Digital Libraries

The Semantic Web: Making sense of it all

Week 6 Min-Yen KAN

Motivation for semantic web

“The primary goal is to make the web more like a library and less like a heap of messy books on the floor.”

-Tim Bray, Textwise consultant

The Web can reach its full potential only if it becomes a place where data can be shared and processed by

automated tools as well as by people.

- Semantic Web Activity Statement

So, what is it anyways?

An extension of current the web that allows:

Exchange of data

By software agents

Will allow agents to reason

Needs to be able to seamlessly exchange data

Other examples:
Services off the desktop

Or perhaps on different desktops…

This is great, why didn’t we think of it sooner?!?

As a community, we have been trying:

Doug Lenat spun off a company to try to capture commonsense knowledge in a huge knowledge representation project.

1,000,000 assertions captured.

Cyc knows that trees are usually outdoors, that once people die they stop buying things, and that glasses of liquid should be carried right-side-up.

What’s different about the Semantic Web?

CYC and the Semantic Web

One company

Centralized

First-order logic

Complex KR language

Authoritative data

Many companies

Distributed

Left up to agent

Simple KR language

Noisy data

Implementing SW

For the semantic web to function, computers must have access to:

structured collections of information

and sets of inference rules that they can use to conduct automated reasoning.

Adding logic to the Web — the means to use rules to make inferences

Uses XML and RDF as a framework

Semantic Web problem

Islands of XML from disparate web services

Example : Tori Amos

Up to consumer to put these chunks together

Situation analogous to pre-web hypertext systems and RDBMS today

Semantic Web problem

TAP Goal

Create a coherent semantic web from disparate chunks

Effectively make the web a giant distributed DB

Why --- Bringing the Internet to programs

SW Layer Cake

Resource Description Framework

Encodes knowledge in sets of triples

A document makes assertions that:

particular things (people, Web pages or whatever)

have properties (such as "is a sister of," "is the author of")

with certain values (another person, another Web page).

RDF Model

A model for representing named properties and property values

models the equivalence relation

Simply a triple of the form:

RDF / XML: assertion interchange

Simplified XML Syntax for RDF

Encodes RDF as machine parsable XML

Verbose, not really readable by humans

Note: counter to what one of XML’s primary motivations.

RDF and XML are complementary:

XML only gives structure (validating with a DTD)

RDF adds to XML the ability to encode simple propositions

RDF Schema – Basis for ontology

RDF with XML: encode assertions

Still need to be able to exchange and reason on the data

To build the necessary ontology, RDF Schema was designed to be a simple data typing model for RDF

RDF Schema Core classes,properties,constraints

rdfs:Resource

rdfs:Property

rdfs:Class

rdf:Type

rdfs:subClassOf

rdfs:PropertyOf

rdfs:ConstraintResource

rdfs:ConstraintProperty

rdfs:range

rdfs:domain

RDF Schemas

The first three most important concepts in RDF datatyping schema:

Resource (rdfs:Resource)

are objects that are uniquely identified by an URI

Note: URI not URL. Question: What is a URI?

Property (rdf:Property)

express the relationships of values associated with resources

Class (rdfs:Class)

are resources denoting a set of resources

RDF schema example

Book rdf:type rdfs:Class .

:bookTitle rdf:type rdf:Property .

:bookTitle rdfs:domain :Book .

:bookTitle rdfs:range rdfs:Literal .

:MyBook rdf:type :Book .

:MyBook :bookTitle “My Book”

There’s a type of resource called “Book”

There a type of property called “BookTitle”

“BookTitle”s are a property of “Book”s

… and they can take a literal string value

MyBook is a type of Book

MyBook’s title is “My Book”

What about incompatible schemas? SW’s Answer: OWL

RDF Schema is fine if one person/organization is authoring all of SW

Inconsistencies among different authors

OWL strengthens RDF Schema with some 30 additional interchange properties

References

SW ontology development information (DAML):

http://derpi.tuwien.ac.at/~andrei/daml.htm

Introduction to RDF Schema

http://www.dlib.org/dlib/may98/miller/05miller.html

RDF and RDF schema

www.wastl.net/download/slides/rdf_overview.pdf

OWL

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

To think about…

What are XML namespaces and how do they figure into the RDF syntax?

Minimalist architecture makes the web scalable, will it make the SW workable?

SW is not (yet fully) standardized

Help everyone out and see what you can contribute!

What’s your prediction when the SW will “arrive”?


	“The primary goal is to make the web more like a library and less like a heap of messy books on the floor.”
	-Tim Bray, Textwise consultant

	The Web can reach its full potential only if it becomes a place where data can be shared and processed by
	automated tools as well as by people.
	- Semantic Web Activity Statement


An extension of current the web that allows:

	Exchange of data
		By software agents
		Will allow agents to reason

	Needs to be able to seamlessly exchange data


As a community, we have been trying:

	Doug Lenat spun off a company to try to capture commonsense knowledge in a huge knowledge representation project.
		1,000,000 assertions captured.
		Cyc knows that trees are usually outdoors, that once people die they stop buying things, and that glasses of liquid should be carried right-side-up.

What’s different about the Semantic Web?


	One company
	Centralized
	First-order logic
	Complex KR language
	Authoritative data


	Many companies
	Distributed
	Left up to agent
	Simple KR language
	Noisy data


	For the semantic web to function, computers must have access to:
		structured collections of information
		and sets of inference rules that they can use to conduct automated reasoning.

	Adding logic to the Web — the means to use rules to make inferences

	Uses XML and RDF as a framework


	Islands of XML from disparate web services
	Example : Tori Amos












	Up to consumer to put these chunks together
	Situation analogous to pre-web hypertext systems and RDBMS today


	Create a coherent semantic web from disparate chunks









	Effectively make the web a giant distributed DB
	Why --- Bringing the Internet to programs


	Encodes knowledge in sets of triples

	A document makes assertions that:
	particular things (people, Web pages or whatever)
	have properties (such as "is a sister of," "is the author of")
	with certain values (another person, another Web page).


	A model for representing named properties and property values
		models the equivalence relation
		Simply a triple of the form:


Simplified XML Syntax for RDF
	Encodes RDF as machine parsable XML
	Verbose, not really readable by humans
		Note: counter to what one of XML’s primary motivations.

RDF and XML are complementary:
	XML only gives structure (validating with a DTD)
	RDF adds to XML the ability to encode simple propositions


	RDF with XML: encode assertions
		Still need to be able to exchange and reason on the data

		To build the necessary ontology, RDF Schema was designed to be a simple data typing model for RDF


	rdfs:Resource
	rdfs:Property
	rdfs:Class
	rdf:Type
	rdfs:subClassOf
	rdfs:PropertyOf
	rdfs:ConstraintResource
	rdfs:ConstraintProperty
	rdfs:range
	rdfs:domain


The first three most important concepts in RDF datatyping schema:
	Resource (rdfs:Resource)
		are objects that are uniquely identified by an URI
			Note: URI not URL. Question: What is a URI?
	Property (rdf:Property)
		express the relationships of values associated with resources
	Class (rdfs:Class)
		are resources denoting a set of resources





	Book rdf:type rdfs:Class .
	:bookTitle rdf:type rdf:Property .
	:bookTitle rdfs:domain :Book .
	:bookTitle rdfs:range rdfs:Literal .
	:MyBook rdf:type :Book .
	:MyBook :bookTitle “My Book”
	There’s a type of resource called “Book”
	There a type of property called “BookTitle”
	“BookTitle”s are a property of “Book”s
	… and they can take a literal string value
	MyBook is a type of Book
	MyBook’s title is “My Book”


	RDF Schema is fine if one person/organization is authoring all of SW
		Inconsistencies among different authors
	OWL strengthens RDF Schema with some 30 additional interchange properties


	SW ontology development information (DAML):
		http://derpi.tuwien.ac.at/~andrei/daml.htm
	Introduction to RDF Schema
		http://www.dlib.org/dlib/may98/miller/05miller.html
	RDF and RDF schema
		www.wastl.net/download/slides/rdf_overview.pdf
	OWL
		http://www.w3.org/TR/owl-ref/


	What are XML namespaces and how do they figure into the RDF syntax?

	Minimalist architecture makes the web scalable, will it make the SW workable?
	SW is not (yet fully) standardized
		Help everyone out and see what you can contribute!
	What’s your prediction when the SW will “arrive”?