Information Technology Reference
In-Depth Information
Process Driven Services
are cataloged and used to identify services to be
implemented.
There are also several methods using ele-
ments from product line engineering to manage
the service specifications and production of new
service variants based on analyzed needs. In these
approaches, the services are seen as reusable ap-
plication elements, which can be used to build
new applications. (Adam & Doerr, 2008; Moon,
Hong, & Yeom, 2008)
Overall, the methods above try to identify
common process elements within the enterprise,
some utilizing also familiar patterns from other
enterprises to support the work.
Process driven SOA is a popular approach, and
business processes can be seen as an ideal source
for reusable services (Papazoglou et al., 2006;
Van Nuffel, 2007). Various methods have been
proposed to map and align services with business
processes, but the field is somewhat dispersed
with various engineering approaches and a vast
number of different business process and workflow
modeling languages.
The survey conducted by Papazoglou et al.
(2004) identified two basic classes for complex
web services: programmatic and interactive web
services. Programmatic services encapsulate
atomic business logic functionality to be used by
other applications to build new applications. In-
teractive services include the logic for interacting
with a user through the presentation service of a
web application. The logic can contain the multi-
step behavior of an interactive business process.
Patterns can be used to identify services in
generic problem areas as long as they fit into the
pre-defined scenarios. Different levels of patterns
have been proposed in the area. For example
Endrei at al. (2004) propose business patterns to
be used to identify services in common business
scenarios, while Zdyn et al. (2007) use more
primitive software patterns to build processes out
of building blocks.
SOMA (service-oriented modeling and archi-
tecture) is a software development method for
SOA-based solutions containing a set of methods
to support all phases of the SOA development (A.
Arsanjani et al., 2008). SOMA provides several
complementary methods to identify flows by
analyzing business goals, business processes, as
well as existing IT assets. However, details of the
method have not been published.
Lo et al. (2008) propose a reference catalogue
approach, which consists of two parts: a set of
reference business models and a set of business
service patterns. The needed business services
Data Driven Services
The basic idea behind the data driven SOA or
“Information as a Service” (IaaS) approach (Dan,
Johnson, & Arsanjani, 2007) is to decouple the
data and the business logic allowing systems to
share the same data and data access logic. This ap-
proach has gained a great deal of interest lately, and
also the market for IaaS tools is growing rapidly.
Forrester predicts the market to exceed Enterprise
Information Integration (EII), Enterprise Applica-
tion Integration (EAI) and replication markets in
size in the future (Forrester, 2008).
A survey by Papazgolou et al. (2004) lists three
types of informational services: content services,
information aggregation services and third-party
information syndication services.
SOMA also provides several complementary
methods to analyze information, a method called
“domain decomposition” being the most interest-
ing from our point of view (Arsanjani et al., 2008).
In this method, the enterprise is first partitioned
into functional areas, and then, business entities
are identified within the areas. Different variations
of the business entities are identified to ensure the
reusability of the service design.
Data warehousing is another approach where
data is gathered across the enterprise into a cen-
tralized database from where it can be read for
Search WWH ::
Custom Search