Information Technology Reference
In-Depth Information
LAOS
dictates the composition of the instances of these
models, e.g. the fact that presentation descriptions
need to be in XML, fixed by a DTD. Moreover,
XAHM is highly reliant on mathematical mod-
els, graph theory and probability computations.
The user model contains, in addition to data on
the current profile, probability distribution func-
tions that map a user over a number of profiles.
Moreover, adaptation is represented as a func-
tion defined on a three-dimensional input-output
space: the
user's behaviour
, the
technology
and
the
external environment
. Finally, the
application
domain
is composed of a
graph-based layered
model
for describing the logical structure of the
hypermedia and
XML-based models
for describing
the metadata for basic information fragments, as
well as elementary
abstract concepts
connected
via weighted, dynamically computed links for
navigation between elements (that transform into
probabilities of users actually choosing those
paths).
The main advantage of XAHM is that it is the
first attempt to create elegant mathematical model-
ling of the adaptation process; another advantages
is that of allowing the adaptation in three dimen-
sions (Cannataro,
et al.,
2001): the behaviour of
the user (i.e., preferences and activity history); the
technology dimension (operating system, internet
connection, access device, etc.); and the external
environment (weather, time-zone, geographical
location, etc.), which are not sufficiently treated
and separated in previous models. However, the
main disadvantage is that it hides adaptation and
personalization, partially in the user model (via
probability density computations), partially in
the application domain model (where weights are
probabilistically computed between navigational
elements), and finally, in the adaptation model.
This distribution of adaptation is hard to follow,
and tools based on it can be difficult to handle by
teachers, for instance. An example of a tool based
on XAHM is the Java Adaptive Hypermedia Suite
(JAHS) (Cannataro & Pugliese, 2002)
The LAOS framework (Cristea & De Mooij, 2003)
is a general framework for authoring adaptive
hypermedia, based mainly on the AHAM model,
presenting however some of the features of the
WebML language with which it shares the pre-
sentation model. It consists of a
Domain Model
(DM), a
Goal and Constraints Model
(GM), a
User Model
(UM), an
Adaptation Model
(AM)
and a
Presentation Model
(PM). LAOS differs
from other models by introducing the goal and
constraints model. This layer supports the original
aim of adaptive hypermedia from the perspec-
tive of the designer (or teacher, in educational
environments, hence pedagogic information, or
business logics for commercial sites), something
that was missing in previous models (Cristea &
De Mooij, 2003).
Furthermore, the LAOS AM model is differ-
ent from that of AHAM. The adaptation model
is based on the three layer LAG model (Cristea
& Verschoor, 2004) for authoring adaptation,
which allows different entry and reuse levels for
adaptation specification, depending on whether the
author has programming skills or not. Thus, the
initial threshold for creating adaptation is lowered.
The major difference between LAOS and
AHAM (and other models) is a higher level of
reuse, due to the clear separation of
primitive in-
formation (content)
and
presentation-goal related
information,
such as pedagogical information in
educational systems and prerequisites. For in-
stance, since prerequisites are not hard-wired in
the domain model, elements of the domain can be
used in different settings and sequences to those
initially intended. In this way LAOS facilitates
a high degree of information reuse by separating
information from its specific context. This separa-
tion is expressed by having two different models,
instead of one: a
domain model
(DM) and a
goal
and constraints model
(GM). The separation
can be understood easily if we use the following
Search WWH ::
Custom Search