pith. sign in

arxiv: 1401.4254 · v1 · pith:MP7WFUIXnew · submitted 2014-01-17 · 💻 cs.SE

Goal-oriented Composition of Software Process Patterns

classification 💻 cs.SE
keywords processpatternscompositionprojectsoftwarecustom-tailoreddescriptionsdevelopment
0
0 comments X
read the original abstract

The development of high-quality software or software-intensive systems requires custom-tailored process models that fit the organizational and project goals as well as the development contexts. These models are a necessary prerequisite for creating project plans that are expected to fulfill business goals. Although project planners require individual process models custom-tailored to their constraints, software or system developing organizations also require generic processes (i.e., reference processes) that capture project-independent knowledge for similar development contexts. The latter is emphazised by assessment approaches (such as CMMI, SPICE) that require explicit process descriptions in order to reach a certain capability or maturity level. Among other concepts such as polymorphism, templates, or generator-based descriptions, software process patterns are used to describe generic process knowledge. Several approaches for describing the architecture of process patterns have already been published (e.g., [7]). However, there is a lack of descriptions on how to compose process patterns for a specific decelopment context in order to gain a custom-tailored process model for a project. This paper focuses on the composition of process patterns in a goal-oriented way. First, the paper describes which information a process pattern should contain so that it can be used for systematic composition. Second, a composition method is sketched. Afterwards, the results of a proof-of-concept evaluation of the method are described. Finally, the paper is summarized and open research questions are sketched.

This paper has not been read by Pith yet.

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.