Semantic Web Presentation

schoolmistInternet and Web Development

Oct 22, 2013 (4 years and 6 months ago)


The W3C’s Semantic Web

Kyle Mosack

The Semantic Web

"The Semantic Web is not a separate Web
but an extension of the current one, in
which information is given well
meaning, better enabling computers and
people to work in cooperation." It is a
source to retrieve information from the web
(using the web spiders from RDF files) and
access the data through Semantic Web
Agents or Semantic Web Services.”

Source: "The Semantic Web" by Tim Berners
Lee, James Hendler, and Ora Lassila, Scientific
American, 2001

What It Could Be

Machine Readable Data View

Having data on web be defined an linked in a way that
can be read by machine for automation, integration,
and re
use across different applications

Intelligent Agent View

Agents retrieve and manipulate pertinent information

Distributed Data View

Provide sufficient flexibility to be able to represent all
databases and logic rules to link them together

Automated Infrastructure View

Current web lacks an easy automation framework

What it Could Be

Servant of Humanity View

Enable Web applications to automatically collect ,
integrate, and process information and interoperate
with other applications

Better Annotation View

Annotations expressed in a machine

and linked together

Improved Searching View

Access Web content by concepts instead of keywords

Web Services View

Expand services from existing web by automating
services with Web agents

Semantic Web Layer Cake

Semantic Web Layer Cake

Readable Description Framework

Designed for specific data about specific

Can represent Data and Metadata

Moves proprietary data to a form
computers can analyze

Recommendation from W3C for
Semantic Web


To be successful RDF must be able to

Describe most kinds of data that will be

Describe structural design of data sets

Describe relationships between bits of data


Creates a data model with triplets

Subject, predicate (property), object (property

These are statements about resources

Identified by URI

Easily convertible

Not limited by predefined database values


RDF Graph with Anonymous Nodes

RDF Applications


Uses XUL (extensible user interface language)

Uses RDF as a source for listings and other control
information that defines which XUL files to use for
specific XUL interfaces

RSS (Really Simple Syndication)

Helps spread summaries of personal blogs cheaply and


XML exchange format

One of the few RDF applications that is distributed
over the Web rather than being used locally, unlike
Mozilla application

Can be seen as an early, maybe primitive, Semantic Web Application

RDF Applications

RDF for annotations:

Experimental scheme by W3C for annotating
Web pages

Bibliographic meta data: Dublin Core

Provides practical standard terms applicable to
nearly any published work

: Fusing Information

DAML (DARPA Agent Markup Language)

Ability to make up web pages so that information
about the page can be extracted in a uniform way
and combined with other web pages

RDF and the Semantic Web

Abilities beyond conventional database

Combine data with different data sets that don’t follow
the same data models

Add data that doesn’t fit the table structures

Exchange data with any other application that can
handle RDF

Use an RDF processor that can do logical reasoning to
discover unstated relationships

Use someone else’s ontology to learn more about own

Add statements about publications and references that
have been defined somewhere else on the Web

Do all these things using well defined standards, so
wide range of applications can process the data


Potential Problems

Many stem from nature of the Web

Incomplete information

Contradictory or unreliable information

Full first order logic requires an ability to
generate general statements about the
whole table

No way to negate statements

May not be powerful enough


Study of existence or being

The kinds of things that can be talked
about in a system or context

Provides the means to classify these

Name and label them

Kinds of organization include

Lists, hierarchies, and trees


The ontology of a complicated semantic
system can capture enough knowledge so
a computer can perform everyday

To define a set of classes that together cover
a domain of interest

Framework provides syntax, vocabulary,
and some pre
defined terms

Framework is an ontology for constructing

Ontology Considerations

Merging Ontology

Terms and classes can by understood by more
than one ontology

Can be accomplished by using same system
(like OWL)

Even if able to merge, inconsistencies could
jeopardize reliability

Imports of second ontology should be kept
simple and be done so in small sections

Merging Ontology

Ontology Considerations

Problems with importing a large

Vocabulary may change over time once
ontology is developed

Already committed to the vocabulary that is

Can be limited by designer but no solutions
have been made available

Ontology Languages

Frameworks with Web
like uses


Resource Description Framework Schema

Base RDF language for describing ontology

Built on top of RDF

RDF makes statements about resources, making assertions
about a subject

Every RDFS statement is a legal RDF Statement

RDF classes and properties

Together with standard classes, possible to give basic
characteristics of classes for an ontology

Not enough power to express many constraints or logical

Ontology Languages

OWL (Web Ontology Language)

Final Recommendation by W3C

Built to standardize a more capable Ontology
Framework than RDFS

Need to restrict Cardinality

Express optionally

Combine classes

Ontology Summary

To be useful for the semantic Web, an ontology
anguage must be:

Able to reference concepts defined elsewhere on

Sharable over the Web

Be able to work with one or more languages

Able to merge several different

Widely accepted as a standard

Expressive enough for serious use

Support logical functions that are needed to conduct
business of the semantic Web

Last two points are questioned abilities of OWL


Uses in the Semantic Web

Applying and evaluating rules

Inferring facts that haven’t been explicitly stated

Explaining why a particular conclusion has been reached

Detecting contradictory statements and claims

Specifying ontology and vocabulary of all kinds

Representing knowledge

Describing the kinds of things that may be said about a subject

How those statements are to be understood

The statement and execution of queries to obtain
information from stores of data

Combining information from distributed sources in a
coherent way


There is considerable risk that an


will absorb contradictory or
incorrect information

Most reasoning systems can not capture

Ones that are able to can not do so in a uniform,
easy to read way

More current versions of RDF define ways
to understand a collection of RDF
statements that can deal with the possibility
of contradictory information

Requires more computing power


Logic and Ontology

Ontology defines concepts and terms

Logic provides ways to make statements that
define the use of concepts and terms

To reason about collections of statements that use
the concepts and terms


Logic and Representing Knowledge

Logic is a formal discipline dealing mostly with
formal language that can express a subset of
everything that can be articulated using
natural language

Formal description of data and information
naturally involves the use of logic



Logical descriptions of information to be
retrieved from a database

Queries will need to operate across
distributed sources of data to be effective in
the Semantic Web

Needs to reconcile the differences in ontology and
deal with problems of contradictory data


Problems of Semantic Web logic looks to
deal with

When trying to decide what data should be
imported from remote database

Size of the knowledge base might be too large and
overwhelm resources

Importing data without duplicating knowledge

How much is interconnected?

Don’t want to import automatically because of a
formation of any new data

Importing unreliable information

Contaminate good data


With development in utility of Web, trust
needs to be established between system
and user


Identity: Who are you?

Why should I trust you?

Who else trusts you?

How much should I trust you?

How do I know that you said what you
claimed you have said?



How much confidence should I place in what
you say?

What should I believe when different facts
don’t agree

How much should my prior beliefs influence
my confidence in what you say?

How can I establish the correct degree of
belief for a given set of information?

Tools of Trust


Private and Public


Special summary of a document or message
that can not be reversed to the original, no
key needed

Slight differences in messages amount to large
differences in digest

Similarity of digests can not be used to
predict similarity in messages

Tools of Trust

Public Key Infrastructure (PKI)

Widespread system

Certificate Authority (CA) issues digital certificate

CA trustworthy source

CA signs certificate for other

Creates chain of certifications, eventually amounting
to a size that alleviates issues of trust

Potential problems

If CA private key compromised, entire chain untrustworthy

Large cross
certified chains can become unmanageable


may have different standards

Human error and fraud

Digital Signatures

Authenticated by CA or CA chain


The Semantic Web

Work Cited

, T. B.
Explorer's guide to the semantic web
. Manning Publications,
2004. Print.