lizin

g U

RLs

UM

BC

UM

BC

an H

onor

s U

nive

rsity

in M

aryl

and

an H

onor

s U

nive

rsity

in M

aryl

and

4747

DA

ML

DA

ML --

SS

DAM

LD

AML --

S is

an

onto

logy

for

des

crib

ing

prop

ertie

s S

is a

n on

tolo

gy f

or d

escr

ibin

g pr

oper

ties

and

capa

bilit

ies

of w

eb s

ervi

ces

and

capa

bilit

ies

of w

eb s

ervi

ces

Des

ider

ata:

Des

ider

ata:

Ease

of

expr

essi

vene

ssEa

se o

f ex

pres

sive

ness

Enab

les

auto

mat

ion

of s

ervi

ce u

se b

y ag

ents

Enab

les

auto

mat

ion

of s

ervi

ce u

se b

y ag

ents

Enab

les

reas

onin

g ab

out

serv

ice

prop

ertie

s an

d En

able

s re

ason

ing

abou

t se

rvic

e pr

oper

ties

and

capa

bilit

ies

capa

bilit

ies

Also

app

ropr

iate

for

des

crib

ing

serv

ices

in a

Al

so a

ppro

pria

te f

or d

escr

ibin

g se

rvic

es in

a

mob

ile/p

erva

sive

com

putin

g en

viro

nmen

tm

obile

/per

vasi

ve c

ompu

ting

envi

ronm

ent

See

http

://d

aml.o

rg/s

ervi

ces/

See

http

://d

aml.o

rg/s

ervi

ces/

UM

BC

UM

BC

an H

onor

s U

nive

rsity

in M

aryl

and

an H

onor

s U

nive

rsity

in M

aryl

and

4848

DA

ML

DA

ML --

S com

pon

ents

S com

pon

ents

Serv

ice

prof

ile (

wh

at it

doe

s)Se

rvic

e pr

ofile

(w

hat

it d

oes)

For s

ervi

ce re

gist

ratio

n, d

isco

very

and

mat

chin

g.Fo

r ser

vice

regi

stra

tion,

dis

cove

ry a

nd m

atch

ing.

Hig

hH

igh --

leve

l des

crip

tion

of s

ervi

ce a

nd p

rovi

der w

ith a

(hum

an re

adab

lele

vel d

escr

iptio

n of

ser

vice

and

pro

vide

r with

a (h

uman

read

able

) ) de

scrip

tion

of s

ervi

ce, a

spe

cific

atio

n of

func

tiona

litie

s pr

ovi

desc

riptio

n of

ser

vice

, a s

peci

ficat

ion

of fu

nctio

nalit

ies

prov

i ded

and

oth

er

ded

and

othe

r fu

nctio

nal a

ttrib

utes

.fu

nctio

nal a

ttrib

utes

.Fu

nctio

nal p

rope

rties

sup

port

com

posi

tion:

inpu

ts, o

utpu

ts, p

rec

Func

tiona

l pro

perti

es s

uppo

rt co

mpo

sitio

n: in

puts

, out

puts

, pre

c ond

ition

s on

ditio

ns

and

effe

cts.

an

d ef

fect

s.

Serv

ice

mod

el (h

ow it

wor

ks)

Serv

ice

mod

el (h

ow it

wor

ks)

For s

ervi

ce in

voca

tion,

com

posi

tion,

inte

rope

ratio

n, m

onito

ring,

For s

ervi

ce in

voca

tion,

com

posi

tion,

inte

rope

ratio

n, m

onito

ring,

……C

ompo

site

pro

cess

es a

re b

uild

usi

ng s

eque

nce,

ifC

ompo

site

pro

cess

es a

re b

uild

usi

ng s

eque

nce,

if-- th

enth

en-- e

lse,

fork

, etc

.el

se, f

ork,

etc

.

Serv

ice

grou

ndin

g (h

ow to

acc

ess)

Serv

ice

grou

ndin

g (h

ow to

acc

ess)

Spe

cific

atio

n of

ser

vice

acc

ess

info

rmat

ion

(com

mun

icat

ion

prot

oS

peci

ficat

ion

of s

ervi

ce a

cces

s in

form

atio

n (c

omm

unic

atio

n pr

oto c

ols,

co

ls,

trans

port

mec

hani

sms,

etc

.) w

hich

cou

ld b

e vi

a S

OA

P, H

TTP

form

s,tra

nspo

rt m

echa

nism

s, e

tc.)

whi

ch c

ould

be

via

SO

AP

, HTT

P fo

rms,

Java

Ja

va

RM

I, R

PC, e

tc.

RM

I, R

PC, e

tc.

Page 13: The Semantic web as part of thefinin/talks/swsw02/rutgers.pdf · 2002. 10. 24. · Next: overview of Semantic Web UMBC an Honors University in Maryland 11 W3C’s Semantic Web Goals

NSF

IDM

Wor

ksho

p M

ay 6

, 200

2Ti

m F

inin

, mai

lto:fi

nin@

umbc

.edu 13

UM

BC

UM

BC

an H

onor

s U

nive

rsity

in M

aryl

and

an H

onor

s U

nive

rsity

in M

aryl

and

4949

Tru

st?

Tru

st?

UM

BC

UM

BC

an H

onor

s U

nive

rsity

in M

aryl

and

an H

onor

s U

nive

rsity

in M

aryl

and

5050

W3

C W

eb O

nto

logy

W3

C W

eb O

nto

logy

Wor

kin

g G

rou

pW

orki

ng

Gro

up

The

WO

WG

is w

orki

ng o

n a

reco

mm

enda

tion

The

WO

WG

is w

orki

ng o

n a

reco

mm

enda

tion

for

the

"Web

Ont

olog

y La

ngua

ge":

fo

r th

e "W

eb O

ntol

ogy

Lang

uage

": O

WL

OW

L 56

Mem

bers

fro

m 3

0 W

3C O

rgan

izat

ions

56

Mem

bers

fro

m 3

0 W

3C O

rgan

izat

ions

C

ompa

nie

s:C

ompa

nie

s:Ag

fa, D

aim

ler

Agfa

, Dai

mle

r --Ch

rysl

er, E

DS,

Fuj

itsu,

Hew

lett

Chry

sler

, ED

S, F

ujits

u, H

ewle

tt-- P

acka

rd,

Pack

ard,

IB

M, I

ntel

, IVI

S, L

ucen

t, N

etw

ork

Infe

renc

e, N

isus

, Nok

ia, P

hili

IBM

, Int

el, I

VIS,

Luc

ent,

Net

wor

k In

fere

nce,

Nis

us, N

okia

, Phi

li ps,

ps

, St

ilo, S

un, U

nisy

s St

ilo, S

un, U

nisy

s P

ubl

ic S

ecto

r:P

ubl

ic S

ecto

r:D

ISA,

Ele

ctric

ite d

e Fr

ance

, Int

elin

k, I

NTA

P, M

ITRE,

D

ISA,

Ele

ctric

ite d

e Fr

ance

, Int

elin

k, I

NTA

P, M

ITRE,

N

IST

NIS

T R

esea

rch

pro

ject

s/La

bs:

Res

earc

h p

roje

cts/

Labs

:D

FKI,

FZI

, Ibr

ow g

roup

, Sta

nfor

d, U

. D

FKI,

FZI

, Ibr

ow g

roup

, Sta

nfor

d, U

. Br

isto

l, U

. Mar

ylan

d, U

. Sou

thha

mpt

on

Bris

tol,

U. M

aryl

and,

U. S

outh

ham

pton

In

vite

d Ex

pert

s:In

vite

d Ex

pert

s:M

edic

al, D

igita

l Lib

rary

, Def

ense

, Tec

hnic

al

Med

ical

, Dig

ital L

ibra

ry, D

efen

se, T

echn

ical

CoCh

airs

: Ji

m H

endl

er, U

nive

rsity

of

Mar

ylan

d/M

IND

; G

uus

CoCh

airs

: Ji

m H

endl

er, U

nive

rsity

of

Mar

ylan

d/M

IND

; G

uus

Schr

eibe

r, U

niv

of A

mst

erda

m/I

brow

Schr

eibe

r, U

niv

of A

mst

erda

m/I

brow

http

://w

ww

.w3.

org/

2001

/sw

/Web

Ont

/ht

tp:/

/ww

w.w

3.or

g/20

01/s

w/W

ebO

nt/

UM

BC

UM

BC

an H

onor

s U

nive

rsity

in M

aryl

and

an H

onor

s U

nive

rsity

in M

aryl

and

5151

OW

L G

oals

OW

L G

oals

The

WO

WG

has

iden

tifie

d th

e Th

e W

OW

G h

as id

entif

ied

the

follo

win

g go

als

in d

evel

opin

g O

WL

follo

win

g go

als

in d

evel

opin

g O

WL

Shar

ed o

ntol

ogie

s Sh

ared

ont

olog

ies

Ont

olog

y ev

olut

ion

Ont

olog

y ev

olut

ion

Ont

olog

y in

tero

pera

bilit

y O

ntol

ogy

inte

rope

rabi

lity

Inco

nsis

tenc

y de

tect

ion

Inco

nsis

tenc

y de

tect

ion

Bala

nce

of e

xpre

ssiv

ity a

nd s

cala

bilit

y Ba

lanc

e of

exp

ress

ivity

and

sca

labi

lity

Ease

of

use

Ease

of

use

XML

synt

ax

XML

synt

ax

Inte

rnat

iona

lizat

ion

Inte

rnat

iona

lizat

ion

UM

BC

UM

BC

an H

onor

s U

nive

rsity

in M

aryl

and

an H

onor

s U

nive

rsity

in M

aryl

and

5252

OW

L st

atu

s an

d pu

blic

atio

ns

OW

L st

atu

s an

d pu

blic

atio

ns

OW

L is

rou

ghly

equ

ival

ent

to D

AML

with

som

e O

WL

is r

ough

ly e

quiv

alen

t to

DAM

L w

ith s

ome

rena

min

g of

pro

pert

ies

rena

min

g of

pro

pert

ies

Curr

ent

plan

is t

o ha

ve t

hree

com

plia

nce

leve

ls:

Curr

ent

plan

is t

o ha

ve t

hree

com

plia

nce

leve

ls:

OW

L lit

e, O

WL,

OW

L pl

usO

WL

lite,

OW

L, O

WL

plus

Web

Ont

has

pub

lishe

d W

ebO

nt h

as p

ublis

hed

Req

uire

men

tsReq

uire

men

tsfo

r a

Web

Ont

olog

y La

ngua

ge

for

a W

eb O

ntol

ogy

Lang

uage

Fe

atur

e Sy

nops

is

Feat

ure

Syno

psis

for

OW

L Li

te a

nd O

WL

for

OW

L Li

te a

nd O

WL

OW

L W

eb O

ntol

ogy

Lang

uage

1.0

O

WL

Web

Ont

olog

y La

ngua

ge 1

.0 R

efer

ence

Ref

eren

ceO

WL

Web

Ont

olog

y La

ngua

ge 1

.0

OW

L W

eb O

ntol

ogy

Lang

uage

1.0

Abs

trac

t Sy

ntax

Ab

stra

ct S

ynta

x (f

orth

com

ing)

(f

orth

com

ing)

OW

L G

uide

O

WL

Gui

de

Page 14: The Semantic web as part of thefinin/talks/swsw02/rutgers.pdf · 2002. 10. 24. · Next: overview of Semantic Web UMBC an Honors University in Maryland 11 W3C’s Semantic Web Goals

NSF

IDM

Wor

ksho

p M

ay 6

, 200

2Ti

m F

inin

, mai

lto:fi

nin@

umbc

.edu 14

UM

BC

UM

BC

an H

onor

s U

nive

rsity

in M

aryl

and

an H

onor

s U

nive

rsity

in M

aryl

and

5353

OW

L Li

teO

WL

Lite

RD

F Sc

hem

a Fe

atu

res:

RD

F Sc

hem

a Fe

atu

res:

Clas

sCl

ass

rdf:

Prop

erty

rdf:

Prop

erty

rdfs

:sub

Clas

sOf

rdfs

:sub

Clas

sOf

rdfs

:sub

Prop

erty

Of

rdfs

:sub

Prop

erty

Of

rdfs

:dom

ain

rdfs

:dom

ain

rdfs

:ran

gerd

fs:r

ange

Indi

vidu

alIn

divi

dual

Equ

alit

y pr

oper

ties

:Eq

ual

ity

prop

erti

es:

sam

eCla

ssAs

sam

eCla

ssAs

sam

ePro

pert

yAs

sam

ePro

pert

yAs

sam

eInd

ivid

ualA

ssa

meI

ndiv

idua

lAs

diff

eren

tInd

ivid

ualF

rom

diff

eren

tInd

ivid

ualF

rom

Pro

pert

ies

of p

rope

rtie

s:P

rope

rtie

s of

pro

pert

ies:

inve

rseO

fin

vers

eOf

tran

sitit

iveP

rope

rty

tran

sitit

iveP

rope

rty

sym

met

ricPr

opoe

rty

sym

met

ricPr

opoe

rty

func

tiona

lPro

pert

yfu

nctio

nalP

rope

rty

inve

rseF

unct

iona

lPro

pert

yin

vers

eFun

ctio

nalP

rope

rty

allV

alue

sFro

mal

lVal

uesF

rom

som

eVal

uesF

rom

som

eVal

uesF

rom

min

Card

inal

ity (

0/1)

min

Card

inal

ity (

0/1)

max

Card

inal

ity (

0/1)

max

Card

inal

ity (

0/1)

card

inal

ity (

0/1)

card

inal

ity (

0/1)

Hea

der

Info

rmat

ion

:H

eade

r In

form

atio

n:

impo

rts

impo

rts

Dub

lin C

ore

Met

adat

aD

ublin

Cor

e M

etad

ata

vers

ionI

nfo

vers

ionI

nfo

Mis

sin

g:en

umer

ated

cla

sses

, dis

join

tnes

s, u

nion

Of,

inte

rsec

tionO

fco

mpl

emen

tOf,

full

card

inal

ity, …

Mot

ivat

ion:

eas

ier

to im

plem

ent

and

to le

arn,

low

er r

easo

ning

com

plex

ity

UM

BC

UM

BC

an H

onor

s U

nive

rsity

in M

aryl

and

an H

onor

s U

nive

rsity

in M

aryl

and

5454

KR

mee

ts t

he

Web

KR

mee

ts t

he

Web

One

way

to

thin

k ab

out

the

sem

antic

One

way

to

thin

k ab

out

the

sem

antic

web

is t

hat

we

are

crea

ting

a kn

owle

dge

web

is t

hat

we

are

crea

ting

a kn

owle

dge

repr

esen

tatio

n la

ngua

ge f

or t

he W

eb.

repr

esen

tatio

n la

ngua

ge f

or t

he W

eb.

This

is m

ore

than

just

sel

ectin

g an

app

ropr

iate

Th

is is

mor

e th

an ju

st s

elec

ting

an a

ppro

pria

te

KR la

ngua

ge a

nd s

elec

ting

an X

ML

enco

ding

.KR

lang

uage

and

sel

ectin

g an

XM

L en

codi

ng.

The

Web

as

an in

form

atio

n sy

stem

has

man

y Th

e W

eb a

s an

info

rmat

ion

syst

em h

as m

any

sign

ifica

nt p

rope

rtie

s.si

gnifi

cant

pro

pert

ies.

Hig

hly

dist

ribut

edSu

bjec

t to

dis

conn

ectio

ns a

nd o

ther

fai

lure

sM

any

cont

ent

prov

ider

sPa

rtia

l and

inco

nsis

tent

info

rmat

ion

Not

all

info

and

ser

vice

s ca

n be

tru

sted

Dyn

amic

Evol

ving

UM

BC

UM

BC

an H

onor

s U

nive

rsity

in M

aryl

and

an H

onor

s U

nive

rsity

in M

aryl

and

5555

Sem

anti

c W

eb P

rinc

iple

sSe

man

tic

Web

Pri

ncip

les

Ever

ythi

ng is

on

the

web

Ever

ythi

ng is

on

the

web

Peop

le, p

lace

s, t

imes

, thi

ngs

all h

ave

URIs

Peop

le, p

lace

s, t

imes

, thi

ngs

all h

ave

URIs

Part

ial i

nfor

mat

ion

is a

ssum

edPa

rtia

l inf

orm

atio

n is

ass

umed

The

web

priv

ilege

s sc

alab

ility

ove

r in

tegr

ity a

nd t

here

’s a

lway

sTh

e w

eb p

rivile

ges

scal

abili

ty o

ver

inte

grity

and

the

re’s

alw

ays

mor

e an

d ne

w s

tuff

to

find

mor

e an

d ne

w s

tuff

to

find

Trus

t m

odel

s ar

e cr

itica

lTr

ust

mod

els

are

criti

cal

It’s

not

all

true

It’s

not

all

true

Supp

ort

info

rmat

ion

evol

utio

nSu

ppor

t in

form

atio

n ev

olut

ion

Cont

ent

and

cons

ensu

s is

dyn

amic

Cont

ent

and

cons

ensu

s is

dyn

amic

Min

imal

ist

desi

gnM

inim

alis

t de

sign

Mak

e th

e si

mpl

e th

ings

sim

ple,

and

the

com

plex

thi

ngs

poss

ible

. M

ake

the

sim

ple

thin

gs s

impl

e, a

nd t

he c

ompl

ex t

hing

s po

ssib

le.

Stan

dard

ize

no m

ore

than

is n

eces

sary

.St

anda

rdiz

e no

mor

e th

an is

nec

essa

ry.

Com

mon

dat

a m

odel

Com

mon

dat

a m

odel

To s

uppo

rt in

tero

pera

bilit

y an

d kn

owle

dge

shar

ing

To s

uppo

rt in

tero

pera

bilit

y an

d kn

owle

dge

shar

ing

Adap

ted

from

Eric

Mill

er, W

3CU

MB

CU

MB

Can

Hon

ors

Uni

vers

ity in

Mar

ylan

dan

Hon

ors

Uni

vers

ity in

Mar

ylan

d5656

SW is

wor

k in

pro

gres

sSW

is w

ork

in p

rogr

ess

Ther

e ar

e im

port

ant

lang

uage

asp

ects

whi

ch

Ther

e ar

e im

port

ant

lang

uage

asp

ects

whi

ch

need

mor

e w

ork:

rul

es, q

uerie

s, e

tc.

need

mor

e w

ork:

rul

es, q

uerie

s, e

tc.

Man

y to

ols

need

to

be c

reat

ed, e

.g.,

Man

y to

ols

need

to

be c

reat

ed, e

.g.,

Prot

égé

plug

Prot

égé

plug

-- in

for

DAM

L+O

ILin

for

DAM

L+O

ILAn

nota

tion

tool

s An

nota

tion

tool

s

Appl

icat

ions

nee

d to

be

expl

ored

Appl

icat

ions

nee

d to

be

expl

ored

The

W3C

is d

evel

opin

g a

new

SW

lang

uage

The

W3C

is d

evel

opin

g a

new

SW

lang

uage

OW

L: O

ntol

ogy

Web

Lan

guag

eO

WL:

Ont

olog

y W

eb L

angu

age

SW id

eas

will

mig

rate

into

oth

er s

tand

ards

(e.

g.,

SW id

eas

will

mig

rate

into

oth

er s

tand

ards

(e.

g.,

basi

c XM

L, W

SDL,

.NET

)ba

sic

XML,

WSD

L, .N

ET)

Page 15: The Semantic web as part of thefinin/talks/swsw02/rutgers.pdf · 2002. 10. 24. · Next: overview of Semantic Web UMBC an Honors University in Maryland 11 W3C’s Semantic Web Goals

NSF

IDM

Wor

ksho

p M

ay 6

, 200

2Ti

m F

inin

, mai

lto:fi

nin@

umbc

.edu 15

UM

BC

UM

BC

an H

onor

s U

nive

rsity

in M

aryl

and

an H

onor

s U

nive

rsity

in M

aryl

and

5757

DA

ML+

OIL

usa

geD

AM

L+O

IL u

sage

DAM

L+O

IL is

alre

ady

the

mos

t us

ed o

ntol

ogy/

KR

DAM

L+O

IL is

alre

ady

the

mos

t us

ed o

ntol

ogy/

KR

lang

uage

in h

isto

ryla

ngua

ge in

his

tory

Dam

l.org

~ 5

.4M

hits

: av

g. 2

4,30

0/da

y in

Oct

02.

Dam

l.org

~ 5

.4M

hits

: av

g. 2

4,30

0/da

y in

Oct

02.

1.8

x 10

1.8

x 10

18(1

80,0

00 G

b) d

ownl

oade

d(1

80,0

00 G

b) d

ownl

oade

dO

ct 1

6: C

raw

ler

finds

5.9

M D

AML

stat

emen

ts o

n 20

K w

eb p

ages

Oct

16:

Cra

wle

r fin

ds 5

.9M

DAM

L st

atem

ents

on

20K

web

pag

esD

oesn

't in

clud

e m

any

inst

ance

KBs

tie

d to

ont

olog

ies

or m

any

Doe

sn't

incl

ude

man

y in

stan

ce K

Bs t

ied

to o

ntol

ogie

s or

man

y ve

ry la

rge

RD

FSve

ry la

rge

RD

FS-- b

ased

KBs

tha

t in

clud

e so

me

OW

Lba

sed

KBs

that

incl

ude

som

e O

WL

OW

L is

mov

ing

it to

war

ds t

he c

omm

erci

al w

orld

OW

L is

mov

ing

it to

war

ds t

he c

omm

erci

al w

orld

Web

too

l dev

elop

er la

bs:

IBM

, HP,

Sun

, Int

el, F

ujits

uW

eb t

ool d

evel

oper

labs

: IB

M, H

P, S

un, I

ntel

, Fuj

itsu

Cont

ent

prov

ider

s/us

ers:

Dai

mle

rCo

nten

t pr

ovid

ers/

user

s: D

aim

ler --

Chry

sler

, Nok

ia, M

otor

ola,

ED

S,

Chry

sler

, Nok

ia, M

otor

ola,

ED

S,

Agfa

Agfa

Star

ting

to b

e no

ticed

by

thes

auru

s di

strib

utor

s St

artin

g to

be

notic

ed b

y th

esau

rus

dist

ribut

ors

----e.

g.,

e.g.

, N

atio

nal

Nat

iona

lCa

ncer

Ins

titut

e m

etat

hesa

urus

to

be r

elea

sed

in O

WL

Canc

er I

nstit

ute

met

athe

saur

us t

o be

rel

ease

d in

OW

L

UM

BC

UM

BC

an H

onor

s U

nive

rsity

in M

aryl

and

an H

onor

s U

nive

rsity

in M

aryl

and

5858

Lots

of

Ope

n I

ssu

esLo

ts o

f O

pen

Iss

ues

How

exp

ress

ive

shou

ld t

he K

R la

ngua

ge b

e?H

ow e

xpre

ssiv

e sh

ould

the

KR la

ngua

ge b

e?W

hat

kind

of

KR/r

easo

ning

sys

tem

W

hat

kind

of

KR/r

easo

ning

sys

tem

F.

O. l

ogic

, fuz

zy, …

F.O

. log

ic, f

uzzy

, …

On

Web

Ont

olog

ies

On

Web

Ont

olog

ies

One

(e.

g. C

YC)

or m

any

(DAM

L)O

ne (

e.g.

CYC

) or

man

y (D

AML)

If m

any,

com

posa

ble

(IEE

E IF

F) o

r m

onol

ithic

(IE

EE S

UM

O)

If m

any,

com

posa

ble

(IEE

E IF

F) o

r m

onol

ithic

(IE

EE S

UM

O)

Will

gen

eral

“up

per

onto

logi

es”

(e.g

., IE

EE S

UO

) be

use

ful?

Will

gen

eral

“up

per

onto

logi

es”

(e.g

., IE

EE S

UO

) be

use

ful?

Will

indu

stry

buy

in?

Will

indu

stry

buy

in?

Or

cont

inue

to

expl

ore

ad h

oc X

ML

base

d so

lutio

nsO

r co

ntin

ue t

o ex

plor

e ad

hoc

XM

L ba

sed

solu

tions

How

will

it b

e us

ed?

How

will

it b

e us

ed?

As m

arku

p? As

alte

rnat

ive

cont

ent?

Ju

st b

oth

mac

hine

s an

d As

mar

kup?

As

alte

rnat

ive

cont

ent?

Ju

st b

oth

mac

hine

s an

d pe

ople

?pe

ople

?

=>

On

ly e

xper

imen

tati

on w

ill y

ield

an

swer

s.=

> O

nly

exp

erim

enta

tion

will

yie

ld a

nsw

ers.

?

UM

BC

UM

BC

an H

onor

s U

nive

rsity

in M

aryl

and

an H

onor

s U

nive

rsity

in M

aryl

and

5959

Con

clu

sion

s an

d fi

nal

th

ough

tsC

oncl

usi

ons

and

fin

al t

hou

ghts

SW m

ight

be

a ch

ance

for

us

to g

et s

ome

SW m

ight

be

a ch

ance

for

us

to g

et s

ome

AI o

ut o

f th

e la

bAI

out

of

the

lab

Solv

ing

the

sym

bol g

roun

ding

pro

blem

Solv

ing

the

sym

bol g

roun

ding

pro

blem

Ret

hink

ing

agen

t co

mm

unic

atio

nRet

hink

ing

agen

t co

mm

unic

atio

nH

ow d

o w

e ge

t th

ere

How

do

we

get

ther

e

UM

BC

UM

BC

an H

onor

s U

nive

rsity

in M

aryl

and

an H

onor

s U

nive

rsity

in M

aryl

and

6060

The

sym

bol g

rou

ndi

ng

prob

lem

The

sym

bol g

rou

ndi

ng

prob

lem

An a

rgum

ent

agai

nst

hum

anAn

arg

umen

t ag

ains

t hu

man

-- lik

e AI

is

like

AI is

th

at it

’s im

poss

ible

unl

ess

mac

hine

sth

at it

’s im

poss

ible

unl

ess

mac

hine

ssh

are

our

perc

eptio

n of

the

wor

ld.

shar

e ou

r pe

rcep

tion

of t

he w

orld

.A

solu

tion

to t

his

“sym

bol g

roun

ding

A so

lutio

n to

thi

s “s

ymbo

l gro

undi

ngpr

oble

m”

is t

o gi

ve r

obot

s w

ith h

uman

prob

lem

” is

to

give

rob

ots

with

hum

anin

spire

d se

nses

.in

spire

d se

nses

.Bu

t th

e w

orld

we

expe

rienc

e is

det

erm

ined

by

our

sens

es,

But

the

wor

ld w

e ex

perie

nce

is d

eter

min

ed b

y ou

r se

nses

, an

d hu

man

and

mac

hine

bod

ies

may

lead

to

diff

eren

t an

d hu

man

and

mac

hine

bod

ies

may

lead

to

diff

eren

t co

ncep

tions

of

the

wor

ld (

e.g.

Nag

el’s

co

ncep

tions

of

the

wor

ld (

e.g.

Nag

el’s

Wha

t Is

It L

ike

To

Wha

t Is

It L

ike

To

Be a

Bat

?Be

a B

at?

))M

aybe

the

Sem

antic

Web

is a

way

out

of

this

pro

blem

?M

aybe

the

Sem

antic

Web

is a

way

out

of

this

pro

blem

?

MIT

’s C

og

Page 16: The Semantic web as part of thefinin/talks/swsw02/rutgers.pdf · 2002. 10. 24. · Next: overview of Semantic Web UMBC an Honors University in Maryland 11 W3C’s Semantic Web Goals

NSF

IDM

Wor

ksho

p M

ay 6

, 200

2Ti

m F

inin

, mai

lto:fi

nin@

umbc

.edu 16

UM

BC

UM

BC

an H

onor

s U

nive

rsity

in M

aryl

and

an H

onor

s U

nive

rsity

in M

aryl

and

6161

Solv

ing

the

sym

bol g

rou

ndi

ng

prob

lem

Solv

ing

the

sym

bol g

rou

ndi

ng

prob

lem

The

web

may

bec

ome

a co

mm

on w

orld

tha

t bo

th

The

web

may

bec

ome

a co

mm

on w

orld

tha

t bo

th

hum

ans

and

mac

hine

s ca

n un

ders

tand

.hu

man

s an

d m

achi

nes

can

unde

rsta

nd.

Conf

essi

on:

the

web

is m

ore

fam

iliar

and

rea

l to

Conf

essi

on:

the

web

is m

ore

fam

iliar

and

rea

l to

me

than

muc

h of

the

rea

l wor

ld.

me

than

muc

h of

the

rea

l wor

ld.

Phys

ical

obj

ects

can

be

tagg

ed w

ith lo

w c

ost

Phys

ical

obj

ects

can

be

tagg

ed w

ith lo

w c

ost

(e.g

., $0

.05)

tra

nspo

nder

s or

RFI

Ds

enco

ding

(e

.g.,

$0.0

5) t

rans

pond

ers

or R

FID

s en

codi

ng

thei

r U

RIs

thei

r U

RIs

See

HP’

s Co

olto

wn

proj

ect

See

HP’

s Co

olto

wn

proj

ect

http

://c

oolto

wn.

com

/ht

tp:/

/coo

ltow

n.co

m/

UM

BC

UM

BC

an H

onor

s U

nive

rsity

in M

aryl

and

an H

onor

s U

nive

rsity

in M

aryl

and

6262

Ret

hin

kin

g th

e ag

ent

com

mu

nic

atio

n p

arad

igm

Ret

hin

kin

g th

e ag

ent

com

mu

nic

atio

n p

arad

igm

Muc

h m

ulti

Muc

h m

ulti --

agen

t sy

stem

s w

ork

is g

roun

ded

in

agen

t sy

stem

s w

ork

is g

roun

ded

in

Agen

t Co

mm

unic

atio

n La

ngua

ges

(e.g

.,

Agen

t Co

mm

unic

atio

n La

ngua

ges

(e.g

., K

QM

LKQ

ML ,

, FI

PAFI

PA)

and

asso

ciat

ed s

oftw

are

infr

astr

uctu

re.

) an

d as

soci

ated

sof

twar

e in

fras

truc

ture

.Th

is p

arad

igm

was

art

icul

ated

~19

90, a

bout

the

sam

e Th

is p

arad

igm

was

art

icul

ated

~19

90, a

bout

the

sam

e tim

e as

the

WW

W w

as d

evel

oped

.tim

e as

the

WW

W w

as d

evel

oped

.O

ur M

AS a

ppro

ach

has

not

yet

left

the

labo

rato

ry y

et

Our

MAS

app

roac

h ha

s no

t ye

t le

ft t

he la

bora

tory

yet

th

e W

eb h

as c

hang

ed t

he w

orld

.th

e W

eb h

as c

hang

ed t

he w

orld

.

May

be w

e sh

ould

try

som

ethi

ng d

iffer

ent?

May

be w

e sh

ould

try

som

ethi

ng d

iffer

ent?

The

com

mun

icat

ion

MAS

par

adig

m h

as b

een

peer

The

com

mun

icat

ion

MAS

par

adig

m h

as b

een

peer

-- toto--

peer

mes

sage

orie

nted

com

mun

icat

ion

med

iate

d by

pe

er m

essa

ge o

rient

ed c

omm

unic

atio

n m

edia

ted

by

brok

ers

and

faci

litat

ors

brok

ers

and

faci

litat

ors

----an

app

roac

h in

herit

ed f

rom

an

app

roac

h in

herit

ed f

rom

cl

ient

clie

nt-- s

erve

r sy

stem

s.se

rver

sys

tem

s.

UM

BC

UM

BC

an H

onor

s U

nive

rsity

in M

aryl

and

an H

onor

s U

nive

rsity

in M

aryl

and

6363

Ret

hin

kin

g th

e ag

ent

com

mu

nic

atio

n p

arad

igm

Ret

hin

kin

g th

e ag

ent

com

mu

nic

atio

n p

arad

igm

A po

ssib

le n

ew p

arad

igm

?A

poss

ible

new

par

adig

m?

Agen

ts “

publ

ish”

bel

iefs

, req

uest

s, a

nd o

ther

Ag

ents

“pu

blis

h” b

elie

fs, r

eque

sts,

and

oth

er

“spe

ech

acts

” on

web

pag

es.

“spe

ech

acts

” on

web

pag

es.

Brok

ers

“sea

rch”

for

and

“in

dex”

pub

lishe

d Br

oker

s “s

earc

h” f

or a

nd “

inde

x” p

ublis

hed

cont

ent

cont

ent

Agen

ts “

disc

over

” w

hat

peer

s ha

ve p

ublis

hed

on

Agen

ts “

disc

over

” w

hat

peer

s ha

ve p

ublis

hed

on

the

web

and

bro

wse

for

mor

e de

tails

the

web

and

bro

wse

for

mor

e de

tails

Agen

ts “

spea

k fo

r” c

onte

nt o

n w

eb p

ages

by

Agen

ts “

spea

k fo

r” c

onte

nt o

n w

eb p

ages

by

Answ

erin

g qu

erie

s ab

out

them

Answ

erin

g qu

erie

s ab

out

them

Acce

ptin

g co

mm

ents

and

ass

ertio

ns a

bout

the

mAc

cept

ing

com

men

ts a

nd a

sser

tions

abo

ut t

hem

UM

BC

UM

BC

an H

onor

s U

nive

rsity

in M

aryl

and

an H

onor

s U

nive

rsity

in M

aryl

and

6464

How

do

we

get

ther

e fr

om h

ere?

How

do

we

get

ther

e fr

om h

ere?

This

sem

antic

web

em

phas

izes

ont

olog

ies

This

sem

antic

web

em

phas

izes

ont

olog

ies

––th

eir

thei

r de

velo

pmen

t, u

se, m

edia

tion,

evo

lutio

n, e

tc.

deve

lopm

ent,

use

, med

iatio

n, e

volu

tion,

etc

.It

will

tak

e so

me

time

to r

eally

del

iver

on

the

It w

ill t

ake

som

e tim

e to

rea

lly d

eliv

er o

n th

e ag

ent

para

digm

, eith

er o

n th

e In

tern

et o

r in

a

agen

t pa

radi

gm, e

ither

on

the

Inte

rnet

or

in a

pe

rvas

ive

com

putin

g en

viro

nmen

t.pe

rvas

ive

com

putin

g en

viro

nmen

t.Th

e de

velo

pmen

t of

com

plex

sys

tem

s is

bas

ical

ly

The

deve

lopm

ent

of c

ompl

ex s

yste

ms

is b

asic

ally

an

evo

lutio

nary

pro

cess

.an

evo

lutio

nary

pro

cess

.Ran

dom

sea

rch

carr

ied

out

by t

ens

of t

hous

ands

Ran

dom

sea

rch

carr

ied

out

by t

ens

of t

hous

ands

of

res

earc

hers

, dev

elop

ers

and

grad

uate

of

res

earc

hers

, dev

elop

ers

and

grad

uate

st

uden

ts.

stud

ents

.

Page 17: The Semantic web as part of thefinin/talks/swsw02/rutgers.pdf · 2002. 10. 24. · Next: overview of Semantic Web UMBC an Honors University in Maryland 11 W3C’s Semantic Web Goals

NSF

IDM

Wor

ksho

p M

ay 6

, 200

2Ti

m F

inin

, mai

lto:fi

nin@

umbc

.edu 17

UM

BC

UM

BC

an H

onor

s U

nive

rsity

in M

aryl

and

an H

onor

s U

nive

rsity

in M

aryl

and

6565C

limbi

ng

Mou

nt

Impr

obab

leC

limbi

ng

Mou

nt

Impr

obab

le

““ The

she

er h

eigh

t of t

he

The

shee

r he

ight

of t

he

peak

doe

sn't

mat

ter,

so

peak

doe

sn't

mat

ter,

so

long

as

you

don'

t try

to

long

as

you

don'

t try

to

scal

e it

in a

sin

gle

boun

d.

scal

e it

in a

sin

gle

boun

d.

Loca

te th

e m

ildly

slo

ping

Lo

cate

the

mild

ly s

lopi

ng

path

and

, if y

ou h

ave

path

and

, if y

ou h

ave

unlim

ited

time,

the

asce

nt

unlim

ited

time,

the

asce

nt

is o

nly

as fo

rmid

able

as

the

is o

nly

as fo

rmid

able

as

the

next

ste

p.”

next

ste

p.”

----Ric

hard

Ric

hard

D

awki

ns, C

limbi

ng M

ount

D

awki

ns, C

limbi

ng M

ount

Im

prob

able

, Pen

guin

Im

prob

able

, Pen

guin

Bo

oks,

199

6.Bo

oks,

199

6.

UM

BC

UM

BC

an H

onor

s U

nive

rsity

in M

aryl

and

an H

onor

s U

nive

rsity

in M

aryl

and

6666

The

Evol

uti

on o

f U

sefu

l Th

ings

The

Evol

uti

on o

f U

sefu

l Th

ings

The

Evol

utio

n of

Use

ful

The

Evol

utio

n of

Use

ful

Thin

gsTh

ings

, Hen

ry P

etro

ski,

1994

., H

enry

Pet

rosk

i, 19

94.

Prio

r to

the

189

0’s,

pap

ers

Prio

r to

the

189

0’s,

pap

ers

wer

e he

ld t

oget

her

with

w

ere

held

tog

ethe

r w

ith

stra

ight

pen

s.st

raig

ht p

ens.

The

deve

lopm

ent

of “

sprin

g Th

e de

velo

pmen

t of

“sp

ring

stee

l” a

llow

ed t

he in

vent

ion

of

stee

l” a

llow

ed t

he in

vent

ion

of

the

pape

r cl

ip in

189

9.th

e pa

per

clip

in 1

899.

It t

ook

abou

t It

too

k ab

out

25 y

ears

(!)

25 y

ears

(!)

for

for

the

evol

utio

n of

the

mod

ern

the

evol

utio

n of

the

mod

ern

“gem

pap

ercl

ip”,

con

side

red

to

“gem

pap

ercl

ip”,

con

side

red

to

be o

ptim

al f

or g

ener

al u

se.

be o

ptim

al f

or g

ener

al u

se.

UM

BC

UM

BC

an H

onor

s U

nive

rsity

in M

aryl

and

an H

onor

s U

nive

rsity

in M

aryl

and

6767

So, w

e sh

ould

…So

, we

shou

ld …

Star

t w

ith

the

sim

ple

and

mov

e to

war

d th

e co

mpl

exSt

art

wit

h t

he s

impl

e an

d m

ove

tow

ard

the

com

plex

E.g.

, fro

m v

ocab

ular

ies

to F

OL

theo

ries

E.g.

, fro

m v

ocab

ular

ies

to F

OL

theo

ries

Dev

elop

new

cap

abili

ties

Dev

elop

new

cap

abili

ties

E.g.

, rul

es, t

rust

, neg

otia

tion,

aut

omat

ic m

arku

p, …

E.g.

, rul

es, t

rust

, neg

otia

tion,

aut

omat

ic m

arku

p, …

Allo

w m

any

onto

logi

es t

o bl

oom

Allo

w m

any

onto

logi

es t

o bl

oom

Let

natu

ral e

volu

tiona

ry p

roce

sses

sel

ect

cons

ensu

s Le

t na

tura

l evo

lutio

nary

pro

cess

es s

elec

t co

nsen

sus

onto

logi

es.

onto

logi

es.

Supp

ort

dive

rsit

y in

on

tolo

gies

Supp

ort

dive

rsit

y in

on

tolo

gies

Mon

ocul

ture

s ar

e un

stab

le, t

here

sho

uld

be n

o M

onoc

ultu

res

are

unst

able

, the

re s

houl

d be

no

““ TheTh

eon

tolo

gy fo

r X”

onto

logy

for

X”..

The

evol

uti

on o

f po

wer

ful,

mac

hin

e re

adab

le

The

evol

uti

on o

f po

wer

ful,

mac

hin

e re

adab

le

onto

logi

es w

ill t

ake

man

y ye

ars,

may

be g

ener

atio

ns

onto

logi

es w

ill t

ake

man

y ye

ars,

may

be g

ener

atio

ns

But

incr

emen

tal b

enef

its w

ill e

asily

just

ify t

he e

ffor

tBu

t in

crem

enta

l ben

efits

will

eas

ily ju

stify

the

eff

ort

UM

BC

UM

BC

an H

onor

s U

nive

rsity

in M

aryl

and

an H

onor

s U

nive

rsity

in M

aryl

and

6868

For

mor

e in

form

atio

nFo

r m

ore

info

rmat

ion

RDF

RDF

http

://w

ww

.w3.

org/

RD

F/ht

tp:/

/ww

w.w

3.or

g/RD

F/

DAM

L+O

ILD

AML+

OIL

http

://w

ww

.dam

l.org

/ht

tp:/

/ww

w.d

aml.o

rg/

OW

L &

W3C

’s s

eman

tic w

eb a

ctiv

ityO

WL

& W

3C’s

sem

antic

web

act

ivity

http

://w

ww

.w3.

org/

2001

/sw

/ht

tp:/

/ww

w.w

3.or

g/20

01/s

w/

Sem

antic

web

link

sSe

man

tic w

eb li

nks

http

://s

eman

ticw

eb.o

rg/

http

://s

eman

ticw

eb.o

rg/

Nex

t Se

man

tic W

eb C

onfe

renc

eN

ext

Sem

antic

Web

Con

fere

nce

http

://

http

:// i

swc.

sem

antic

web

.org

isw

c.se

man

ticw

eb.o

rg//

Oct

ober

200

3, S

anib

el I

slan

d, S

C.O

ctob

er 2

003,

San

ibel

Isl

and,

SC.

Page 18: The Semantic web as part of thefinin/talks/swsw02/rutgers.pdf · 2002. 10. 24. · Next: overview of Semantic Web UMBC an Honors University in Maryland 11 W3C’s Semantic Web Goals

NSF

IDM

Wor

ksho

p M

ay 6

, 200

2Ti

m F

inin

, mai

lto:fi

nin@

umbc

.edu 18

UM

BC

UM

BC

an H

onor

s U

nive

rsity

in M

aryl

and

an H

onor

s U

nive

rsity

in M

aryl

and

6969