|
Computing in context has become a necessity
in modern and intelligent IT applications. Context is now more than
just
location. It is seen as a multi-dimensional space of environmental
aspects,
even including non-physical facets like emotions. Hence, models for
representing context have evolved from using simple key-value pairs to
using
current methods and techniques derived from artificial intelligence and
knowledge management, e.g., logic, object relationship models, and
ontologies.
Context and context-awareness are
crucial
not only for mobile and ubiquitous computing, but for spanning various
application
areas: collaborative software and web engineering, personal digital
assistants
and peer-to-peer information sharing, health care workflow and patient
control,
and adaptive games and e-Learning solutions. In these areas, context
serves as
a major source for reasoning, decision making, and adaptation, as it
covers not
only application knowledge but also environmental
knowledge.
With the introduction of intelligent
systems and automation, other aspects for IT applications arise that
are
important for the users of such systems and may be solved or supported
by
applying context. A crucial one is the understandability of an IT
system, for
explaining how solutions are found, what the system is doing, and why
it
operates a certain way. Applied methods and given advice have to be
explained,
so that the user can understand the process and agree on decisions.
Another
vital feature is to provide uncertain or blurred information, e.g.,
when using a
tracking system in situations, where either revealing the current
position or
denying access to it would spoil the activity.
Appropriate context management methods are an
important prerequisite for using contextual information. Therefore,
advanced
models, methods, and tools are needed to provide mechanisms and
techniques for
structured storage of contextual information, to provide effective ways
to
retrieve it, and to enable integration of context and application
knowledge.
|
