Adaptive Hypermedia From Concepts to Authoring

Adaptive HypermediaFrom Concepts to Authoring

Peter Brusilovsky

School of Information Sciences

University of Pittsburghpeterb@pitt.edu

http://www.sis.pitt.edu/~peterb/

Adaptive systems

Classic loop “user modeling - adaptation” in adaptive systems

Adaptive software systems

• Intelligent Tutoring Systems– adaptive course sequencing– adaptive . . .

• Adaptive Hypermedia Systems– adaptive presentation– adaptive navigation support

• Adaptive IR systems

• Adaptive . . .

Outline

• Adaptive hypermedia– Where? – Why? – What? – How? – Who?...

• Adaptive presentation

• Adaptive navigation support

Adaptive hypermedia: Why?

Different people are differentIndividuals are different at different times"Lost in Hyperspace”Large variety of usersVariable characteristics of the usersLarge hyperspace

Where it can be useful?

• Web-based Education– ITS, tutorials, Web courses

• On-line information systems– classic IS, information kiosks, encyclopedias

• E-commerce• Museums

– virtual museums and handheld guides

• Information retrieval systems– classic IR, filtering, recommendation, services

Where it can be useful?

• Web-based educationELM-ART, AHA!, KBS-Hyperbook, MANIC

• On-line information systemsPEBA-II, AHA!, AVANTI, SWAN, ELFI, ADAPTS

• E-commerceTellim, SETA, Adaptive Catalogs

• Virtual and real museumsILEX, HYPERAUDIO, HIPS, Power, Marble Museum

• Information retrieval, filtering, recommendationSmartGuide, Syskill & Webert, IfWeb, SiteIF, FAB, AIS

Adapting to what?

• Knowledge: about the system and the subject

• Goal: local and global

• Interests

• Background: profession, language, prospect, capabilities

• Navigation history

Who provides adaptation?

• User

• "Administrator"

• System itself

• Adaptive vs. adaptable systems

What can be adapted?

• Hypermedia = Pages + Links

• Adaptive presentation

– content adaptation

• Adaptive navigation support

– link adaptation

Adaptive

hypermedia

technologies

Adaptive

presentation

Adaptive

navigation support

Direct guidance

Adaptive link

sorting

Adaptive link

hiding

Adaptive link

annotation

Adaptive link

generation

Adaptive

multimedia

presentation

Adaptive text

presentation

Adaptation of

modality

Canned text

adaptation

Natural

language

adaptation

Inserting/

removing

fragments

Altering

fragments

Stretchtext

Sorting

fragments

Dimming

fragments

Map adaptation

Hiding

Disabling

Removal

Adaptive presentation: goals

• Provide the different content for users with different knowledge, goals, background

• Provide additional material for some categories of users– comparisons– extra explanations– details

• Remove or fade irrelevant piece of content• Sort fragments - most relevant first

Adaptive presentation techniques

• Conditional text filteringITEM/IP, PT, AHA!

• Adaptive stretchtextMetaDoc, KN-AHS, PUSH, ADAPTS

• Frame-based adaptationHypadapter, EPIAIM, ARIANNA, SETA

• Full natural language generationILEX, PEBA-II, Ecran Total

Conditional text filtering

If switch is known and user_motivation is high

Fragment 2

Fragment K

Fragment 1

• Similar to UNIX cpp• Universal technology

– Altering fragments– Extra explanation– Extra details– Comparisons

• Low level technology– Text programming

Example: Stretchtext (PUSH)

Example: Stretchtext (ADAPTS)

Adaptive presentation: evaluation

• MetaDoc: On-line documentation system, adapting to user knowledge on the subject

• Reading comprehension time decreased

• Understanding increased for novices

• No effect for navigation time, number of nodes visited, number of operations

Adaptive navigation support: goals

• Guidance: Where I can go? – Local guidance (“next best”)– Global guidance (“ultimate goal”)

• Orientation: Where am I? – Local orientation support (local area)– Global orientation support (whole hyperspace)

Adaptive navigation support

• Direct guidance

• Restricting access– Removing, disabling, hiding

• Sorting

• Annotation

• Generation– Similarity-based, interest-based

• Map adaptation techniques

Example: Adaptive annotation

Annotations for topic states in Manuel Excell: not seen (white lens) ; partially seen (grey lens) ; and completed (black lens)

Adaptive annotation and removing

QuickTime™ and aTIFF (LZW) decompressorare needed to see this picture.

Example: Adaptive annotation

1. Concept role

2. Current concept state

3. Current section state

4. Linked sections state

4

3

2

1

v

Adaptive navigation support: major goals and relevant technologies

Direct

guidance

Sorting Hiding Annotation Map adaptation

Global

guidance

WebWatcher

ITEM/IP

ISIS-Tutor

SHIVA

Adaptive

HyperMan

CID

HYPERFLEX

Local

guidance

Land Use

Tutor

HyperTutor

Adaptive

HyperMan

ELM-PE

Hypadapter

HYPERFLEX

Hypadapter

PUSH

ISIS-Tutor

ELM-ART HYPERCASE

Local

orientation

support

(knowledge)

Hypadapter

ELM-PE

[Clibbon]

HyperTutor

Hypadapter

ISIS-Tutor

ELM-ART

ISIS-Tutor

ITEM/PG

Manuel Excel

Local

orientation

support

(goal)

Hynecosum

HyPLAN

ISIS-Tutor

PUSH

SYPROS

ELM-ART

ISIS-Tutor HYPERCASE

Global

orientation

support

[Clibbon]

Hynecosum

HyperTutor

ISIS-Tutor

SYPROS

ITEM/PG

ISIS-Tutor

ELM-ART

Manuel Excel

HYPERCASE

What can be adapted: links

• Contextual links (“real hypertext”)

• Local non-contextual links

• Index pages

• Table of contents

• Links on local map

• Links on global map

Link types and technologies

Directguidance

Sorting Hiding Annotation Mapadaptation

Contextual links OK (disabling) OK

Non-contextual links OK OK ? OK

Table of contents OK ? OK

Index OK ? OK

Local map OK OK OK OK

Global map OK OK OK OK

Adaptive navigation support: evaluation

• Sorting HYPERFLEX, 1993

• Annotation (colors) and hidingISIS-Tutor, 1995

• Annotation (icons)InterBook, 1997

• HidingDe Bra’s course, 1997

Evaluation of sorting

• HYPERFLEX: IR System– adaptation to user search goal– adaptation to “personal cognitive map”

• Number of visited nodes decreased (significant)

• Correctness increased (not significant)

• Goal adaptation is more effective

• No significant difference for time/topic

Adaptive Hypermedia: our approach

ISIS-Tutor, MSU (1992-1994) ITEM/PG, MSU (1991-1993)

SQL-Tutor, MSU (1995-1998) ELM-ART, Trier (1994-1997)

InterBook, CMU (1996-1998) ELM-ART II, Trier (1997-1998)

ITEM/IP, MSU (1986-1994)

ADAPTS, CMU (1998-1999) COCOA, CTE (1999-2000)

Annotation and hiding: ISIS-Tutor

• An adaptive tutorial for CDS/ISIS/M users• Domain knowledge: concepts and constructs• Hyperspace of learning material:

– Description of concepts and constructs– Examples and problems indexed with concepts

(could be used in an exploratory environment)

• Link annotation with colors and marks• Removing links to “not relevant” pages

Concepts, examples, and problems

Example 2 Example M

Example 1

Problem 1

Problem 2 Problem K

Concept 1

Concept 2

Concept 3

Concept 4

Concept 5

Concept N

Examples

Problems

Concepts

Indexing and navigation

Example 2 Example M

Example 1

Problem 1

Problem 2 Problem K

Concept 1

Concept 2

Concept 3

Concept 4

Concept 5

Concept N

Examples

Problems

Concepts

Student modeling and adaptation

• States for concepts:– not ready (may be hidden)– ready (red)– known (green)– learned (green and ‘+’)

• State for problems/examples:– not ready (may be hidden)– ready (red)– solved (green and ‘+’)

Sample index page (annotation)

Sample index page (hiding)

ISIS-Tutor: Evaluation

• 26 first year CS students of MSU

• 3 groups: – control (no adaptation)– adaptive annotation– adaptive annotation + hiding

• Goal: 10 concepts (of 64), 10 problems, all examples

Experiment design

Educational status of the node

behind the link

A. Non-

adaptive

B. Adaptive

annotation

C. Adaptive

hiding

Outside educational goal NA NA hidden

Within educational goal NA mark "-" mark "-"

Well-learned NA mark "+" mark "+"

Known NA green color green color

Ready-to-be-learned NA red color red color

Not-ready-to-be-learned NA NA hidden

Results: performance

Group Number of steps Time (sec) Concept

repetitions

"Unforced"

concept

repetitions

Task

repetitions

Non-adaptive 81.3 2196 17.3 11.2 6.2

Adaptive 65.2 1418 9.0 5.0 0.8

Restrictive 58.2 1785 8.9 4.8 0.4

Adaptive annotation makes navigation more efficient

The value of adaptivity: steps

30

40

50

60

70

80

90

100

110

120

130

Units

Steps

Hiding

Annotation

Nonadaptive

The value of adaptivity: repetitions

0

2

4

6

8

10

12

14

16

18

Concept repetitions Unforced repetitions

Hiding

Annotation

Nonadaptive

Results: navigation

Group Concept index Concept ->

concept

Concept ->

task

Task index Task ->

concept

Non-adaptive 18.5 8.1 9.8 12.2 2.0

Adaptive 18.4 1.8 6.4 9.6 2.2

Restrictive 14.0 1.6 8.1 7.0 1.8

No effect on navigation patterns due to variety of navigation styles

Results: recall

Group

Recalled concepts

correct

Recalled concepts

incorrect

Recalled links

correct

Recalled links

incorrect

Non-adaptive 7.0 0.7 8.6 5.0

Adaptive 6.2 1.2 7.5 8.5

Restrictive 6.9 0.8 5.7 5.1

No effect on recall

To hide or not to hide?

Additional value of hiding is unclear. Users prefer “freedom”

Group % users visited

non-goal tasks

% users visited

not-ready tasks

% users visited

non-goal

concepts

% users visited

not-ready concepts

visited non-

goal

concepts

Non-adaptive 20 66 33 0 1.0

Adaptive 0 100 80 20 2.8

Restrictive 0 0 0 0 0.0

Evaluation of hiding

• Adaptive course on Hypertext (De Bra)

• Hiding “not ready” links

• Hiding obsolete links

• Small-scale evaluation

• No significant differences

• Students are not comfortable with disappearing links

InterBook: concept-indexed ET

• “Knowledge behind pages”

• Structured electronic textbook (a tree of “sections”)

• Sections indexed by domain concepts– Outcome concepts– Background concepts

• Concepts are externalized as glossary entries

• Shows educational status of concepts and pages

Sections and concepts

Chapter 1

Chapter 2

Section 1.1

Section 1.2

Section 1.2.1 Section 1,2,2

Textbook

Sections and concepts

Chapter 1

Chapter 2

Section 1.1

Section 1.2

Section 1.2.1 Section 1,2,2

Domain model

Concept 1

Concept 2

Concept 3

Concept 4

Concept m

Concept n

Textbook

Indexing and navigation

Chapter 1

Chapter 2

Section 1.1

Section 1.2

Section 1.2.1 Section 1.2.2

Domain model

Concept 1

Concept 2

Concept 3

Concept 4

Concept m

Concept n

Textbook

Glossary view

Navigation in InterBook

• Regular navigation– Linear (Continue/Back)– Tree navigation (Ancestors/Brothers)– Table of contents

• Concept-based navigation– Glossary (concept -> section)– Concept bar (section -> concept)– Hypertext links (section -> concept)

Adaptive navigation support

• Adaptive annotations– Links to sections– Links to concepts– Pages

• Adaptive sorting– Background help

• Direct guidance (course sequencing)– Teach Me

User modeling

• Overlay student model for domain concepts

• Knowledge states for each concept– unknown (never seen)– known (visited some page)– learned (passed a test)

• Information for sections– visited/not visited– time spent

• Information for tests: last answers

Adaptive annotation

• Educational status for concept unknown

known

learned

• Educational status for sections not ready to be learned

ready to be learned

suggested

Adaptive annotation in InterBook

1. State of concepts (unknown, known, ..., learned)

2. State of current section (ready, not ready, nothing new)

3. States of sections behind the links (as above + visited)

3

2

1

v

Bookshelves and books

Book view

Glossary view

Backward learning: “help” and “teach this”

InterBook: Evaluation

• Goal: to find a value of adaptive annotation

• Electronic textbook about ClarisWorks

• 25 undergraduate teacher education students

• 2 groups: with/without adaptive annotation

• Format: exploring + testing knowledge

• Full action protocol

Preferred ways of navigation

0

2

4

6

8

10

12

14

16

18

20

Cell Mean

pCONTINUE

pBACK pTEXT

pCONTENT

pINTRODUCINGpREQUIRING

no

yes

Cell Bar Char t Split By: ANS Inclusion cr iter ia: Hits > 15 from Eklund Separated.stv

0

.1

.2

.3

.4

.5

.6

.7

.8

.9

1

Units

%Annotated % Sequential

no

yes

Box Plot Split By: ANS Inclusion cr iter ia: Hits > 15 from Eklund Separated.stv

.213 .172 .042 17 .043 .742 0

.282 .226 .080 8 .059 .742 0

.151 .072 .024 9 .043 .273 0

.583 .258 .063 17 .097 .913 0

.516 .306 .108 8 .097 .909 0

.643 .207 .069 9 .364 .913 0

Mean Std. Dev. Std. Er ror Count Minimum Maximum # Missing

%Annotated, Total

%Annotated, yes

%Annotated, no

% Sequential, Total

% Sequential, yes

% Sequential, no

Descr iptive Statistics Split By: ANS Inclusion cr iter ia: Hits > 15 from Eklund Separated.stv

The effect of “following green”with adaptation

6

6.5

7

7.5

8

8.5

9

9.5

10

Units

Score

yes, Low-negative

yes, Low-positive

yes, High-positive

Adaptation mechanisms do work!

0

20

40

60

80

100

120

140

160

Units

s1t/ s1 s2t/ s2 s3t/ s3

Box Plot Split By: Par t Inclusion cr iter ia: Hits > 15 from Eklund Separated.stv

Results

• No overall difference in performance

• Sequential navigation dominates

...but ...

• Adaptive annotation encourage non-sequential navigation

• The effect of “following green”

• The adaptation mechanism works well

Where is the magic?

• No magic: Knowledge behind material• Knowledge about domain (subject)• Knowledge about documents

– Simple concept indexing

• Knowledge about students– Learning goal model– Overlay student model

• Straightforward techniques of user modeling and adaptation

Adaptive Hypertext: The Secret

• Adaptive hypertext has knowledge “behind” the pages

• A network of pages like a regular hypertext plus a network of concepts connected to pages

Pool of Learning ItemsDomain Model

Adaptive Hypertext: Design

• Design and structure knowledge space• Design a generic user model• Design a set of learning goals• Design and structure the hyperspace of

educational material• Design connections between the

knowledge space and the hyperspace of educational material

Domain model - the key

Concept 1

Concept 2

Concept 3

Concept 4

Concept 5

Concept N

Domain model - the key

• Knowledge about domain is decomposed into a set of fragments - domain knowledge elements (DKE)– Also called topics, knowledge items, concepts,

learning outcomes…• Most often DKE is just a name denoting a piece of

knowledge, sometimes it has internal structure• Semantic relationships between DKE can be

established

Vector vs. network models

• Vector - no relationships

• Precedence (prerequisite) relationship

• is-a, part-of, analogy: (Wescourt et al, 1977)

• Genetic relationships (Goldstein, 1979)

Vector (set) model

Concept 1

Concept 2

Concept 3

Concept 4

Concept 5

Concept N

Network model

Concept 1

Concept 2

Concept 3

Concept 4

Concept 5

Concept N

Overlay user model: knowledge

Concept 1

Concept 2

Concept 3

Concept 4

Concept 5

Concept N

Simple overlay model

Concept 1

Concept 2

Concept 3

Concept 4

Concept 5

Concept Nyesno

no

noyes

yes

Simple overlay model

Concept 1

Concept 2

Concept 3

Concept 4

Concept 5

Concept Nyesno

no

noyes

yes

Weighted overlay model

Concept 1

Concept 2

Concept 3

Concept 4

Concept 5

Concept N103

0

27

4

Simple goal model

• Learning goal as a set of topics

More complicated models

• Sequence, stack, tree

Indexing: the key to AH

• Types of indexing– Knowledge-based hypertext (concept = node)– Indexing of nodes– Indexing of Fragments

• How to get the hyperspace indexed?– Manual indexing (closed corpus)– Computer indexing (open corpus)

Knowledge-based hypertext

Concept 1

Concept 2

Concept 3

Concept 4

Concept 5

Concept N

Indexing of nodes

Domain model

Concept 1

Concept 2

Concept 3

Concept 4

Concept m

Concept n

Hyperspace

Indexing of page fragments

Fragment 1

Fragment 2

Fragment K

Concept 1

Concept 2

Concept 3

Concept 4

Concept 5

Concept N

NodeConcepts

Overlay model + indexing

• Indexing nodes with concepts– InterBook, ELM-ART, ISIS-Tutor, AHA

• Indexing anchors with concepts– StrathTutor

• Indexing fragments with concepts– MetaDoc, AHA, PT

• Nodes are concepts– InterBook, ELM-ART, ISIS-Tutor

Generalized overlay model

• Knowledge– overlay for set of concepts, network of concepts

• Goals– overlay for set of possible goals, tree of goals

• Stereotypes– overlay for set of stereotypes

Hyperspace structuring

• Concept-based hyperspace• No imposed structure• Hierarchy• ASK approach - conversational

relationships

Indexing with generalized model

• fragments are indexed with goals– PUSH

• nodes are indexed with user’s tasks– HYNECOSUM:

• nodes are indexed with stereotypes– EPIAIM, Anatom-Tutor, C-Book

• goals are nodes– HYPERFLEX

What else?• More functionality with the same knowledge

Adaptive hiding (ISIS-Tutor, AHA!)Adaptive presentation (AHA!)Adaptive Testing (ELM-ART)Adaptive Recommendation (Hirashima)

• More functionality with more knowledge– Domain specific

Knowledge about programming (ELM-ART)

– Domain-neutralAdvanced indexing (ADAPTS and COCOA)

ELM-ART: Lisp ITS on WWW

• Model: adaptive electronic textbook– hierarchical textbook – tests– examples– problems– programming laboratory

• Extra for Web-based teaching– messages to the teacher– chat room

ELM-ART: navigation and testing

Knowledge representation

• Domain knowledge– conceptual network for Lisp– problem solving plans– debugging knowledge

• Student model– overlay model for Lisp concepts– episodic model for problem-solving knowledge

Adaptivity in ELM-ART

• Adaptive navigation support

• Adaptive sequencing

• Adaptive testing

• Adaptive selection of relevant examples

• Adaptive similarity-based navigation

• Adaptive program diagnosis

ANS + Adaptive testing

Adaptive Diagnostics

Similarity-Based Navigation

ELM-ART: Evaluation

• No formal classroom study

• Users provided their experience

• Drop-out evaluation technology

• 33 subjects– visited more than 5 pages– have no experience with Lisp– did not finish lesson 3– 14/19 with/without programming experience

ELM-ART: Value of ANS

With adaptive

annotation

Without adaptive

annotation

NEXT button 21.0 (N = 4) 25.0 (N = 3) 22.7 (N = 7)

No NEXT button 13.8 (N = 5) 9.5 (N = 2) 12.6 (N = 7)

17.0 (N = 9) 18.8 (N = 5) 17.7 (N = 14)

Mean number of pages which the users with no experience in programming languages completed with ELM-ART

ELM-ART: Value of ANS

With adaptive

annotation

Without adaptive

annotation

NEXT button 23.5 (N = 6) 14.0 (N = 3) 20.3 (N = 9)

No NEXT button 22.4 (N = 5) 12.6 (N = 5) 17.5 (N = 10)

23.0 (N = 11) 13.1 (N = 8) 18.8 (N = 19)

Mean number of pages which the users with experience in at least one programming language completed with ELM-ART

Adaptive annotation can:

• Reduce navigation effortsResults are not significant (variety of styles?)

• Reduce repetitive visits to learning pagesSignificant - if applied properly

• Encourage non-sequential navigation

• Increase learning outcomeFor those who is ready to follow and advice

• Make system more attractive for students

What else?• More functionality with the same knowledge

Adaptive hiding (ISIS-Tutor, AHA!)Adaptive presentation (AHA!)Adaptive Testing (ELM-ART)Adaptive Recommendation (Hirashima)

• More functionality with more knowledge– Domain specific

Knowledge about programming (ELM-ART)

– Domain-neutralAdvanced indexing (ADAPTS and COCOA)

ONR research project Architecture for integration of:

– Diagnostics– Technical Information– Performance-oriented Training

Technology investigation & testbed

ADAPTS

ADAPTS: What it is

IETMsTraining

Diagnostics

Use

r M

od

el

Diagnostics

Content

Navigation

What task to doSystem health

What content is applicable to thistask and this user

Levels of detail

How to display this content to this user

Experience,

Preferences,

ASSESSES: DETERMINES:

Adaptive Diagnostics

Personalized Technical Support

Video clips

(Training)Schematics

EngineeringData

Theory ofoperation

Blockdiagrams

Equipment

Simulations

(Training)

EquipmentPhotos

Illustrations

TroubleshootingStep

Troubleshooting step plus

hypermedia support

information, custom-

selected for a specific

technician within a specific

work context.

ADAPTS dynamically assembles custom-selected content.

What’s in adaptive content?

ADAPTS - an adaptive IETM

The result

Maintenance

history Preprocessed,

condition-basedinputs

Technicianand OperatorObservations

Sensor inputs(e.g., 1553 bus)

PersonalizedDisplay

IETM TrainingStretch

text OutlineLinks

Training records

Skill assessment Experience

Preferences

Content NavigationDiagnostics

How do we make decisions?

User Model

Concept A

Concept A

ConceptB

ConceptB

ConceptC

ConceptC

SupportingInformation

Domain model

• Defines relationships between elements of technical information

• Indicates level of difficulty/ detail

• Shows prerequisites

Domain model example

CONCEPTReeling Machine

CONCEPTSonar Data Computer

CONCEPTSonar System

RemovalInstructions

TestingInstructions

IllustratedParts

Breakdown

Principles of

Operation

Principles of

Operation

Principles of

Operation

RemovalInstructions

RemovalInstructions

TestingInstructions

TestingInstructions

IllustratedParts

Breakdown

IllustratedParts

Breakdown

ReelingMachine

ReelingMachine

Sonar SystemSonar System

General Component Location

Principles of operation

Removal instructions

Principles of Operation

System Description DetailsParts List

Power Distribution

Domain model example

PART OF

Testing instructions

SUMMARY

DETAILS

TUTORIAL

User model

• Characterizes user ability at each element of the domain model– Size of model is bounded by domain– Weights on different types of elements account

for learning styles and preferences– Can be time sensitive

• Constrains the diagnostic strategy

User model example

Certified

CONCEPTReeling Machine

CONCEPTSonar Data Computer

CONCEPTSonar System

ROLERemoval

Instructions

ROLETesting

Instructions

ROLEIPB

ReviewedHands-on

Simulation

AT2 Smith

AD2 Jones

Preference

Reviewed

Hands-on+

Certified

Reviewed

Hands-on

Hands-on Reviewed

Reviewed

ROLETheory of Operatio

n

Adaptive content selection

• Information is custom-selected for a user– Level of detail offered depends upon who the

user is (i.e., his level of expertise)– Selected at a highly granular level, e.g., for

each step within a procedure

• Performance-oriented training is presented as part of content

Integrated interface

Summary

• Concept-based approach to adaptive hypertext and adaptive WBS

• Concept indexing: Knowledge behind pages

• Explored– Different levels of model complexity– Different application domains– Different adaptation techniques

The complexity issue

ISIS-Tutor, MSU (1992-1994)

COCOA, CTE (1998-1999)

ELM-ART, Trier (1994-1997)

InterBook, CMU (1996-1998)

ADAPTS, CMU, (1997-1998)ITEM/IP, MSU (1986-1994)

Open Corpus AH

• Current AH techniques are based on manual page/fragment indexing– How to work with open Web and Digital

Libraries?

• Ways being explore– Add open corpus content, but index it manually– Use IR techniques for content-based adaptation

without manual indexing– Use social navigation approaches for adaptation

Social Navigation Support

AH Service: NavEx

AH Service: QuizGuide

More information...

• Adaptive Hypertext and Hypermedia Home Page: http://wwwis.win.tue.nl/ah/

• Brusilovsky, P., Kobsa, A., and Vassileva, J. (eds.) (1998), Adaptive Hypertext and Hypermedia. Dordrecht: Kluwer Academic Publishers.

• Brusilovsky, P. (2001) Adaptive hypermedia. User Modeling and User Adapted Interaction, Ten Year Anniversary Issue (Alfred Kobsa, ed.) 11 (1/2), 87-110

• Brusilovsky, P. (2003) Developing adaptive educational hypermedia systems: From design models to authoring tools. In: T. Murray, S. Blessing and S. Ainsworth (eds.): Authoring Tools for Advanced Technology Learning Environment. Dordrecht: Kluwer Academic Publishers.