Information Technology Reference
In-Depth Information
sis did not uncover any completely new services
compared to the project findings.
Additionally, the feasibility of the prioritization
model was evaluated by comparing how well the
categorization matched with assumed reusability
of each service candidate. Service candidates were
assumed to be reusable only if the key project
personnel were able to name either a system or
another active project where the service could
be used as well. As a result, 12 out of 17 global
service candidates were seen reusable in other
projects as well and further analysis showed that
the scope of the case project itself included a new
operative system to support all four processes from
where the remaining five global service candidates
were identified. Similarly only one out of 14 local
service candidates was estimated to be reusable
outside the project scope.
In this case study, QSE was able to provide a
service catalogue comparable with the one made by
experts with traditional means. Also the concept of
estimating the reusability of services based on their
occurrences in the process descriptions seemed
to be promising in the case project. As a result,
QSE seems to provide a promising way of doing
service elicitation in early phases of the project,
at least when used with high quality raw material.
faulty or incomplete source material will produce
a faulty and incomplete analysis. In real life, the
business process descriptions rarely cover all of
the processes and should not be used as the only
source for service engineering. A good practice
is to utilize several complementary service iden-
tification techniques. However, intensive use of
business process descriptions may motivate busi-
ness stakeholders to prepare the descriptions with
more enthusiasm.
Various different tools are available to au-
tomate some parts of the service modeling and
implementation tasks, but the most difficult part
remains manual. The elicitation of the business
needs and componentization of the business itself
is not something one can really fully automate.
However, there are tools supporting qualitative
research and grounded theory methodology in
general and these tools can be very helpful for
QSE practitioners as well.
CONCLUSION
In this chapter, we have introduced QSE, a method
for Qualitative Service Elicitation that applies the
qualitative research approach to service elicita-
tion. The use of QSE in service-oriented systems
development allows more consistent quality of
analysis and enforces developers to concentrate
on reusability aspects of services.
Based on the experiences from the case
study, it is feasible to use QSE as a systematic
and practical method for service elicitation with
results comparable to similar analysis carried out
by experts with traditional methods. The use of
business process level concepts and vocabulary
in the service candidates made it easy for require-
ment engineers to putting uses in place of usages
for the service candidates outside the case project
context. Also the idea of prioritization of services
based on how often the related concepts appear in
the business processes seemed to be promising,
although it should be enhanced to take the rela-
FUTURE RESEARCH DIRECTIONS
The strength of QSE is its generic nature; it can be
expanded with new service types with character-
istics identified in other projects, it is technology
neutral and the analyzed source material can be
virtually in any human readable format. It would
be also beneficial, if the laborious service engi-
neering work could be done by less experienced
analysts by following well-defined procedures.
One of our future research directions will be to
test how easily junior requirement engineers can
adopt QSE.
Similarly to grounded theory, QSE depends
purely on the material being analyzed. Thus,
Search WWH ::
Custom Search