Faster Maintenance with Augmented Reality

In the not-too-distant future, it might be possible to slip on a pair of augmented-reality (AR) goggles instead of fumbling with a manual while trying to repair a car engine. Instructions overlaid on the real world would show how to complete a task by identifying, for example, exactly where the ignition coil was, and how to wire it up correctly.

A new AR system developed at Columbia University starts to do just this, and testing performed by Marine mechanics suggests that it can help users find and begin a maintenance task in almost half the usual time.

AR has long shown potential for both entertainment and practical applications, and the first commercial applications are starting to appear in smart phones, thanks to cheaper, more compact computer chips, cameras, and other sensors. So far, however, these apps have been mainly limited to providing directions. But researchers are also working on many practical applications, including ways to help with specific repair and maintenance tasks.



 The Columbia researchers worked with mechanics from the U.S. Marine Corps to measure the benefits of using an AR headset when performing repairs to a light armored vehicle. Currently, Marine mechanics have to refer to a technical manual on a laptop while performing maintenance or repairs inside the vehicle, which has many electric, hydraulic, and mechanical components in a tight space

TICNO ...RSS

رح بسيط بالفيديو لاهم خدمات web2.0

RSS



as google reader





Blogs



Online Photo Sharing in Plain English




Social Bookmarking in Plain English





Social Media




Twitter in Plain English

Artificial Intelligence in Education

systems

1 0

References

1. Alpert, S. R., Singley, M. K., and Fairweather, P. G.: Deploying Intelligent Tutors on the

Web: An Architecture and an Example. International Journal of Artificial Intelligence in

Education

http://cbl.leeds.ac.uk/ijaied/abstracts/Vol_10/alpert.html

2. Anderson, J. R. and Reiser, B.: The LISP tutor. Byte

3. André, E., Rist, T., and Müller, J.: WebPersona: A Life-Like Presentation Agent for

Educational Applications on the World-Wide Web. In: Brusilovsky, P., Nakabayashi, K. and

Ritter, S. (eds.) Proc. of Workshop "Intelligent Educational Systems on the World Wide

Web" at AI-ED'97, 8th World Conference on Artificial Intelligence in Education, Kobe,

Japan, ISIR (1997) 78-85, available online at

http://www.contrib.andrew.cmu.edu/~plb/AIED97_workshop/Andre/Andre.html

4. Anjaneyulu, K.: Concept Level Modelling on the WWW. In: Brusilovsky, P., Nakabayashi,

K. and Ritter, S. (eds.) Proc. of Workshop "Intelligent Educational Systems on the World

Wide Web" at AI-ED'97, 8th World Conference on Artificial Intelligence in Education,

Kobe, Japan, ISIR (1997) 26-29, available online at

http://www.contrib.andrew.cmu.edu/~plb/AIED97_workshop/Anjaneyulu.html

5. Barr, A., Beard, M., and Atkinson, R. C.: The computer as tutorial laboratory: the Stanford

BIP project. International Journal on the Man-Machine Studies

6. Brusilovsky, P.: Intelligent tutoring systems for World-Wide Web. In: Holzapfel, R. (ed.)

Proc. of Third International WWW Conference (Posters), Darmstadt, Fraunhofer Institute

for Computer Graphics (1995) 42-45

7. Brusilovsky, P.: Methods and techniques of adaptive hypermedia. User Modeling and User-

Adapted Interaction

8. Brusilovsky, P.: Adaptive educational systems on the World Wide Web. In: Ayala, G. (ed.)

Proc. of Workshop "Current Trends and Applications of Artificial Intelligence in Education"

at the 4th World Congress on Expert Systems, Mexico City, Mexico, ITESM (1998) 9-16

9. Brusilovsky, P., Eklund, J., and Schwarz, E.

developing adaptive courseware. Computer Networks and ISDN Systems.

291-300

10. Brusilovsky, P. and Pesin, L.: An intelligent learning environment for CDS/ISIS users. In:

Levonen, J. J. and Tukianinen, M. T. (eds.) Proc. of The interdisciplinary workshop on

complex learning in computer environments (CLCE94), Joensuu, Finland, EIC (1994) 29-33,

available online at http://cs.joensuu.fi/~mtuki/www_clce.270296/Brusilov.html

11. Brusilovsky, P., Ritter, S., and Schwarz, E.: Distributed intelligent tutoring on the Web. In: du

Boulay, B. and Mizoguchi, R. (eds.) Artificial Intelligence in Education: Knowledge and

Media in Learning Systems. IOS, Amsterdam (1997) 482-489

12. Brusilovsky, P., Schwarz, E., and Weber, G.: ELM-ART: An intelligent tutoring system on

World Wide Web. In: Frasson, C., Gauthier, G. and Lesgold, A. (eds.) Intelligent Tutoring

Systems. Lecture Notes in Computer Science, Vol. 1086. Springer Verlag, Berlin (1996)

261-269

13. Burns, H. L. and Capps, C. G.: Foundations of intelligent tutoring systems: An introduction.

In: Polson, M. C. and Richardson, J. J. (eds.): Foundations of intelligent tutoring systems.

Lawrence Erlbaum Associates, Hillsdale (1988) 1-19

14. Carver, C. A., Howard, R. A., and Lavelle, E.: Enhancing student learning by incorporating

student learning styles into adaptive hypermedia. In: Proc. of ED-MEDIA'96 - World

Conference on Educational Multimedia and Hypermedia, Boston, MA, AACE (1996) 118-

123

10 (1999) 183-197, available online at10, 4 (1985) 159-1758, 5 (1976) 567-5966, 2-3 (1996) 87-129: Web-based education for all: A tool for30, 1-7 (1998)

Augmented Reality Goggles

I held a black-and-white square of cardboard in my hand and watched as a dragon the size of a puppy

 appeared on top of it and roared at me. I watched a tiny Earth orbit around a real soda can, saw virtual balls fall through a digital gap in a table, and viewed a life-sized virtual human sitting in an empty chair.
What made these impressive special effects possible was a pair of augmented reality (AR) glasses—specifically, the Wrap 920AR glasses from Vuzix. Whereas virtual reality shows you only a digital landscape, augmented reality (AR) mixes virtual information, like text or images, into your view of the real world in real-time.
In the last few years, AR has started appearing on smart phones. In that context, software superimposes information on top of your view of the world as seen through the device's screen. But AR eyewear, which provides a more immersive experience, has been confined to academic research and niche applications like medical and military training. That's been largely because older AR hardware has been so bulky and has cost tens of thousands of dollars.
The Wrap 920AR from Vuzix, based in Rochester, New York, costs $1,995—about half the price of other AR goggles with similar image resolution. The company hopes that the glasses will appeal to gamers, animators, architects, and software developers, and it has developed software for building AR environments, which is included with the glasses.

s, which have never been used for

teaching real distance classes. The rest of them, a handful of systems mainly from ELM-ART

and AHA families, were used in a few relatively small classes. At the same time, none of the

dozens of commercial and "university-grown" Web courseware systems that are used in

hundreds of real distance courses applies adaptive and intelligent technologies. Does it mean that

research and practice in Web-based education area will never merge together?

The position of the author is the following. Web-based education itself is relatively young.

Until now different companies producing Web-based education systems were able to compete

on the market with their simple non-adaptive systems. However, a number of research level

systems have already clearly demonstrated the benefits of adaptive and intelligent technologies.

As long as the competition on the market of Web-based educational system will increase, “being

adaptive” or “being intelligent” will become an important factor for winning the customers.

Traditional Web-based education companies will start to include adaptive and intelligent

functionality. Research teams with solid experience in using adaptive and intelligent technologies

will found startup companies to bring their technology to the market. The first technologies to be

used in commercial systems will probably be sequencing technologies (page sequencing and

question sequencing) since they match very well to the current structure of Web-based education

systems. Next will come the turn of adaptive navigation support and model matching. Problemsolving

support technologies will stay on research level for longer, though we could expect the

market debut of small Web-based tutors that are aimed to support teaching a fragment of some

subject. I hope that the next five years will show us a number of examples of commercial-level

adaptive and intelligent systems as well as many new and exciting developments on the research

level.

System Ref. Adaptive

sequencing

Adaptive

navigation

support

Problem

solving

support

Intelligent

solution

analysis

Adaptive

presentation

ELM-ART [12] Page Annotation Partial Server Some

ELM-ART-II [53] Course,

tests

Annotation Partial Server Some

PATInterBook

[11] Page,

remedial

Annotation Partial Server Some

VC Prolog

Tutor

[39] Task,

remedial

Server

Table 1

hypermedia and ITS functionality

Adaptive and intelligent technologies in Web-based educational systems that combine adaptive

System Ref. Adaptive sequencing Adaptive

navigation

support

Adaptive

presentation

InterBook [9] Page Annotation Some

AST [46] Course Annotation Some

ADI [43] Course

(knowledge+interests)

Annotation Some

Adaptive and intelligent technologies for web-based education

roup.

7

Adaptive collaboration support is a very new adaptive technology which was developed

within last 5 years along with development of networked educational systems. The goal of

adaptive collaboration support is to use system's knowledge about different students to form a

matching group for different kinds of collaboration. The pioneering non-WBE (i.e., non-Web, or

non-educational) examples of adaptive collaboration support are known for already a few years.

These examples include forming a group for collaborative problem solving at a proper moment

of time [25; 26] or finding the most competent peer to answer a question about a topic (i.e.

finding a person with a model showing good knowledge of this topic) [31]. Less than two years

ago Brusilovsky [8] predicted that adaptive collaboration support will become a popular

technology. This prediction came true almost immediately. Now we can list already several real

examples of adaptive collaboration support in WBE context. The group from University of

Saskatchevan has extended their original workplace-oriented peer-help technology developed for

PHelpS system [21; 31] to the WBE context in their Intelligent Helpdesk system [22]. Another

similar system was developed and evaluated in the University of Central Florida [32]. In addition

to that, the group in the University of Duisburg known for their pioneering work on adaptive

collaboration support [25] have recently suggested a complete framework for implementation of

intelligent support techniques for distributed internet-based education. This framework can

naturally support their original adaptive collaboration support techniques and provides a

framework for exploring other model matching techniques.

Intelligent class monitoring is also based on the ability to compare records of different

students. However, instead of searching for a match, it search for a mismatch. The goal is to

identify the students who have learning records essentially different from those of their peers.

These students may be different from others in many ways. They cold be progressing too fast, or

too slow, or simply have accessed much less material than others. In any case, these students

need teacher's attention more than others - to challenge those who can, to provide more

explanations for those who can't, and to push those who procrastinate. In a regular classroom the

teacher can simply track students attendance and activity to find students who need special

attention. In a Web-based classroom, the teacher in the best case has only logging data - tables

with numbers which are very hard to grasp. At the same time, the need to identify a small subset

of students who need help more than others is more important. In WBE context, communication

between teacher and students is usually more time consuming and a distance teacher simply can't

individually address more than a small subset of the class. The system HyperClassroom [36]

provides an interesting example of using fuzzy mechanisms to identify deadlocked students in a

WBE classroom. At the time of writing, it is the only example of the intelligent class monitoring

technology known to the author.

Adaptive and intelligent technologies for web-based education

comparison with two-level sequencing in most ITS): the best page is simply selected from the

6

set of acceptable pages using some heuristics. We refer to this way of sequencing as page

sequencing. InterBook and ELM-ART provide good examples of this technology. However, the

difference between these two technologies starts to disappear in the Web context. Web-based

ITS systems are naturally moving to hypermedia platform representing at least some part of the

learning material as a hyperspace. As long as some type of educational material (presentations,

problems, and questions) is represented as a set of nodes in hyperspace, sequencing of it

becomes indistinguishable from direct guidance. To stress this similarity we have represented

adaptive sequencing and adaptive navigation support with direct guidance in the same column of

the tables.

The most popular form of ANS on the Web is annotation. It was used first in ELM-ART

[12] and since that applied in all descendants of ELM-ART such as InterBook, AST, ADI, ACE,

and ART-Web as well as in some other systems such as WEST-KBNS and KBS HyperBook.

ELM-ART and InterBook also use adaptive navigation support by sorting. Another popular

technology is hiding and disabling (a variant of hiding that keeps link visible but does not let the

user to proceed to the page behind the link if this page is not ready to be learned). The options

are either to make the link completely non-functional (nothing happens when the user clicks on

it) as implemented, for example, the Remedial Multimedia System [4] or to show the user a list

of pages to be read before the goal page as done in Albatros [29]. Tables 1 and 2 list all major

systems that use adaptive navigation support and indicates the type of adaptation.

The goal of

page to the user's goals, knowledge and other information stored in the user model. In a system

with adaptive presentation, the pages are not static, but adaptively generated or assembled from

pieces for each user. For example, with several adaptive presentation techniques, expert users

receive more detailed and deep information, while novices receive more additional explanation.

Adaptive presentation is very important in WWW context where the same "page" has to suit to

very different students. Only two Web-based AES implement full-fledged adaptive presentation:

PT [28] and AHA [16]. Both these systems apply a flexible but low-level conditional text

technique. Some other systems use adaptive presentation is special contexts. Medtec [19] is able

to generate adaptive summary of book chapters. MetaLinks can generate a special preface to a

content page depending on where the student came from to this page. ELM-ART, AST,

InterBook and other descendants of ELM-ART use adaptive presentation to provide adaptive

insertable warnings about the educational status of a page. For example, if a page is not ready to

be learned, ELM-ART and AST insert a textual warning at the end of it and InterBook inserts a

warning image in a form of a red bar. A very interesting example of adaptive presentation is

suggested in WebPersona project [3] where an individualized presentation of information in an

educational hypertext is performed by a life-like agent.

the adaptive presentation technology is to adapt the content of a hypermedia

2.3 Web-inspired technologies in Web-based education

The last group of technologies is probably the most exciting one since these technologies

has almost no roots in pre-internet educational systems. Currently this group include only one

technology. We call this technology

because the essence of this technology is the ability to analyze and match student models of

many students at the same time. Traditional adaptive and intelligent educational systems has no

opportunity to explore this technology since they usually work with one student (and one student

model) at a time. On the contrary, in the WBE context this opportunity happens naturally

because student records are usually stored centrally on a server (at least for administrative

reasons). It provides an excellent framework for developing various adaptive and intelligent

technologies that can make some use of matching student models of different students. So far,

we have identified two examples of student model matching, which we call

collaboration support

other and probably could be considered as different technologies within the student model

matching group.

student model matching (or simply model matching)adaptiveand intelligent class monitoring. These examples quite differ from each

The social shaping of technology

5

functionality is implemented in Java and works on the client side, and another part works on the

server side. The parts communicate over the Internet. While the pure Java solution looks simpler

(just a new language to build an AIES), the client-server architecture offers a more attractive

choice for developing Web-based tutors. It is a definite choice for porting a standalone

interactive tutor on the Web. D3-WWW-Trainer [20] and AlgeBrain [1] demonstrate how to reuse

the intelligent functionality of an earlier standalone tutor by changing it to a server-side

application and developing a relatively thin "brainless" Java client that implements interface

functions and communicates with an intelligent server. Event relatively small newly implemented

interactive tutors such as ADIS [50] and ILESA [30], which could be easily implemented in pure

Java, can benefit from client-server architecture for such reasons as central student modeling.

Finally, an overhead of the client-server approach (the need to have a distributed system) is not

very big since Java naturally supports several ways of client-server communications -

HTTP/CGI, sockets, or RMI/CORBA. We think, that the client-server architecture will become

very popular in the coming years as a standard way of implementing Web-based interactive

tutors and a way to implement all kinds of highly interactive Web-based AIES. We already see

examples of using it for implementing pen-based interface in WITS-II [27] and an animated

pedagogic agent Vincent in TEMAI [38].

2.2 Adaptive hypermedia technologies in Web-based education

Adaptive hypermedia is a relatively new research area [7]. Adaptive hypermedia systems

apply different forms of user models to adapt the content and the links of hypermedia pages to

the user. We distinguish two major technologies in adaptive hypermedia: adaptive presentation

and adaptive navigation support. Education always was one of the main application areas for

adaptive hypermedia. A number of standalone (i.e., non-Web-based) adaptive educational

hypermedia systems was built between 1990 and 1996. First Web-based AIES that use adaptive

hypermedia technologies were reported in 1996 [12; 17]. Since that the Web has become the

primary platform for developing educational adaptive hypermedia systems.

The goal of the

hyperspace orientation and navigation by changing the appearance of visible links. Adaptive

navigation support (ANS) can be considered as a generalization of curriculum sequencing

technology in a hypermedia context. It shares the same goal - to help students to find an "optimal

path" through the learning material. At the same time, adaptive navigation support has more

options than traditional sequencing: it can guide the students both directly and indirectly. In a

WWW context where hypermedia is a basic organizational paradigm, adaptive navigation

support can be used very naturally and efficiently. There are several known ways to adapt the

links [7]. Two examples of ANS-based standalone systems are ISIS-Tutor [10] with adaptive

hiding and adaptive annotation and Hypadapter [24] with adaptive hiding and adaptive sorting.

The three ways that are most popular in Web-based AIES are direct guidance, adaptive link

annotation, and adaptive link hiding.

Direct guidance implies that the system informs the student which of the links on the current

page will drive him or her to the "best" page in the hyperspace (which page is "best" is decided

on the basis of student's current knowledge and learning goal). Often, if a link to the next best

page is not presented on the current page, the system can generate a dynamic "next" link. As we

can see, adaptive navigation support with direct guidance is almost equivalent to curriculum

sequencing technology. There are some differences though (in addition to the different origin). A

page suggested by a direct guidance technology is always a page of the existing hyperspace. The

student usually could reach this page in one or several steps without the system guidance. The

guidance just helps the student to realize that this page is "best" and to get there fast. In an ITS

with adaptive sequencing a "page" with next best task or presentation could be completely

generated from system's knowledge, thus the student has no ways to get to this material others

than using sequencing. Also, direct guidance usually applies a one level sequencing mechanism

(in comparison with two-level sequencing in most ITS): the best page is simply selected from the

adaptive navigation support technology is to support the student in

The social construction of technological systems: New directions in the sociology and history of

4

Interactive problem solving support

Instead of waiting for the final solution, this technology can provide a student with intelligent

help on each step of problem solving. The level of help can vary: from signaling about a wrong

step, to giving a hint, to executing the next step for the student. The systems which implement

this technology (often referred to as

understand them, and use this understanding to provide help and to update the student model.

The classic example is the LISP-TUTOR [2]. This technology is also represented by a number

of Web-based AIES (Tables 1 and 3).

is a more recent and a move powerful technology.interactive tutors) can watch the actions of the student,

The example-based problem solving

helping students to solve new problems not by articulating their errors, but by suggesting them

relevant successful problem solving cases from their earlier experience (it could be examples

explained to them or problems solved by them earlier). An example is ELM-PE [51]. In the Web

context, this technology is implemented in ELM-ART [12] and ELM-ART-II [53].

In the area of traditional ITS, the interactive problem solving support technology absolutely

dominates. Interactive problem solving support is an ultimate goal of almost any ITS, while

intelligent analysis of student solutions is often considered imperfect (and example based

problem solving support is too rare to consider as a competitor). Again, the Web context changes

the situation. Both intelligent analysis of student solutions and example based problem solving

support appears to be very natural and useful in Web context. Both technologies are passive

(works by student request) and can be relatively easy implemented on the Web using a CGI

interface. Moreover, an old standalone AIES, which uses these technologies, could be relatively

easy ported to the Web by implementing a CGI gateway to the old standalone program. It is not

surprising that these technologies were among the first implemented on the Web. An important

benefit of these two technologies in the Web context is their low interactivity: both usually

require only one interaction between browser and server for a problem solving cycle. This is very

important for the case of slow Internet connection. These technologies can provide intelligent

support when a more interactive technology will be hardly useful. Currently, these technology

dominates in Web context over more powerful and interaction hungry interactive problem

solving support.

Interactive problem solving support technology is the last ITS technology migrated to the

Web. The problem here is that the "fast-track" approach of implementing Web-based ITS

(developing a CGI interface to an older standalone ITS) used in pioneer systems does not work

properly for this technology. It could be well illustrated by the PAT-Online system [41], which

was probably the first trial to implement interactive problem solving support on the Web. This

system uses a form-based CGI-AppleScript interface to a standalone Practical Algebra Tutor

(PAT) system. Since CGI interface is passive, the Web version of the system had to provide a

"submit" button for the student to get the feedback from the system. Naturally, it also added

another feature, which was essential for students with a slow Internet connection: a possibility to

request a feedback once after performing several problem solving steps. As a result, PAT-Online

moved to the category of an intelligent problem analyzers, more exactly, to a subcategory of

analyzers that are capable to analyze incomplete solutions (ELM-ART also belongs to this

subcategory). The intelligent analyzers of this subcategory can be placed between traditional

analyzers and interactive tutors (in Tables 1 and 3 they are marked with keyword "partial",

however, they can't be considered as real interactive tutors).

A real interactive tutor is expected to be not only interactive, but also active. It should not

sleep from one help request to another, but instead should be able to monitor what the student is

doing and instantly react to errors. It simply can't be implemented with the traditional server-side

CGI interactivity and requires client-side interactivity based on Java. Java technology has

matured very recently. Two years ago the review [8] named it as a prospective platform for Webbased

AIES and mentioned only three Java-based systems. Now Java provides a reliable solution

for Web-based interactive tutors. To be more exact, Java offers two different solutions. One

solution is a tutor implemented completely in Java. It could be a Java applet working in a

browser, or a Java application. Another solution is a distributed client-server tutor where a part of

technology is the newest one. This technology is

Adaptive and intelligent technologies for web-based education

3

Prolog Tutor, and Remedial Multimedia System) can perform passive remedial sequencing.

Among active sequencing systems, only a handful of systems such as ELM-ART-II, AST, ADI,

ART-Web, ACE, KBS-Hyperbook, and ILESA are able to perform intelligently both high and

low level sequencing. Others, like Manic, leave a choice of activity within a topic to the user. Vice

versa, some systems, like Medtec, leave a choice of a topic to the user but can generate an

adaptive sequence of problems within the topic. Most of the systems supports sequencing with

fixed learning goal (equals to the whole course). Only a few systems support adjustable learning

goals enabling a teacher (as in DCG) or a student (as in InterBook and KBS Hyperbook) to

select an individual goal. The student can choose a goal as a subset of domain concepts

(InterBook) or a project (KBS Hyperbook).

Active sequencing in most of the systems is driven by the students knowledge (more exactly,

by the difference between student's knowledge and global goal). A few systems and projects,

however, experiment with the use of students’ preferences on the type and media of available

learning material to drive sequencing of tasks within a topic [14; 15; 45]. Two interesting cases

of sequencing could be found in DCG and SIETTE systems. DCG [49] can perform advanced

sequencing of educational material adapted to a learning goal. However, the sequencing is

performed before students start working with the system producing a static Web-based course.

SIETTE [40] is an example of a Web-based adaptive testing system. The only kind of learning

material it possesses is questions. The only thing it can do is to generate an adaptive sequence of

questions to assess student's knowledge. Systems like SIETTE are incomplete by their nature

and have to be used as components in distributed Web-based AIES.

While curriculum sequencing could be considered as the oldest ITS technology (it was

implemented in almost all first ITS), for about 20 years it was a Cinderella among other

technologies. Very little attention was devoted to it. Mainstream ITS research were centered

around problem solving support technologies (which will be analyzed below). Problem solving

support was considered as a main duty of an ITS, while delivery and sequencing of education

material was though to be performed outside the system (usually, by a human teacher). Naturally,

almost no ITS includes educational material itself (other than a set of problems). The situation

with Web-based AIES is very different. In the context of Web-based education a solid amount

of educational material (usually structured as a hyperspace) is one of the main attractions of an

educational system. In this context (with its "lost in hyperspace" problem), curriculum

sequencing technology becomes very important to guide the student through the hyperspace of

available information. This technology is also natural and easy to implement on the Web: all

knowledge could be located on the server and all sequencing could be done by a CGI-script. It's

not surprising that, it is not only the oldest, but also the most popular technology of Web-based

AIES.

2.1.2 Problem solving support technologies

As it is mentioned above, for many years, problem solving support was considered as a main

duty of an ITS system and a main value of an ITS technology. We have identified three problem

solving support technologies: intelligent analysis of student solutions, interactive problem solving

support, and example-based problem solving support. All these technologies can help a student

in a process of solving an educational problem, but they do it by different ways.

Intelligent analysis of student solutions

problems no matter how these answers were obtained. To be considered as intelligent, a solution

analyzer has to decide whether the solution is correct or not, find out what exactly is wrong or

incomplete, and possibly identify which missing or incorrect knowledge may be responsible for

the error (the last functionality is referred as knowledge diagnosis). Intelligent analyzers can

provide the student with extensive error feedback and update the student model. The classic

example is PROUST [Johnson, 1986 #681. As it could be seen from the Tables 1 and 3, a

number of Web-based AIES implement intelligent analysis of student solutions.

deals with students' final answers to educational

Adaptive and intelligent technologies for web-based education

AIES:

intelligent technologies applied in Web-based AIES systems were directly adopted from either

the ITS area or the adaptive hypermedia area. As long as Web-based AIES research get more

mature, it will produce original technologies inspired by the Web context. At least one of these

Web-inspired technologies could already be identified (model matching). This section provides a

review of existing technologies grouped by its origin. For each technology we list existing Webbased

AIES and projects, which implements variations of this technology and discuss the ways

to implement it on the Web.

intelligent tutoring systems (ITS) and adaptive hypermedia systems. Most of adaptive and

2.1 ITS technologies in Web-based education

Intelligent tutoring systems is a traditional area of research that investigates problems of

developing AIES [13]. The goal of various ITS is the use the knowledge about the domain, the

student, and about teaching strategies to support flexible individualized learning and tutoring. A

review of existing intelligent tutoring systems performed by the author in 1990 helped to identify

three core ITS technologies: curriculum sequencing, intelligent analysis of student's solutions,

and interactive problem solving support. All these technologies were implemented in numerous

ITS. Since 1990, only one new technology (example-based problem solving support) was added

to the set to classify a functionality that was not covered by the core three. While the proposed

set of ITS technologies could be considered subjective and incomplete, it turned out to be very

useful for classifying existing Web-based AIES. Web-based AIES that use traditional ITS

technologies are usually called Web-based ITS. First Web-based ITS were reported in 1995-

1996 [6; 12; 34; 37]. These systems still constitute a rather small stream inside the ITS area.

2.1.1 Curriculum sequencing

The goal of the

technology) is to provide the student with the most suitable individually planned sequence of

knowledge units to learn and sequence of learning tasks (examples, questions, problems, etc.) to

work with. In other words, it helps the student to find an "optimal path" through the learning

material. The classic example is the BIP system [5]. There are two essentially different kinds of

sequencing: active and passive. Active sequencing implies a

concepts or topics to be mastered). Systems with active sequencing can build the best individual

path to achieve the goal. Passive sequencing (which is also called

technology and does not require an active learning goal. It starts when the user is not able to

solve a problem or answer a question (questions) correctly. Its goal is to offer the user a subset

of available learning material, which can fill the gap in student's knowledge of resolve a

misconception. For active sequencing systems, it makes sense to distinguish systems with fixed

and adjustable learning goal. Most of existing systems can guide their students to the fixed

learning goal - the whole set of domain concepts. A few systems with adjustable learning goal let

a teacher or a student to select a subset of the whole set of concepts as the current learning goal.

In most of ITS systems with sequencing it is possible to distinguish two levels of sequencing:

high and low. High-level sequencing or

next concept, set of concepts, topic, or lesson to be taught. Low-level sequencing or

sequencing

and low level sequencing are often performed by different mechanisms. In many ITS systems

only one of these two mechanisms are intelligent, for example, a lesson is selected by a student,

while learning tasks within this lesson are adaptively selected by the system. Some systems can

only manipulate the order of task of one particular kind: usually problems or questions. In this

case it could be also called problem or question sequencing.

Sequencing is currently the most popular technology in Web-based AIES. Almost all kinds

of sequencing

curriculum sequencing technology (also referred to as instructional planninglearning goal (a subset of domainremediation) is a reactiveknowledge sequencing determines next learning subgoal:taskdetermines next learning task (problem, example, test) within current subgoal. High

Adaptive and Intelligent Technologies for Web-based Education

Peter Brusilovsky

Carnegie Technology Education and

Human-Computer Interaction Institute

Carnegie Mellon University

Pittsburgh, PA 15213, USA

plb@cs.cmu.edu

Abstract:

context of Web-based distance education. We analyze what kind of technologies are

available right now, how easy they can be implemented on the Web, and what is the

place of these technologies in large-scale Web-based education.

The paper provides a review of adaptive and intelligent technologies in a

1 Introduction

Web-based education (WBE) is currently a hot research and development area. Benefits of

Web-based education are clear: classroom independence and platform independence. Web

courseware installed and supported in one place can be used by thousands of learners all over the

world that are equipped with any kind of Internet-connected computer. Thousands of Web-based

courses and other educational applications have been made available on the Web within the last

five years. The problem is that most of them are nothing more than a network of static hypertext

pages. A challenging research goal is the development of advanced Web-based educational

applications that can offer some amount of adaptivity and intelligence. These features are

important for WBE applications since distance students usually work on their own (often from

home). An intelligent and personalized assistance that a teacher or a peer student can provide in a

normal classroom situation is not easy to get. In addition, being adaptive is important for Webbased

courseware because it has to be used by a much wider variety of students than any

"standalone" educational application. A Web courseware that is designed with a particular class

of users in mind may not suit other users.

Since the early days of the Web, a number of research teams have implemented different

kinds of adaptive and intelligent systems for on-site and distance WBE. The goal of this paper is

to provide a brief review of the work performed so far in his area. The review is centered on

different adaptive and intelligent

compatibility with earlier papers on adaptive hypermedia [7] and Web-based ITS [6]. By

adaptive and intelligent technologies we mean essentially different ways to add adaptive or

intelligent functionality to an educational system. A technology usually could be further

dissected into finer grain techniques and methods, which corresponds to different variations of

this functionality and different ways of its implementation. In the next section we analyze what

kind of technologies are available right now, and how easy they can be implemented on the Web.

After that we discuss what is the place of these technologies in large-scale Web-based education.

technologies. We stay on the level of technologies to provide

2 Web-based educational systems: a review of technologies

Web-based Adaptive and Intelligent Educational Systems (AIES) are not an entirely new

kind of