A Software Engineering Framework for Context-Aware Service-Based Processes in Pervasive Environments - Advanced Design Approaches to Emerging Software Systems - page 111

Table 2. IEC for different context primitives

Implicit Existence Condition (IEC) the OCM containment hierarchy in depth-

first order. Children of context primitives
whose ECs evaluate to false are not visited
because they will be removed.

Context primitives that are not explicitly an-
notated will have implicit EC. The IEC for a
context primitive can be provided based on the
existence conditions of other context primitives
and on the syntax and semantics of the OCM.
For example, according to the ontology syntax,
an object property requires a class at each of its
ends. Thus, a reasonable choice of IEC for an
object property would be the conjunction of the
ECs of both classes. This way, removing any of the
classes will also lead to the removal of the object
property. IECs reduce the necessary annotation
effort of the user.
Table 2 shows our choice of IECs for the
context primitives. An IEC for a given primitive
is assumed based on its type.

• Removal Analysis: Removal analysis in-

volves computing IECs. The IECs can be
computed in a single additional pass after
evaluating explicit ECs. In addition, in
this step all the individuals and statements
whose subjects are included in the ele-
ments to be removed are also marked to be
removed. For example, if the Room entity
is known to be removed, all its individu-
als and all triples whose subject is of type
Room should be marked to be removed.

• Primitive Removal: In this step, primi-

tives whose ECs are false are removed.

• Applying Reasoning: In order to interpret

the remaining context information from the
perspective specified by the context feature
configuration, it is necessary to apply the
corresponding remaining rules. The result
of the reasoner will be the context product.

Context Information Generation

A context information generation process involves
computing MSs and ECs, and removing elements
whose ECs are false. The complete context prod-
uct instantiation algorithm can be summarized
as follows:

In the implemented prototype we use rule-
based inference reasoners. Different rule-based
systems provide different logical inference support
for context reasoning. To reason about ontologies,
a description logic reasoner, namely Pellet is ap-

• Evaluation of MSs and explicit ECs:

The evaluation is done while traversing