Information Technology Reference
In-Depth Information
3.1 Service Knowledge Base
In the following, we formally define the service knowledge base as
.
We refer to
D
as a set of terminological concepts describing an abstract repre-
sentation of a service development methodology. With
F
, we describe a set of
factual concepts that are mapped to related terminological concepts in
D
.
KB
=
D, F
Definition 1 (Methodology Representation
D
).
We define the abstract
representation
D
as a directed graph
D
=(
T,R,E
)
with
T
representing a set
of terminological concepts
T
=
{
t
1
,...,t
n
}
,
R
denoting a set of relationships
R
=
and
E
a set of directed edges between two terminological
concepts belonging to a specific relationship such that
E
=
{
r
1
,...,r
m
}
{
e
1
,...,e
k
}
with
e
i
=(
t
o
,r
y
,t
p
)
,
0
≤
i
≤
k
,
0
≤
o, p
≤
n
,
0
≤
y
≤
m
.
Example 1 (Methodology Representation
D
).
We use the example of termino-
logical concepts as described in Section 2.1 to define the conceptual part
D
=
(
T,R,E
) of the service knowledge base
KB
D, F
. As such, the representa-
tion
D
describes a child-relationship of the concept ”Business Object Node” to
the concept ”Business Object”.
=
T
:=
{
t
1
,t
2
}
=
{
Business Object
,
Business Object Node
}
,
R
:=
{
r
1
}
{
containsBON
}
,
E
:=
{
e
1
}
with
e
1
:= (
t
1
,r
1
,t
2
)=(
Business Object
,
containsBON
,
Business Object Node
)
=
Definition 2 (Factual Concepts
F
).
We define
F
as a set of factual concepts
F
=
{
f
1
,f
2
,...,f
m
}
. We further define a mapping
Φ
:
F
→
T
,
Φ
(
f
)=
t
for
f
∈
F, t
∈
T
, such that
∀
f
∈
F
:
∃
t
∈
T
:
Φ
(
f
)=
t
. Furthermore, for each
t
∈
T
we denote the (possibly empty) subset
F
t
⊂
F
such that
∀
f
∈
F
t
:
Φ
(
f
)=
t
.
Obviously these subsets are distinct for different
t
, i.e.
∀
t
i
,t
j
∈
T
:
t
i
=
t
j
→
F
t
i
∩
F
t
j
=
∅
Example 2 (Factual Concepts
F
).
Referring to the examples of factual concepts
asshowninSection2.1,weuse
F
to represent a set of factual concepts, i.e. ”Sales
Order”, ”Purchase Order” and ”Item”. We further have distinct subsets
F
t
1
and
F
t
2
of
F
, whereas ”Sales Order” and ”Purchase Order” represent
F
t
1
and ”Item”
forms
F
t
2
. The mapping
Φ
describes the relationship of factual concepts in
F
t
1
and
F
t
2
to
T
, which practically relates ”Sales Order” and ”Purchase Order” to
”Business Object” and ”Item” to ”Business Object Node”.
F
:=
{
f
1
,f
2
,f
3
}
=
{
Sales Order
,
Purchase Order
,
Item
}
,
Φ
(
f
1
):=
t
1
→
Φ
(
Sales Order
)=
Business Object
Φ
(
f
2
):=
t
1
→
Φ
(
Purchase Order
)=
Business Object
Φ
(
f
3
):=
t
2
→
Φ
(
Item
)=
Business Object Node
F
t
1
:=
{
f
1
,f
2
}
=
{
Sales Order
,
Purchase Order
}
,
F
t
2
:=
{
f
3
}
=
{
Item
}
3.2 Service Signature Automaton
In this section, we use the notation of a non-deterministic finite automaton
(NFA) with
ε
-moves to formally define a language of accepted Enterprise Services
