Information Technology Reference
In-Depth Information
•
a conceptual context representation model
at a high level of abstraction, whose vis-
ibility could be properly propagated up to
applications;
al. 2009) are an object-role based model, a spatial
model and an ontology based model. The object-
role based approach supports various stages of
the software engineering process. Its weakness is
a 'flat' information model, i.e. all of the context
types are represented as atomic facts. The spatial
context models are well suited for context-aware
applications that are mainly location-based, like
many mobile applications. The main consideration
of the spatial context model is the choice of the
underlying location model. Relational location
models are easier to build up than geographic
location models as they provide a simple means
to map data and Global Positioning Systems
(GPS) data. The drawback is the effort that the
special context model takes to gather and keep
the location data of the context information up
to date. As an example of spatial context models,
the spatial application programming model, intro-
duced in (Meier, et al 2008), uses a small set of
predefined types for composing information and
context. The spatial programming model supports
a topographical approach for modelling a space,
i.e. the context of actors (e.g. sensors, devices,
systems and users) is modelled as a geometric
shape which is based on a sequence of coordinates.
This enables actors to independently define and
use potentially overlapping spatial context in a
consistent manner, when the relationships between
spatial objects are defined implicitly, i.e. as the
positions of the spatial objects shape within the
coordinate system. Moreover, the programming
model defines a set of types for modelling data,
i.e. the various roles that spatial objects and their
context information may have within a space. In
addition, the programming model supports context
along the dimension of time, defined by a set of
attributes. The approach is similar to ours, but
relies on a specific programming model and not
a common ontology of shared information.
•
a set of middleware services to effectively
manage the context;
•
appropriate specification models and ser-
vices to define/enforce the context-de-
pendent adaptation strategies for smart
applications.
In the following, we will focus on models for
representing the context and enforcing context-
aware adaptation strategies, while leaving context
management issues out of the scope of this section.
CONTEXT MODELS
Context has many definitions in literature. Dey
& Abowd (1999) define context as follows:
'Context is any information that can be used to
characterize the situation of an entity. An entity is
a person, place, or object that is considered to be
relevant to the interaction between the user and
the application, including the user and applica-
tions themselves.' Understanding of the context
information has heavily improved over the last ten
years. Recently published journal articles indicate
that knowledge on the specification, modeling
and usage of context information might be mature
enough for the realization of context aware smart
space applications. Typically, context information
has three dimensions; the physical, computational
and user context (Bettini et al. 2009). In order to
assist achieving interoperability on the levels that
concern the context data and change of context
(see Table 1), the context specification shall
(Preuveneers & Berbers, 2008) i) have a com-
prehensive domain coverage and terminology; ii)
be expressive and without semantic ambiguity;
iii) be processed without complexity; and iv) be
evolvable. The three types of context modeling
and reasoning approaches analyzed in (Bettini et
Search WWH ::
Custom Search