Adaptive Navigation for Learners in Hypermedia is Scaffolded
Navigation
Roland Hu¨bscher1 and Sadhana Puntambekar2
1 Department of Computer Science and Software Engineering, Auburn University, 107 Dunstan Hall, Auburn, AL
36849-5347, U.S.A.,
roland@eng.auburn.edu,
WWW home page: http://iis.cse.eng.auburn.edu/~roland/
2 School of Education, Program in Instructional Media and Technology, University of Connecticut, U-64, 249
Glenbrook Ave., Storrs, CT 06269-2004, U.S.A.,
sadhana@uconnvm.uconn.edu,
WWW home page: http://www.sp.uconn.edu/~sadhana/
Abstract. Adaptive navigation support can be of great help in large hypermedia systems supporting
learners as well as users searching for specific information. A wide variety of adaptive mechanisms have
been implemented in existing adaptive hypermedia systems that provide better and better suggestions
to the user what hyperlinks to follow. We suggest that adaptive navigation support should scaffold a
learner in an educational hypermedia system to select the appropriate links. We show that this implies
that selecting a link is an educationally relevant activity that should not always be reduced to a trivial
task by powerful adaptive mechanisms. It follows that learners require sometimes different kinds of
adaptive navigation support than users looking for information. Finally, we will suggest how to extend
current mechanisms to provide scaffolded navigation support to learners.
1 Introduction
With the advent of the World-Wide Web, hypermedia systems have become a widely used and dominating
way of providing information and educational content to users. Hypermedia has been recognized as having
great potential in providing content to learners because relationships between concepts can be made explicit
with hyperlinks, and the same material can be organized along different dimensions presenting the material
to be learned from different views [1]. The non-linear nature of hypertext environments offers opportunities as
well as certain difficulties for learning, thus making the design of such systems both complex and challenging.
The flexible nature of hypertext makes it necessary for designers to provide learners with some kind of
navigational support. Researchers believe that learning from hypertext puts a greater cognitive load on
learners [2]. Readers have to acquire specific strategies such as knowing where they are, deciding where to
go next and building a cognitive representation of the network structure, in order to cope with the specific
constraints of a non-linear presentation. Hyperlinks allow each individual learner to traverse and explore
the content in a way that fits his or her interests and learning goals at any particular time. In a hypertext
system, the reader is actively engaged in creating both meaning and structure.
The reader constantly makes decisions about where to go next. However, this added flexibility, compared
to books which are often read in a more or less linear fashion, can also cause problems seriously impacting
the pedagogical benefits of a hypermedia system. It is quite easy for the learner to lose orientation and
therefore not knowing how the current page fits into the big picture and what hyperlinked path to follow.
Adaptive hypermedia attempts to solve these problems by individualizing the presentation of the content
for each user and by providing personalized navigation support. The goal of both approaches, personalized
presentation and navigation, is to reduce the cognitive load for the user. Learners with different goals and
knowledge may be interested in different information or prefer examples from different domains. They also
may want navigate along different paths depending on their goals and prior knowledge. This is especially
important in educational systems to make sure the learner sees those content relevant to his or her learning
goals. Furthermore, adaptive navigation can help prevent users from getting lost in large systems.
Adaptive hypermedia systems (AHS) approach these problems with adaptive presentation of the infor-
mation and adaptive navigation. Adaptive presentation mechanisms decide how the information is delivered

most appropriately to the user considering factors like user expertise, media preferences, learning goals, in-
terests, and so forth stored in a user model [3]. The user model is learned by the AHS over time using various
information sources like questionnaires, quizzes, navigation behavior of the user, and others. The basic idea
of adaptive navigation is to constrain the options of where to go next to a smaller set of relevant choices
making sure the user will end up at a hypermedia page that is appropriate to his or her task.
Examples of navigation support are the “next” or “continue” links which are, according to the AHS, the
best and possibly only choice [4], sorted lists of links where earlier links are better choices [5], annotated
links reflecting some important status like ‘the user is ready to follow it or not’ or ‘it leads to redundant or
irrelevant material’ [6], or link hiding where only those links are made accessible that the system considers
relevant to the user [7].
Each of these mechanisms is based on a user model which is a description of the user’s preferences,
knowledge, skills, characteristics, learning goals, etc. and a pedagogical model that suggests what the user
should visit under what circumstances [8]. Too often, the pedagogical model is rather implicit and built into
the adaptive algorithm [9]. Much of the current research focusses on providing better user models and better
algorithms for providing appropriate navigation support by ordering the content according to some scheme.
In this paper, we would like to revisit the assumption that the better suggestions we can give to the
user, the better the system serves its purpose. Of course, on the one hand, it is obvious that the better the
adaptive mechanisms are, the better the system is. On the other hand, it is not obvious that more accurate
advice is indeed better for a user—or, more specifically, a learner—under all circumstances.
This paper is not about what can and cannot be done. It is about how adaptive navigation should be
designed for specific users, especially for learners. We do not need to propose new mechanisms—many good
ones already exist. However, we want to make explicit that efficiently getting to the right page should often
not be the overriding goal of adaptive navigation support for learners.
Consider the following hypothetical AHS. Let’s assume that the user model and adaptive mechanisms
are so good that the system can almost always provide exactly one link to the page that is best for the user
to visit next. In an information-seeking task, this is of course great because the user will find immediately
what he was looking for, possibly without ever thinking about what he was looking for at all. But is this
also a good system for a learner? We will come back to this question at the end of this paper.
2 Adaptive Navigation
Throughout this paper, we will consider two types of tasks—learning and information seeking–that cover
a large class of activities AHS are used for. A learner who uses an AHS to learn about a certain concept
executes a learning task. It is the system’s goal to make sure that the learner will understand well enough
the concept and all the necessary prerequisite concepts. On the other hand, if the user is only interested
in finding the relevant information, then we consider the task to be supported by the AHS an information-
seeking task. Many systems support both types of tasks, however, it is not necessarily the case that both
types of tasks should get the same adaptive navigation support.
Next, we will review existing adaptive hypermedia mechanisms. Then we will discuss what navigation
support information-seeking and learning tasks require. We will argue that the objectively best suggestions
the adaptive mechanism can provide are not necessarily the ones from which learners benefit the most.
2.1 Adaptive Mechanisms
Adaptive navigation deals with the problem of the user having to select a link among the many possibilities.
Often, there are too many possibilities and it is very difficult for the user to choose an appropriate link.
Adaptive navigation reduces the number of choices using various mechanims. Some of the most frequently
used ones are link ordering, link hiding, link annotation and the use of the “next” link [10]. All these
mechanisms constrain how and from how many links the user can choose to go to the next page in the
hypermedia system.
The most restrictive mechanism is the “next” or “continue” link that the AHS recommends as leading to
the most relevant next page. This allows the user to turn to the next page as easily as in a book, i.e., there is
no need to think about where to go next. If this were the only link provided, the hypermedia system would