7/29/2019 23_Melecon_2000

1/4

Studentmodule

Teachermodule

Domain knowledgemodule

User interfacemodule

Figure 1. Typical ITS Architecture

Abstract Asynchronous distance education delivery systems

do not require real-time student-human teacher interaction

thus enabling students to use tutoring resources anytime and

anywhere. Among various possibilities for implementing

asynchronous distance education delivery computer supported

ones are nowadays the most popular. Categories thereof are

intelligent tutoring systems that are used for supporting and

improving the process of learning and teaching in arbitrarydomain knowledge. In such education models the role of

human teacher is replaced with a computer tutor basically

consisting of a domain knowledge base as well as of both a

student module with an appropriate student model and a

teacher strategy guiding the learning and teaching process.

The paper describes an authoring shell Distributed Tutor-

Expert System (DTEx-Sys) developed for asynchronous

distance education purposes, as a secondary knowledge source

for teachers and their students in secondary and primary

schools.

Index Terms asynchronous distance education, authoring

shells, intelligent tutoring systems.

I. INTRODUCTION

The growing popularity and ease of access to the World

Wide Web (WWW) stimulates present day interest in

systems for self-learning in general and distance learning(DL) in particular. The Web provides itself as an

outstanding delivery medium, thus acting as both a provider

of content and of subject matter. On the other hand HTML

technology allows and easy way to deliver all kinds of text,graphics, sound and video offer.

It is claimed that the Web represents a well-chosen

technology for implementing asynchronous distant

education delivery systems. Such systems do not require

real-time student/human teacher interaction hence enablingstudents to use tutoring resources anytime and anywhere.

Moreover, students within asynchronous group learning

environments do not need to manage their schedules.Learning networks based on asynchronous communication

offer unique opportunities for active participation. Unlike

the traditional classroom, students within on-line courses

have access to the airtime they want or need, enabling everyone of them to have a voice. Asynchronicity provides each

Marko Rosi and Slavomir Stankov are with the Faculty of Natural

Sciences, Mathematics and Education, University of Split, HR-21000

Split, Croatia (e-mail: marko.rosic{stankov}@pmfst.hr).

Vlado Glavini is with the Faculty of Electrical Engineering and

Computing, University of Zagreb, HR-10000 Zagreb, Croatia (e-mail:vlado.glavinic@fer.hr).

student the time to reflect, formulate ideas and compose

responses thoughtfully, hence enabling the elevation of the

quality of student interaction and participation.

Among various possibilities for implementing

asynchronous distance education delivery computersupported ones are nowadays the most popular. A category

thereof are intelligent tutoring systems (ITS) that are used

for supporting and improving the process of learning andteaching in arbitrary domain knowledge. In such educationmodels the role of human teacher is replaced with a

computer tutor, basically consisting of a domain knowledge

base as well as of both a student module with anappropriate student model and a teacher strategy guiding

the learning and teaching process [1]. An ITS takes into

consideration the knowledge about what to teach (subject

matter domain knowledge), the way to teach (pedagogicalstrategy teacher), as well as the relevant information

about the student being taught (student). As shown in

Figure 1, there are four main interconnected modules of an

intelligent tutoring system: (i) domain knowledge module

with the domain knowledge base, (ii) teacher module forguiding the teaching and learning process, (iii) student

module with information that is specific to each individualstudent, and a (iv)suitable user interface module enabling

interaction among student, teacher and domain knowledge

[1,2].

Teaching and learning are expressed by a four-phase

functional cycle: (1) didactics, (2) perception, (3) diagnostic

and evaluation and (4) help and remediation [3]. Thisprocess is a self-adaptable one and is performed according

to the assumed pre-knowledge, capabilities and students'

behavior.

Intelligent Tutoring Systems for Asynchronous

Distance EducationMarko Rosi, Slavomir Stankov, and Vlado Glavini,Member, IEEE

7/29/2019 23_Melecon_2000

2/4

The biggest problem in building an ITS is its high costsresulting from the inherently cyclic and time consuming

design process. In order to overcome this problem another

approach has been chosen, namely to create particular ITSs

from flexible shells acting as program generators. Such

authoring shells should show design usability andflexibility so as (i) to allow different representations of

problem areas, and (ii) to enforce ease-of-use when

developing an ITS for a particular problem area [4].

The asynchronous nature of ITSs makes them a goodcandidate for use throughout the Web, since they facilitate

instruction offer anywhere and to anyone and for an

arbitrary domain.

In this paper we present a Web-oriented authoring shell Distributed Tutor-Expert System (DTEx-Sys), which is

developed for asynchronous distance education purposes.

DTEx-Sys is designed as a 3-tier client-server architecture

[5] where the intelligent tutoring functions are separatedfrom the user interface and the knowledge base(s). The

system functionality comprehends knowledge base access

for arbitrary domain knowledge, along with testing,

diagnosing and evaluation of students work. TEx-Sysfeatures Web-based tutoring by means of standard browsers

and an interactive hypermedia learning environment, and

enables a particularly easy two-way communicationbetween students and teachers, as well as students and

students.

II. WEB ORIENTED INTELLIGENT TUTORING SYSTEMS

The Web has demonstrated itself as one of the most popular

education supporting technologies on the Internet.Thousands of Web-based courses have been made available

within the last five years already, unfortunately most of

them using static hypertext pages [6]. In contrast, there arepresently only few technologies for implementing ITSs on

the Web. Frequently used technologies are those enabling

dynamic generation of Web document contents, hereincluding the following ones: Common Gateway Interface

(CGI), Active Server Pages (ASP) and Java Server Pages

(JSP). ASP and JSP are better solutions than CGI, because

they are faster and support interoperability with

components (Component Object Model or JavaBeanscomponents). Another group of technologies for developing

distributed ITSs are those based on distributed objects suchas Common Object Request Broker Architecture (CORBA),Distributed Component Object Model (DCOM) and Java

Remote Method Invocation (Java RMI). Additionally, for

operating the clients side, Java applets and Active Xcomponents are also used.

A. Web-Based Authoring Shells

A number of Web-based authoring shells six of them

are considered in the following, with the purpose of

illustrating design decisions and implementationtechnologies.

ELM-ART (ELM Adaptive Remote Tutor) is a Web basedITS to support the learning process of LISP programming.

It is based on the on-site ELM-PE (ELM-Programming

Environment) intelligent learning environment used to

teach an introductory LISP course at the University of Trier

[7]. The course materials in ELM-ART are presented as anon-line intelligent textbook with a hypermedia integrated

problem solving environment. The system is implemented

using Common Lisp Hypermedia Server functions.

CALAT (Computer Aided Learning and Authoringenvironment for Tele-education) consists of an ITS shell on

the Web server side and of a multimedia scene viewer on

the client side. This system is built with the aim ofachieving a high degree of adaptation to individual learners

by dynamically selecting the subject matter. Besides, it

includes a courseware authoring system that can be readily

used by courseware specialists even without system or

programming skills [8,9]. CALAT uses CGI in combination

with appropriate plug-ins.

Lee and Wang [10] proposed an intelligent hypermedia

learning system that can perform intelligent as well as

individual instruction based on the combination of CORBAand the Web. It provides instructional units for diagnosing

students response and allows dynamic instruction

strategies developed according to a students progress.

Another system with a Web oriented intelligent learning

environment is VALIENT (Virtual Adaptive LearningIntegrated Environments using Net-based Technologies)

[11]. Its authors aim to combine the strengths of the Web asa mechanism for delivering conventional instructional

material with the learning opportunities available inintelligent learning environments. VALIENT uses JSPtechnologies and Java applets.

The IDLE system (Intelligent Distance Learning

Environment) supports the development of adaptive Web

courseware in the areas of artificial intelligence (especially

knowledge engineering) and cognitive sciences

(psycholinguistics and neuro-psychology). Theimplemented prototype is a multi-agent portable system

enforcing access restriction and using multimedia effects

[12]. There are both a UNIX and a PC version, the UNIXone using CGI, while the PC one using JSP technology.

The MANIC (Multimedia Asynchronous NetworkedIndividualized Courseware) [13] project is targeted towards

facilitating individualized asynchronous distance learning

based on new technologies like high-speed computernetworks and hypermedia instructional material (text,

graphics, animations) presentation. Its authors concentrate

the research especially on the effective use of Webhypermedia technology for delivering to students stored

instructional courses, such as class notes/overheads and

audio/video of classroom lectures. MANIC is implementedwith CGI, using Real Audio and Real Video Server.

7/29/2019 23_Melecon_2000

3/4

Figure 4. Access to knowledge base

Knowledge basesaccess module

Consultation module

Testing module Knowledgebases

DTEx-Sys

Student

Teacher

Figure3. Structure of DTEx-Sys shell

III. DISTRIBUTED TUTOR-EXPERT SYSTEM (DTEX-SYS)

Our DTEx-Sys is built by keeping in mind universality and

application quality thus resulting in generality of use. A

brief consideration of DTEx-Sys universality featuresshows (i) the support for availability to a wide group of

users and (ii) the possibility of its use for arbitrary domainknowledge. A coarse-grained decomposition of DTEx-Sysuniversality results in the following elements:

user time independence when accessing systemservices,

user location independence when using systemservices,

independence of client computer system platform, limited requirements upon client computer system, and

uniform way of storing knowledge regardless of thesubject matter in question.

Regarding application quality the following elements are

considered:

the system is adaptable to individual users needs, learning is performed within a hypermedia

environment,

the system supports a teacher consultation option,

it offers a user supervision option, it a incorporates a user test feature,

it provides tested users with recommendations forfuture work,

the system is compatible with resources pertaining toother distance learning systems, and

it is suitable for future refinement.

The system has been designed according to the general

principles of student knowledge and skills acquisition, aswell as of the methods and techniques of knowledge

representation. Of course, DTEx-Sys builds upon the

experiences of an earlier on-site version denoted TEx-Sys(Tutor-Expert System) [2], whose main objective has been

the generation of hypermedia-rich learning environments

for teaching and learning basic principles in control theory

and their application in natural, technical and social

systems.

All of DTEx-Sys functions are accessible through standardWeb browsers. Advantages of such an approach include the

following ones: the Web is a generally accepted service for envisaged

users of DTEx-Sys,

Web browsers are free of charge software products, Web browsers enable client platform independence, Web browsers allow for an interactive hypermedia

learning environment,

the two-way student-teacher as well as student-studentcommunication can be made extremely easy using

Internet resources, and

there are many other educational resources located onthe Internet which can additionally be used by DTEx-

Sys.

DTEx-Sys is developed as 3-tier client-server architecture[5] where the intelligent tutoring functions are separated

from the user interface and the knowledge base as shown in

Figure 2. Semantic networks with frames are the formalism

for knowledge representation in DTEx-Sys. Nodes

represent domain knowledge objects, while links showrelations between objects. Additionally nodes from the

knowledge base can have any of the following structural

attributes: textual descriptions with hypertext, pictures,(sequence of) slides, animations (moving pictures and

sound) and Internet links.

DTEx-Sys users are supported in the process of learning

and teaching with the following services: knowledge

base(s) access, testing knowledge, reception of test results

with recommendation for future work and consultation withinstructors. These services are of course provided by their

respective system modules, as shown in Figure 3: (i)

knowledge access module, (ii) testing module and (iii)consultation module. All the modules exhibit appropriate

user interfaces. E.g. the user interface for accessing the

knowledge bases shows domain knowledge objects along

Figure 2. DTEx-Sys 3-tier architecture

Knowledgebases

Userinterface

DTEx-Sysfunctions

7/29/2019 23_Melecon_2000

4/4

Figure 5. Results of test and recommendations for

future work (rating 4 roughly equivalent to B)

with structure attributes assigned to the selected object, seeFigure 4. Testing is implemented using Web quizzes: after

testing the student receives as feedback both the rating and

recommendations for future work, see Figure 5. All of these

activities are under control of the DTEx-Sys shell.

DTEx-Sys is implemented with Microsoft technology,

therefore including the Windows DNA (Distributed

interNet Applications Architecture) model. The contents of

Web documents that are distributed to clients is generated

dynamically.

IV. CONCLUSION

In this paper we discuss authoring shells presently being a

very promising CAI tool class for insuring Web-basedasynchronous distant education. After shortly reviewing six

such systems known from the literature, we describe thebasic architecture and outline the implementation of the

authoring shell Distributed Tutor-Expert System (DTEx-

Sys). DTEx-Sys has been in use at the Faculty of NaturalSciences, Mathematics and Education, University of Split,

for some time already. E.g. it has been used for supporting

both the course "Computer Aided Instruction" during thepast academic year for several study groups, and the

realization of a number of diploma works for various

domain knowledges. Beside that, in our teaching practicewe use DTEx-Sys for improving the learning and teaching

process in regular and further education, where it serves asa secondary knowledge source for teachers and their

students in secondary and primary schools.

ACKNOWLEDGMENTS

This work has been carried out within projects 036033

Architectural Elements for Regional Information

Infrastructures jointly funded by the Ministry of Science

and Technology of the Republic of Croatia and the Istrian

County, and 177010 Independence of Student Using New

Information Technology funded by the Ministry of Science

and Technology of the Republic of Croatia.

REFERENCES

[1] H. L. Burns and C. G. Capps: Foundations of Intelligent TutoringSystems: An Introduction, in M. C. Polson, J. J. Richardson, Eds.:

Foundations of Intelligent Tutoring Systems, Lawrence ErlbaumAssociates Publishers, 1988, pp. 1-18.

[2] S. Stankov:Isomorphic Model of the System as the Basis of TeachingControl Principles in the Intelligent Tutoring System, PhD Diss.,

Faculty of Electrical Engineering, Mechanical Engineering and

Naval Architecture, University of Split, Split, Croatia 1997 (in

Croatian).[3] M. C. Angelides, I. Garcia: Towards an Intelligent Knowledge

Based Tutoring System for Foreign Language Learning, CIT

Journal of Computing and Information Technology, Vol. 1, No. 1,

March 1993, pp. 15-28.[4] M. Barton. Authoring Shells for Intelligent Tutoring Systems, 7th

World Conference on Artificial Intelligence in Education,

Washington, DC, USA, August 16-19, 1995,

ht t p: / / www. pi t t . edu/ ~al / ai ed/ bar t on. html [5] J. Edwards: 3-Tier Client/Server At Work, Rev. Ed., John Wiley

&Sons, Inc., NewYork, 1999

[6] P. Brusilovsky: Adaptive Educational Systems on the World-Wide-Web: A review of Available Technologies,ITS98 The Workshop:

Intelligent Tutoring Systems on the World Wide Web, San Antonio,

TX, USA, August 16-19, 1998. ht t p: / / www. aml . cs.umass . edu/ ~st er n/ webi t s/ i t swor kshop/ br usi l ovsky. ht ml

[7] P. Brusilovsky, E. Schwartz, and G. Weber: ELM-ART: AnIntelligent Tutoring System on World Wide Web, ITS96 TheWorkshop: Architecture and Methods for Designing Cost-Effective

and Reusable ITS, Montreal, June 10-14, 1996. ht t p: / / www.cont r i b. andr ew. cmu. edu/ ~pl b/ I TS96. ht ml

[8] M. Kiyama, S. Ishiuchi, K. Ikeda, M. Tsujimoto, Y. Fukuhara: AVisual Authoring Tool for the Web-Based Intelligent CAI and its

Evaluation, Proc. ED-MEDIA 98 World Conference onEducational Multimedia and Hypermedia, Freiburg, Germany, June

20-25, 1998, pp. 778-783.[9] K. Nakabayashi, Y. Koike, M. Maruyama, H. Touhei, S. Ishiuchi, Y.

Fukuhara: An Intelligent Tutoring System on World-Wide Web:

Towards an Integrated Learning Environment on a Distributed

Hypermedia, in H. Maurer, Ed.: Educational Multimedia andHypermedia, Proc. ED-MEDIA 95. World Conference on

Educational Multimedia and Hypermedia, Graz, Austria, June 17-21,

1995, pp. 488-493.

[10] S. H. Lee and C. J. Wang: Intelligent Hypermedia Learning Systemon the Distributed Environment, Proc. ED-MEDIA 97 World

Conference on Educational Multimedia and Hypermedia, Calgary,Alberta, Canada, June 14-19, 1997, pp. 780-785.

[11] L. Hall, A. Gordon: Synergy on the Net: Integrating the Web andIntelligent Learning Environments, ITS98 The Workshop:

Intelligent Tutoring Systems on the World Wide Web, San Antonio,

TX, August 16-19, 1998.ht t p: / / wwwaml . cs . umass. edu/ ~st er n/ webi t s/ i t swor kshop/ gordonhal l . ht ml

[12] T. Gavrilova, T. Chernigovskaya, A. Voinov, S. Udaltsov: IntelligentDevelopment Tool for Adaptive Courseware on WWW, 1999.ht t p: / / www. csa. r u/ I nst / gorb_dep/ ar t i f i c / I A/ cali sce98. htm

[13] M. K. Stern: The Difficulties in Web-Based Tutoring, and SomePossible Solutions, 8th World Conference of the AIED Society,

Kobe, Japan, 18-22 August 1997.ht t p: / / www. cont r i b. andr ew. cmuedu/ ~pl b/ AI ED97_woekshop/ Ster n. html