PPT

Foundations of Adaptive Hypermedia


Heiko Spallek, DMD, PhDCenter for Dental Informatics

University of Pittsburgh, School of Dental Medicine, Pittsburgh, PA

March 30, 2005

Dental Informatics Seminar (2202)



Agenda Background

e-learning hype reality check current status of dental online courses

Adaptive Hypermedia definitions, concepts usage in education, information retrieval, kiosk & help systems

Conceptual architecture Technical implementation In-class assignment Evaluation Conclusions


Hype About E-learning

General number of online courses 50,000 100,000 between 1998 and 1999 * online learning market: $2.2 billion in 2001, estimated: $11.4 billion by 2003 * e-learning conference in Europe attracted 10,000 attendees last year

Dentistry 1998: 158 online continuing dental education (CDE) courses from 32 providers 2001: just DentalxChange.com: 428 courses in 27 different categories

* International Data Corp.


Reality check“E-learning is the revenge of the nerds.”

evaluation of 158 online continuing dental education uncovered major deficiencies in existing Web-based CDE courses *

no author indicated relation between course length and credit hours varies

market for course shells immature no standards → risk of losing content no peer-review process for online learning resources → no academic credit usually no formative or summative evaluation high drop-out rate (based on meta analysis of multiple studies) in dentistry/medicine: no one has reported profits yet (break even for online course: 300-500 participants)

* Schleyer T, Pham T. Online continuing dental education. Journal of the American Dental Association 1999;130(June):848-854,877-878


Current state-of-the-art dental online course collection of static Web pages

translations of existing lecture notes existing course manuals existing papers

= first generation online course model

Course material is still implicitly oriented for a traditional on-campus audience!


Generations of online courses

1. Generation static pages

2. Generation adapted by the instructor to different audiences (static versions of the same course saved at different locations)

3. Generation adaptive to the individual learner


Adaptive Hypermedia (AH)

- new direction of research - crossroads of hypermedia and user modeling- model of the goals, preferences, knowledge of each individual user- use of model throughout interaction with user- adapt to the needs of user

Brusilovsky 1996, Brusilovsky 2001


Definition User Model- “mental state” of the user *- AH system keeps a model of the user - to adapt teaching sequence, presentation - for each Knowledge Item (KI)

- individual user model

- stores some value

- estimation of the user’s knowledge level of this concept **

* De Bra and Calvi 1998

** Brusilovsky 1996


AH ― Areas of Application

• educational hypermedia• online information systems• online help systems• information retrieval hypermedia• institutional hypermedia• systems for managing personalized views in information spaces

Brusilovsky, P. (2001) Adaptive hypermedia. User Modeling and User Adapted Interaction, Ten Year Anniversary Issue (Alfred

Kobsa, ed.) 11 (1/2), 87-110


Adaptivehypermediatechnologies

Adaptivepresentation

Adaptivenavigation support

AdaptiveMultimedia

presentation

Adaptive textpresentation

Adaptationof modality

Canned textadaptation

Naturallanguage adaptation

Inserting/Removingfragments

Alteringfragments

Sortingfragments

Dimmingfragments

Strechtext

Direct guidance

Adaptive linksorting

Adaptive linkhiding

Adaptive linkgeneration

Map annotation

hiding

disabling

removal

The updated taxonomy of adaptive hypermedia technologies; Brusilovsky, P. (2001) Adaptive hypermedia. User Modeling and User Adapted Interaction, Ten Year Anniversary Issue (Alfred Kobsa, ed.) 11 (1/2), 87-110

Taxonomy of AH Technologies


Curriculum Sequencing

active - implies a learning goal- best individual path to achieve the goal

passive- reactive technology- offers the user a subset of available learning material fill the gap in student's knowledge


AH Technologies in Web-based Educationadaptive navigation support technology

- changing the appearance of visible links

= generalization of curriculum sequencing technology in a hypermedia context

- direct guidance, adaptive link annotation, adaptive link hiding

adaptive presentation technology

- adapt the content of a Web page to the user's goals, knowledge and other information stored in the user model

- pages are adaptively generated or assembled from pieces for each user


Conceptual architecture

1. What features of the learner can be used as a source of the adaptation?

2. What can be adapted in a technical feasible manner?


1. What features of the learner can be used as a source of the adaptation?

user’s knowledge- pre-test results- completion of quizzes throughout the course

user’s interests/goals- selection of a specific learning goal

user’s traits- learning style


2. What can be adapted in a technical feasible manner?

user’s learning goal compilation of KIs

individual curriculum (active sequencing)

system’s estimation of the learner’s knowledge level of each KI stored modifies content (adaptive presentation)


Adaptiveness

adaptive text presentation- inserting/removing of fragments- altering fragments- stretch text

adaptive navigation support - adaptive links: hiding- adaptive link annotation- sitemap adaptation

admission to final quiz only after mastering of all offered KIsforced repetition of KI example-based learning versus concept-based learning


AH system stores information

User ModelUser Info Course Info

Learning Goal Knowledge Items

A G B D E H O J L T

User

provides enrollment information

Demographic Information

Demographic

Access

Privacy

AccessRightsPrivacyInformationLearningGoal

performes pre-test

A

Estimation of user’s knowledge

Knowledge Assessment

GB

DE H

OJ

L

T

Learning style

Final test

Test history

provides traits

LearningStyle AH system

compiles KIsAH system

compiles pre-testEstimation of

user’s knowledgeAH system adjusts KIs

used for adaptation effect

AH system compiles curriculum

AH system adjusts user model

Test results are used for revising the user model.

O

Eligibility for Final Test

A G B E HJ

T

Final test

AH system compiles finalAH system

evaluates final

Test results


Technical Implementation

Learner Environment

Authoring Tool


Authoring Tools general course setting views of KIs

- abbreviated version of the core explanation (short)- for novice users (basic)- very detailed high-level textual explanation (main)- for graduate level knowledge (advanced)

multiple-choice questions fill-in-the-blank questions descriptive / online examples prerequisites KI metadata: keywords, semantic

Teaching Tools

student registration, performance, question analysis


Learner Environment

General design approach follows Mark Weiser’s idea:

“The most profound technologies are those that disappear. They weave themselves into the fabric of everyday life until they are indistinguishable from it.”

Scientific American, 265(3):94--104, September 1991


Learner Environment

Overview page- background about the subject of the course- explanation about the concept of AH- introduction to personal tutor - acknowledgements


Learner Environment

Emotional Intelligence?

show emotional behavior (e.g. the persona-effect, help relieve frustration, recover to a positive emotional state)

beneficial for tutoring agents and learning software C Elliot, J.C. Lester, J. Rickel. Integrating affective computing into animated tutoring agents

In Proceedings of the IJCAI Workshop on Animated Interface Agents

Nagoya, Japan, 1997, 113–121

improve user acceptanceJ. Klein. Computer Response to User Frustration. TR480, MIT, 1998


Learner Environment

- user model is dynamic and is continuously revised- not based on the content reviewed

Beaumont 1994

Boyle and Encarnacion 1994

Vassileva 1996

Henze et al. 2002


User ModelUser Info Course Info

Learning Goal Recompile Knowledge Items

A U F

User

Demographic

Access

Privacy

LearningGoal

A

Estimation of user’s knowledge

GB

D

OJ

Learning style

Test history

AH system re-compiles curriculum

User changes learning goal.

K RE

L

STM

H

New curriculum


Annotations

- feedback is encouraged - annotations = informal continuous formative evaluation *

all KIs are annotatable by learners

* Henze et al. 2002


Design

- cross platform, all browsers- screen resolutions with a minimum of 800x600 pixel- Guidelines of the Design of Educational Software *

* ANSI Standards Committee on Dental Informatics


Technology

Purpose Technology

Web server for hosting IIS on MS-Windows 2000 Server

Programming language Scripting language: PHP version 4 and Perl-compatible regular expression

Database (highly granular storage) MySQL version 3.23.47 on Linux

Validation of user input JavaScript

Authentication Cookies


Users AnnoUsers

AnnoKI

Descriptive Examples

Prerequisites

KIQuestions

KIDescript

Ex

QuestionsLog

KIAnnotation

User Model

Questions


PHP Scripting// check what is up with prerequisites

if($prereq!="disregard"){

$queryPrereq = "select Requires from AH_Prerequisites where KI='$KIID'";

$dbPrereq->query($queryPrereq);

if($dbPrereq->num_rows()){

$querygetKI = "select AH_Knowledge_Items.KI_title, AH_Knowledge_Items.KIID from AH_Prerequisites, AH_Knowledge_Items where AH_Prerequisites.KI='$KIID' and AH_Prerequisites.Requires=AH_Knowledge_Items.KIID";

$dbgetKI->query($querygetKI);

$KItitlewithlink="";

while($dbgetKI->next_record()):

$vKI_title = $dbgetKI->Record["KI_title"];

$vKIID = $dbgetKI->Record["KIID"];

$score=getscore($vKIID, $UserID);


Evaluation

1. initial review (cognitive walkthroughs )

2. alpha-test (task-oriented)

3. formative evaluation (heuristic evaluation & end-user testing)

4. summative evaluation (end-user testing) [planned]


Cognitive Walkthroughs

Goal: solve anticipated problems before bringing the course to users for testing

- performed independent and under observation of the developer

Results (encountered problems were fixed immediately )- confusing labels of fields- confusing options during enrollment- missing guidance on start page- repetition of test questions- inadequate feedback for wrong answers


Alpha Test

task-oriented alpha test- complete the course - determine deficiencies (educational, instructional design, technical,

design and layout)- send an unstructured informal feedback

Results (encountered problems were fixed immediately )- general feedback was positive- personal tutor concept was beneficial- 12 specific usability, functionality issues were discovered (regarding: change

of setting, annotations, quiz, flow, sequencing)


Formative Evaluation I

Usability Inspection: Heuristic Evaluation

- general heuristic for UI design (Jakob Nielsen)- heuristic extracted from ANSI standard “Guidelines for the Design of

Education Software”http://di.dental.pitt.edu/edswstd/


In-Class Assignment

Evaluate the pre- and post-questionnaires− Problems/Critique− Order of questions− Comparing terms− Formatting− Disclaimer


Formative Evaluation II

End-user testing− separation between delivery (media) and instructional method

− use of a validated instrument

Evaluation of– participants’ reactions

– achievement of program objectives

• changes in student learning

• values

• motivations

• ability to transfer knowledge acquired outside the instructional setting


Conclusions

• applying of established principles of AH research• state-of-the-art programming environment• separation of content and design• Web-based delivery system For the first time, a dental educational delivery system adapts itself

dynamically using active and passive curriculum sequencing and adaptive presentation.

authoring process = challenge ** Brusilovsky, Eklund, and Schwartz 2002

* De Bra and Calvi 1998

PPT

Documents

hardware events

layers of components

components frame

types of components

wires components

lower level commands

internal tasks commands

data char