RDFS, Taxonomy, and Ontology

Chapter 4: Building the Foundation of the Semantic Web

Eng. Dr. Tiroshan Madushanka

Why Do We Need RDFS?

The Missing Vocabulary

The Problem with RDF Alone

While RDF can describe resources in a structured way, it lacks a critical component: vocabulary definition.

                    Key Questions RDF Cannot Answer:
                    Where is the class SLR defined? What does it look like?
Are there superclasses or subclasses of SLR?
What properties can be defined for this class?
What relationships exist between classes?

                

RDFS: The Solution

RDFS provides a language to create vocabularies that define classes, subclasses, properties, and their relationships—making the web truly machine-processable at a global level.

What is RDFS?

RDF Schema Explained

RDFS Definition

A vocabulary description language for RDF resources
Defines classes, subclasses, and properties
Associates properties with classes they describe
Adds semantics by specifying property meanings and value types
Written in RDF using the same triple/graph model

The Great News!

RDFS is written in RDF, so it's not as scary as you might think. It uses the same familiar structure you already know.

Key Insight

RDFS can be viewed as an extension of RDF. Together, RDF + RDFS push the Internet one step further toward machine-readability—a step that cannot be accomplished by RDF alone.

RDFS + RDF = Enhanced Reasoning

One More Step Toward Machine-Readability

A Simple Vocabulary Example

Consider a camera vocabulary with the following structure:

📷 Camera

💾 Digital

🎞️ Film

🔭 SLR

📸 Point-and-Shoot

Interactive Reasoning Example

Given:

Nikon D70 is an instance of SLR
SLR is a subclass of Digital
Digital is a subclass of Camera

Machine Can Conclude:

Nikon D70 is also a Digital camera
Nikon D70 is also a Camera
Search for "Digital" should include Nikon D70

                    Key Point: The machine performs this reasoning automatically by combining RDF triples with the RDFS vocabulary structure!
                

Core Elements of RDFS

The Building Blocks

Core Classes

rdfs:Resource - The root class of all classes
rdfs:Class - The class of classes
rdf:Property - The class of properties
rdfs:Datatype - The class of datatypes

Core Properties

rdfs:subClassOf - Defines class hierarchy
rdfs:subPropertyOf - Defines property hierarchy

Core Constraints

rdfs:domain - Specifies which class a property describes
rdfs:range - Specifies the type of values a property can have

Defining Classes in RDFS

Creating Your Vocabulary

Basic Class Definition

<?xml version="1.0"?>
<rdf:RDF xmlns:rdf="http://www.w3.org/1999/02/22-rdf-syntax-ns#"
         xmlns:rdfs="http://www.w3.org/2000/01/rdf-schema#"
         xml:base="http://example.org/camera#">

    <rdfs:Class rdf:ID="Camera">
    </rdfs:Class>

    <rdfs:Class rdf:ID="Digital">
        <rdfs:subClassOf rdf:resource="#Camera"/>
    </rdfs:Class>

    <rdfs:Class rdf:ID="SLR">
        <rdfs:subClassOf rdf:resource="#Digital"/>
    </rdfs:Class>

</rdf:RDF>

                    Key Points:
                    Use rdfs:Class to define a new class
Use rdf:ID to name the class
Use rdfs:subClassOf to create class hierarchies
Classes without subClassOf are direct subclasses of rdfs:Resource

                

Defining Properties in RDFS

Describing Class Characteristics

Property Definition with Domain and Range

<rdf:Property rdf:ID="owned_by">
    <rdfs:domain rdf:resource="#SLR"/>
    <rdfs:range rdf:resource="#Photographer"/>
</rdf:Property>

<rdf:Property rdf:ID="has_spec">
    <rdfs:domain rdf:resource="#SLR"/>
    <rdfs:range rdf:resource="#Specifications"/>
</rdf:Property>

                    Understanding Domain and Range
                    rdfs:domain - Specifies which class this property can describe
rdfs:range - Specifies what type of values the property can have

                

Interpretation

Property owned_by:

Can only be used to describe instances of SLR class
Its value must be an instance of Photographer class

RDFS vs XML Schema

Different Purposes, Different Goals

XML Schema

Validates XML documents
Ensures syntax is legal
Defines allowed structure and data types
About validation and syntax
Document-centric

RDF Schema

Provides vocabulary for RDF
Defines classes and relationships
Associates properties with classes
About semantics and meaning
Knowledge-centric

Key Distinction

XML Schema: "Is this document syntactically correct?"

RDFS: "What does this mean, and how do concepts relate?"

RDFS is About Semantics

Semantics (meaning) is expressed by:

Specifying properties and their value types
Following the resource-property-value structure
Enabling machines to understand relationships

What is an Ontology?

From Vocabulary to Knowledge Representation

W3C Definition

"An ontology defines the terms used to describe and represent an area of knowledge."

Key Characteristics of Ontologies

Domain-specific - Represents a specific area of knowledge
Terms and relationships - Defines classes and how they relate
Hierarchical structure - Superclasses and subclasses
Properties - Describe features and attributes
Machine-understandable - Structured for computational processing

Examples of Domains

Photography, Medicine, Real Estate, Education, E-commerce, Biology, Music, etc.

Camera Ontology Structure:

Classes: Camera, Digital, Film, SLR, Point-and-Shoot, Photographer
Properties: owned_by, has_spec, pixel, model
Relationships: Hierarchies, ownership, specifications

Ontology vs Taxonomy

Understanding the Difference

Taxonomy

Classification-focused
Defines classes and subclasses
Hierarchical relationships
No property definitions
Simpler structure

Ontology

Classification + Properties
Defines classes and their properties
Multiple relationship types
Property constraints (domain/range)
Richer, more expressive

Simple Rule

Taxonomy: "What belongs where?" (Classification only)

Ontology: "What is it, what properties does it have, and how does it relate to others?"

Example

If you remove all properties from the camera ontology, leaving only classes and subclass relationships, you have a taxonomy, not an ontology.

Reasoning Power of RDFS

Simple, Yet Powerful

Four Main Reasoning Mechanisms

Domain inference - Determine resource type from property domain
Range inference - Determine value type from property range
Class hierarchy - Infer superclass memberships
Property hierarchy - Infer from subPropertyOf relationships

Interactive Example

Given RDF Statement:

NikonD70 owned_by LiyangYu

Machine Reasoning:

Fact: owned_by has domain SLR
Conclusion: NikonD70 is an SLR
Fact: SLR is subclass of Digital
Conclusion: NikonD70 is Digital
Fact: owned_by has range Photographer
Conclusion: LiyangYu is a Photographer
Fact: Photographer is subclass of Person
Conclusion: LiyangYu is a Person

                    Key Insight: All this reasoning happens automatically by following the resource-property-value structure!
                

Benefits of Ontologies

Why Build Them?

Major Benefits

Shared understanding - Common definitions of key concepts
Knowledge reuse - Share and extend domain knowledge
Explicit assumptions - Make domain assumptions clear
Machine understanding - Encode semantics for computers
Automatic processing - Enable large-scale intelligent agents

Real-World Impact

Ontologies enable:

Smarter search engines that understand meaning
Automatic data integration across systems
Intelligent assistants that reason about information
Semantic interoperability between applications

The Ultimate Goal

Make distributed information on the Internet machine-friendly and machine-processable, enabling a true Semantic Web where computers can understand, reason about, and act on information automatically.

Limitations of RDFS

Good, Better, Best - More is Needed

RDFS is Powerful, But...

There are gaps that prevent even richer ontologies and reasoning:

What RDFS Cannot Express

Class equivalence - Cannot say "SLR" and "Single-Lens-Reflex" are the same
Cardinality constraints - Cannot specify "exactly one pixel value"
Property characteristics - Cannot define transitive, symmetric properties
Class operations - Cannot create unions, intersections, complements
Disjointness - Cannot say "Digital and Film are mutually exclusive"
Property restrictions - Limited ability to constrain property values

The Solution: OWL

Web Ontology Language (OWL) extends RDFS to provide:

Richer vocabulary for complex relationships
More expressive constraints
Enhanced reasoning capabilities

Key Takeaways

Essential Concepts to Remember

Core Concepts

RDFS provides vocabulary that RDF alone cannot
Ontologies encode knowledge in machine-understandable ways
Semantics through structure - Resource-property-value enables reasoning
Hierarchies matter - Subclass/subproperty relationships enable inference
Domain and range are key to automatic type inference

The Big Picture

RDF: Describes resources in structured triples

RDFS: Defines vocabulary (classes, properties, relationships)

Together: Enable machines to understand and reason about web content

Result: One more critical step toward the Semantic Web vision

Next Step

Learn OWL (Web Ontology Language) to express even richer ontologies with advanced reasoning capabilities!

Thank You!

Questions & Discussion

Summary

RDFS extends RDF with vocabulary definition capabilities
Ontologies make domain knowledge explicit and machine-processable
Simple reasoning enables powerful inference on the Semantic Web
OWL provides even more expressive power (coming next!)

Remember

"The Semantic Web is not about making machines think like humans,
but about encoding knowledge so machines can help humans think better."