Adaptive Hypermedia: From Intelligent Tutoring Systems to Web-Based Education

Peter Brusilovsky

Carnegie Technology Education and
HCI Institute, Carnegie Mellon University

4615 Forbes Avenue, Pittsburgh, PA 15213, USA

plb@cs.cmu.edu

Abstract. Adaptive hypermedia is a new area of research at the crossroads of hypermedia and adaptive systems and. Education is the largest application area of adaptive hypermedia systems. The goals of this paper are to provide a brief introduction into adaptive hypermedia and supply the reader with an organized reading on adaptive educational hypermedia. Unlike some other papers that are centered around the current state of the field, this paper attempts, from one side, to trace the history adaptive educational hypermedia in connection with intelligent tutoring systems research and, from another side, draft its future in connection with Web-based education.

1 Introduction

Adaptive hypermedia is a relatively new direction of research on the crossroads of hypermedia and user modeling. One limitation of traditional "static" hypermedia applications is that they provide the same page content and the same set of links to all users. If the user population is relatively diverse, a traditional system will "suffer from an inability to be all things to all people". For example, a traditional educational hypermedia system will present the same static explanation and suggest the same next page to students with widely differing educational goals and knowledge of the subject. Similarly, a static electronic encyclopedia will present the same information and same set of links to related articles to readers with different knowledge and interests. Finally, a static virtual museum will offer the same "guided tour" and the same narration to visitors with very different goals and background knowledge.

Adaptive hypermedia is an alternative to the traditional "one-size-fits-all" approach in the development of hypermedia systems. Adaptive hypermedia systems build a model of the goals, preferences and knowledge of each individual user, and use this model throughout the interaction with the user, in order to adapt the hypertext to the needs of that user. For example, a student in an adaptive educational hypermedia system will be given a presentation which is adapted specifically to his or her knowledge of the subject [10], and a suggested set of most relevant links to proceed further [4]. An adaptive electronic encyclopedia will personalise the content of an article to augment the user's existing knowledge and interests [26]. A virtual museum will adapt the presentation of every visited object to the user's individual path through the museum [30].

2 What Can Be Adapted in Adaptive Hypermedia

A typical hyperdocument consists of a set of nodes or "pages" connected by links. Each page contains some local information and a number of links to related pages. Hypermedia systems can also include special navigation tools such as table of contents, index, and map that could be used to navigate to all accessible pages. What can be adapted here are the page (content-level adaptation) and the appearance and behavior of the links (link-level adaptation). In adaptive hypermedia literature they are referred respectively as adaptive presentation and adaptive navigation support .

The goal of the adaptive presentation is to adapt the content of a hypermedia page to the user's goals, knowledge and other information stored in the user model. There could be multiple reasons to use adaptive presentation. Two typical cases in the area of education are comparative explanations and explanation variants. The idea of comparative explanations is to connect new content to the existing knowledge of the student. A page can have one or more hidden comparative explanation fragments that comparing some aspects of new topic with relevant aspects of other topic [15]. For example, "while" loop in C++ language could be compared with the same construct in Pascal. Only students with relevant previous knowledge will (knowledge of Pascal loops in our example) will see the comparison. The idea of explanation variants is to use essentially different variants of explanations for users with different level of knowledge of the topic. A system can store several variants for some parts of the page content. For example, a variant prepared for a medical experts can use extensive Latin terminology, while a variant prepared for novices can use everyday names for illnesses and body parts [2; 17; 34].

The goal of adaptive navigation support is to help users to find their paths in hyperspace by adapting link presentation and functionality to the goals, knowledge, and other characteristics of an individual user. It is typically done by one of the following ways:

A number of interesting forms and applications of adaptive presentation and adaptive navigation support were developed since 1990. A more comprehensive review can be found in [3].

3 From Intelligent Tutoring Systems to Adaptive Hypermedia

Education was always the most popular application area for adaptive hypermedia systems. A number of interesting methods and techniques of adaptive hypermedia were originally developed for in various adaptive educational hypermedia systems. In turn, most of the early research on adaptive educational hypermedia was inspired by the area of intelligent tutoring systems [2; 7; 15; 17; 20; 22; 31] and were born in a trial to combine an intelligent tutoring system (ITS) and an educational hypermedia.

In the early times of ITS, most of these systems provide little or no learning material. The most important duty of an ITS was to support a student in the process of problem solving. It was assumed that the required knowledge is acquired outside of the system, for example, by attending a lecture or reading a textbook. Along with the growth of computer capabilities more and more ITS developers found it reasonable to provide an ITS and a learning material in electronic form in one package. Very soon it became clear that hypertext or hypermedia provides the best option for organizing on-line learning material. A combination of an ITS and a learning material organized as hypermedia was a natural starting point for the research on adaptive educational hypermedia. A number of research groups has independently realized that a hypermedia system coupled with an ITS can offer more functionality than a traditional static educational hypermedia.

Adaptive presentation came first. Adaptive presentation was the natural and the simplest way to make the hypermedia component of the system to use some knowledge about individual students represented in a student model of ITS. A number of adaptive presentation methods and techniques were explored in early projects. In particular, comparative explanations method was used in Lisp-Critic [15] and explanation variants method was used in Lisp-Critic [15], Anatom-Tutor [2], and SYPROS [17].

In our ITS for programming domain ITEM/IP [6] we have explored several adaptive presentation methods including explanation variants. Our goal was to achieve a gradual transformation of the learning material from an explanation-rich textbook to a concise manual. We have developed a simple but powerful technique known as "conditional text". With this technique, all information about that could be presented on a page is divided into several chunks of texts. Each chunk is associated with a condition on the state of user knowledge stored in the user model. When assembling a page for presentation the system selects only the chunks with true condition. This technique is a low-level technique (it requires some "programming" work from the author to set all the required conditions) but it is also very flexible. By choosing appropriate conditions on the knowledge level of the current concept and related concepts represented in the user model we were able to implement several adaptive presentation methods. A simple example is hiding chunks that contain additional explanations if the user's knowledge of the current concept is good enough, or turning on a chunk with comparative explanations if the corresponding related concept is already known. This conditional text technique was later independently developed by Kay and Kummerfeld [21] and De Bra [10] and became quite popular in Web-based adaptive systems.

The work on adaptive navigation support in educational hypermedia was influenced by research on curriculum sequencing. Curriculum sequencing is one of the oldest ITS technologies. The goal of the curriculum sequencing 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. Early ITS with curriculum sequencing were able to sequence only one kind of learning tasks - problems to for the student to solve [1; 25]. More recent ITS such as ITEM/IP [8], TOBIE [38] and ECAL [14] were able to deal with more rich educational material. The early work on adaptive navigation support in educational hypermedia was simply a trial to apply the ideas of sequencing in a hypermedia context. From the first sight, a dynamic linear sequence of learning tasks produced by a sequencing-based ITS and a static network of educational hypermedia pages looks like two contradictory approaches to organizing access to the learning material. However, these approaches are really complementary. The key is that a typical sequencing engine can do more than just selecting the "next best" task. On the way to the "best", such an engine can usually classify all available tasks into non-relevant and relevant candidates. For example, a task can be considered non-relevant if it was already completed in the past or if it is not ready to be learned due to the lack of prerequisite knowledge and experience. After excluding non-relevant tasks a sequencing engine use some approach to pick up the best of relevant tasks. In a hyperspace of learning material where each learning task is represented by a separate page an ability to distinguish "ready", "not-ready", or "best" tasks is a direct precondition for adaptive navigation support.

In our systems ITEM/PG [6] and ISIS-Tutor [5] we explored several ways of adaptive navigation support. We have used direct guidance in the form of "teach me" button to provide a one-click access to the next best task. We have used adaptive annotation to color-code the links to "ready", "not-ready", and "already learned" tasks. In one of the versions of ISIS-Tutor we have applied adaptive link removal to remove all links to not-ready tasks. From our point of view a sequencing-based adaptive navigation support in educational hypermedia is "best of both worlds". Choosing next task in an ITS with sequencing is based on machine intelligence. Choosing next task in a traditional hypermedia is based on human intelligence. Adaptive navigation support is an interface that can integrate the power of machine and human intelligence: a user is free to make a choice while still seeing an opinion of an intelligent system. From this point of view we can speculate that adaptive navigation support is a natural way to add some intelligence to adaptive hypermedia system. It is not surprising that several research groups have independently developed major adaptive navigation support techniques such as direct guidance [42], hiding [10; 31], and annotation [12].

4 Adaptive Hypermedia for Web-based Education

The year of 1996 could be considered a turning point in adaptive hypermedia research. The key factor here is the rapid increase in the use of the Word Wide Web. The Web, with its clear demand for adaptivity, served to boost adaptive hypermedia research, providing both a challenge and an attractive platform.

We know only four Web-based adaptive educational hypermedia systems developed by 1996: ELM-ART [7], InterBook [4], PT [21], and 2L670 [11]. These "classic" systems have influenced a number of more recent systems. The Web platform enabled these systems to live much longer than similar pre-Web systems and influence a number of more recent systems. In particular, ELM-ART gave a start to a whole tree of systems including InterBook, AST, ADI, ART-WEB, and ACE. It is not surprising that all adaptive educational hypermedia systems developed since 1996 are Web-based systems. Examples are: Medtech [13], AST [36], ADI [33], HysM: [23], AHM [32], MetaLinks [27], CHEOPS [28], RATH [19], TANGOW [9], Arthur [16], CAMELEON [24], KBS-Hyperbook [18], AHA! [10], SKILL [29], Multibook [37], ACE [35], ART-Web [41].

The introduction of the Web has impacted not only on the number of adaptive educational hypermedia systems, but also on the type of systems being developed. All the early systems were essentially lab systems, built to explore some new methods, which used adaptivity in an educational context. In contrast, a number more recent systems provide complete frameworks and even authoring tools for developing Web-based courses. The appearance of a number of authoring tools is not only indicative of the maturity of adaptive educational hypermedia, but also a response to a Web-provoked demand for user-adaptive distance education courses.

Existing adaptive hypermedia frameworks such as InterBook, ART-Web, ACE, AHA!, SKILL, MetaLinks or Multibook are getting strikingly close to commercial tools for developing Web-based courses such as WebCT [40] or TopClass [39]. Developers of adaptive hypermedia frameworks are clearly interested in making their systems suitable for handling real Web courses. From another side, developers of commercial course management systems are becoming interested in adaptive and personalized systems. In this situation we could hope that adaptive hypermedia technology that was originally developed inside the area of ITS will soon be used in commercial-strength Web-based systems to deliver thousands of real world courses to students all over the world.

References

1. Barr, A., Beard, M., and Atkinson, R. C.: The computer as tutorial laboratory: the Stanford BIP project. International Journal on the Man-Machine Studies 8, 5 (1976) 567-596

2. Beaumont, I.: User modeling in the interactive anatomy tutoring system ANATOM-TUTOR. User Modeling and User-Adapted Interaction 4, 1 (1994) 21-45

3. Brusilovsky, P.: Methods and techniques of adaptive hypermedia. User Modeling and User-Adapted Interaction 6, 2-3 (1996) 87-129

4. Brusilovsky, P., Eklund, J., and Schwarz, E.: Web-based education for all: A tool for developing adaptive courseware. Computer Networks and ISDN Systems. 30, 1-7 (1998) 291-300

5. 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

6. Brusilovsky, P., Pesin, L., and Zyryanov, M.: Towards an adaptive hypermedia component for an intelligent learning environment. In: Bass, L. J., Gornostaev, J. and Unger, C. (eds.) Human-Computer Interaction. Lecture Notes in Computer Science, Vol. 753. Springer-Verlag, Berlin (1993) 348-358

7. 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

8. Brusilovsky, P. L.: A framework for intelligent knowledge sequencing and task sequencing. In: Frasson, C., Gauthier, G. and McCalla, G. I. (eds.) Intelligent Tutoring Systems. Springer-Verlag, Berlin (1992) 499-506

9. Carro, R. M., Pulido, E., and Rodrígues, P.: TANGOW: Task-based Adaptive learNer Guidance on the WWW. Computer Science Report, Eindhoven University of Technology, Eindhoven (1999) 49-57

10. De Bra, P. and Calvi, L.: AHA! An open Adaptive Hypermedia Architecture. The New Review of Hypermedia and Multimedia 4 (1998) 115-139

11. De Bra, P. M. E.: Teaching Hypertext and Hypermedia through the Web. Journal of Universal Computer Science 2, 12 (1996) 797-804, available online at http://www.iicm.edu/jucs_2_12/teaching_hypertext_and_hypermedia

12. de La Passardiere, B. and Dufresne, A.: Adaptive navigational tools for educational hypermedia. In: Tomek, I. (ed.) Computer Assisted Learning. Springer-Verlag, Berlin (1992) 555-567

13. Eliot, C., Neiman, D., and Lamar, M.: Medtec: A Web-based intelligent tutor for basic anatomy. In: Lobodzinski, S. and Tomek, I. (eds.) Proc. of WebNet'97, World Conference of the WWW, Internet and Intranet, Toronto, Canada, AACE (1997) 161-165

14. Elsom-Cook, M. T. and O'Malley, C.: ECAL: Bridging the gap between CAL and intelligent tutoring systems. Computers and Education 15, 1 (1990) 69-81

15. Fischer, G., Mastaglio, T., Reeves, B., and Rieman, J.: Minimalist explanations in knowledge-based systems. In: Proc. of 23-th Annual Hawaii International Conference on System Sciences, Kailua-Kona, HI, IEEE (1990) 309-317

16. Gilbert, J. E. and Han, C. Y.: Arthur: Adapting Instruction to Accommodate Learning Style. In: Bra, P. D. and Leggett, J. (eds.) Proc. of WebNet'99, World Conference of the WWW and Internet, Honolulu, HI, AACE (1999) 433-438

17. Gonschorek, M. and Herzog, C.: Using hypertext for an adaptive helpsystem in an intelligent tutoring system. In: Greer, J. (ed.) Proc. of AI-ED'95, 7th World Conference on Artificial Intelligence in Education, Washington, DC, AACE (1995) 274-281

18. Henze, N., Naceur, K., Nejdl, W., and Wolpers, M.: Adaptive hyperbooks for constructivist teaching. Künstliche Intelligenz , 4 (1999) 26-31

19. Hockemeyer, C., Held, T., and Albert, D.: RATH - A relational adaptive tutoring hypertext WWW-environment based on knowledge space theory. In: Alvegård, C. (ed.) Proc. of CALISCE'98, 4th International conference on Computer Aided Learning and Instruction in Science and Engineering, Göteborg, Sweden (1998) 417-423

20. Hohl, H., Böcker, H.-D., and Gunzenhäuser, R.: Hypadapter: An adaptive hypertext system for exploratory learning and programming. User Modeling and User-Adapted Interaction 6, 2-3 (1996) 131-156

21. Kay, J. and Kummerfeld, B.: User models for customized hypertext. In: Nicholas, C. and Mayfield, J. (eds.): Intelligent hypertext: Advanced techniques for the World Wide Web. Lecture Notes in Computer Science, Vol. 1326. Springer-Verlag, Berlin (1997)

22. Kay, J. and Kummerfeld, R. J.: An individualised course for the C programming language. In: Proc. of Second International WWW Conference, Chicago, IL (1994), available online at http://www.ncsa.uiuc.edu/SDG/IT94/Proceedings/Educ/kummerfeld/kummerfeld.html

23. Kayama, M. and Okamoto, T.: A mechanism for knowledge-navigation in hyperspace with neural networks to support exploring activities. 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) 41-48

24. Laroussi, M. and Benahmed, M.: Providing an adaptive learning through the Web case of CAMELEON: Computer Aided MEdium for LEarning on Networks. In: Alvegård, C. (ed.) Proc. of CALISCE'98, 4th International conference on Computer Aided Learning and Instruction in Science and Engineering, Göteborg, Sweden (1998) 411-416

25. McArthur, D., et al.: Skill-oriented task sequencing in an intelligent tutor for basic algebra. Instructional Science 17, 4 (1988) 281-307

26. Milosavljevic, M.: Augmenting the user's knowledge via comparison. In: Jameson, A., Paris, C. and Tasso, C. (eds.) User Modeling. Springer-Verlag, Wien (1997) 119-130

27. Murray, T., Condit, C., and Haugsjaa, E.: MetaLinks: A preliminary framework for concept-based adaptive hypermedia. In: Proc. of Workshop "WWW-Based Tutoring" at 4th International Conference on Intelligent Tutoring Systems, San Antonio, TX (1998), available online at http://www-aml.cs.umass.edu/~stern/webits/itsworkshop/murray.html

28. Negro, A., Scarano, V., and Simari, R.: User adaptivity on WWW through CHEOPS. Computing Science Reports, Eindhoven University of Technology (1998) 57-62

29. Neumann, G. and Zirvas, J.: SKILL - A scallable internet-based teaching and learning system. In: Maurer, H. and Olson, R. G. (eds.) Proc. of WebNet'98, World Conference of the WWW, Internet, and Intranet, Orlando, FL, AACE (1998) 688-693, available online at http://nestroy.wi-inf.uni-essen.de/Forschung/Publikationen/skill-webnet98.ps

30. Oberlander, J., O'Donell, M., Mellish, C., and Knott, A.: Conversation in the museum: experiments in dynamic hypermedia with the intelligent labeling explorer. The New Review of Multimedia and Hypermedia 4 (1998) 11-32

31. Pérez, T., Gutiérrez, J., and Lopistéguy, P.: An adaptive hypermedia system. In: Greer, J. (ed.) Proc. of AI-ED'95, 7th World Conference on Artificial Intelligence in Education, Washington, DC, AACE (1995) 351-358

32. Pilar da Silva, D., Durm, R. V., Duval, E., and Olivié, H.: Concepts and documents for adaptive educational hypermedia: a model and a prototype. Computing Science Reports, Eindhoven University of Technology, Eindhoven (1998) 35-43

33. Schöch, V., Specht, M., and Weber, G.: "ADI" - an empirical evaluation of a tutorial agent. In: Ottmann, T. and Tomek, I. (eds.) Proc. of ED-MEDIA/ED-TELECOM'98 - 10th World Conference on Educational Multimedia and Hypermedia and World Conference on Educational Telecommunications, Freiburg, Germany, AACE (1998) 1242-1247

34. Specht, M. and Kobsa, A.: Interaction of domain expertise and interface design in adaptive educational hypermedia. Computer Science Report, Eindhoven University of Technology, Eindhoven (1999) 89-93

35. Specht, M. and Oppermann, R.: ACE - Adaptive Courseware Environment. The New Review of Hypermedia and Multimedia 4 (1998) 141-161

36. Specht, M., Weber, G., Heitmeyer, S., and Schöch, V.: AST: Adaptive WWW-Courseware for Statistics. In: Brusilovsky, P., Fink, J. and Kay, J. (eds.) Proc. of Workshop "Adaptive Systems and User Modeling on the World Wide Web" at 6th International Conference on User Modeling, UM97, Chia Laguna, Sardinia, Italy, (1997) 91-95, available online at http://www.contrib.andrew.cmu.edu/~plb/UM97_workshop/Specht.html

37. Steinacker, A., Seeberg, C., Rechenberger, K., Fischer, S., and Steinmetz, R.: Dynamically generated tables of contents as guided tours in adaptive hypermedia systems. In: Proc. of ED-MEDIA/ED-TELECOM'99 - 11th World Conference on Educational Multimedia and Hypermedia and World Conference on Educational Telecommunications, Seattle, WA, AACE (1998)

38. Vassileva, J.: Dynamic CAL-courseware generation within an ITS-shell architecture. In: Tomek, I. (ed.) Computer Assisted Learning. Springer-Verlag, Berlin (1992) 581-591

39. WBT Systems: TopClass, Dublin, Ireland, WBT Systems (1999) available online at http://www.wbtsystems.com/

40. WebCT: World Wide Web Course Tools, Vancouver, Canada, WebCT Educational Technologies (1999) available online at http://www.webct.com

41. Weber, G.: ART-WEB, Trier, University of Trier (1999) available online at

42. Zeiliger, R.: Adaptive testing: contribution of the SHIVA model. In: Leclercq, D. and Bruno, J. (eds.): Item banking: Interactive testing and self-assessment. NATO ASI Serie F, Vol. 112. Springer-Verlag, Berlin (1993) 54-65