[Back]

A. Mazak, B. Kratzwald:
"The Weighted Decision Matrix: Tracking Design Decisions in Service Compositions";
in: "Proceedings of the 7th IEEE 7th International Conference on Service-Oriented Computing and Applications", issued by: IEEE; IEEE Computer Society, 2014, 169 - 174.

Often, there is a lack of efficient procedures in order to identify design errors when transforming customer requirements into servic-oriented solution. A well-established approach supporting this kind of transformation in classical requirements engineering is the traceability matrix which allows tracing the coverage of customer requirements by system components in general, or services in particular. The matrix shows which services realize which customer requirements. However, the matrix does not show the relevance of a service in order to meet a certain quality aspect (e.g. Security, usability, maintainability, etc.). Since this information is not explicitly covered, it is the system designer who has to keep this knowledge in mind when designing an appropriate set of services that build the overall system. Unfortunately, stakeholders not involved in the design decisions, e.g. Customers, Project managers and developers, have a hard time to understand the significance of the coverage of customer requirements by services in order to meet desired quality aspects. This may cause misinterpretations especially in the early stages of the service system development life-cycle. In this paper, we present a heuristic-based approach to continually monitor and control design decisions and their effects on certain quality aspects. This approach extends the traceability matrix by a weighted decision-matrix.

Decision making, service-oriented architecture, software quality, customer requirements

http://dx.doi.org/10.1109/SOCA.2014.5

http://publik.tuwien.ac.at/files/PubDat_234757.pdf