Social, Personalized Lifelong Learning - E-Infrastructures and Technologies for Lifelong Learning - page 96

hyperlinks, and personalization is a user-based
adaptation. In this section we examine the most
important frameworks for personalization on the
web, in order to consider the different aspects
of adaptation and personalization on the one
hand, and on the other to select a platform on
which to base social extensions. Many adaptive
(educational) hypermedia systems have been
launched since the early 1990s; however, until
the late 1990s, there was no structural design or
standard model for learning adaptive hypermedia
systems. One of the first models designed was
the Adaptive Hypermedia Application Model
(AHAM) (De Bra et al ., 1999), followed by the
Web Modeling Language (WebML) (Ceri et al.,
2000), the Goldsmiths Adaptive Hypermedia
Model (GAHM) (Ohene-Djan, 2000), the Munich
reference model (Koch, 2001), the XML Adap-
tive Hypermedia Model (XAHM) (Cannataro et
al., 2002), the LAOS framework (Cristea & De
Mooij, 2003), and the Generic Adaptivity Model
(GAM) (De Vrieze, 2004). The goal of each of
these models is to record important concepts in
current adaptive (educational) hypermedia sys-
tems, such as the node/link structure, user model,
adaptation patterns and presentation settings. In
this section, we analyze the similarities and dif-
ferences between these models.

2. the user model , containing concepts with
attributes, used to store user preferences or
other information (such as knowledge-of or
interest-in domain model concepts); and
3. the adaptation model , which consists of
adaptation rules that use the attribute values
of concepts in the user model in order to
determine if and how to present concepts
and links from the domain model.

The main advantages of AHAM are that it is a
relatively simple model which allows for separa-
tions of concerns. The separation into layers helps
to define the main components that need to be
created by an author. However, AHAM does not
make full use of other potential advantages of the
separation into layers: for instance, reusability is
not supported. In principle, having separate layers
would allow for one domain model to be used in
different adaptation or user models. However, this
is not possible in AHAM, due to the fact that the
adaptation rules apply to concrete domain model
concepts, and cannot be re-applied to others.
Moreover, reusability would mean that authors
could be assigned different roles on each layer,
and this would speed up the development process
by enabling developers to work in parallel on the
different layers - which is not possible in AHAM
due to the interdependencies between the layers.
An example system based on AHAM is AHA!
(De Bra & Ruiter, 2001), proposed by Eindhoven
University of Technology.

The Adaptive Hypermedia
Application Model (AHAM)

AHAM (De Bra et al., 1999) is based on the
Dexter model (Halasz, 1994), a reference model
for hypertext systems. AHAM divides adaptive
(educational) hypermedia systems into three lay-
ers: the run-time layer , the storage layer and the
within-component layer , connected by the inter-
faces presentation specifications and anchoring .
The focus of AHAM is the storage layer with its
three sub-models:

The Munich Reference Model

The Munich Reference Model (Koch, 2001), de-
veloped at the Ludwig-Maximilians University of
Munich, also extends the Dexter storage layer with
user and adaptation models , and has a run-time
layer , a storage layer and a component layer . It
is very similar to AHAM, but its main differences
are (Koch, 2001) that it:

1. the domain model , consisting of a set of
concepts and concept relationships;