Practice Selection Framework
Artikel i vetenskaplig tidskrift, 2012

Knowledge management (KM) in software engineering and software process improvement (SPI) are challenging. Most existing KM and SPI frameworks are too expensive to deploy or do not take an organization's specific needs or knowledge into consideration. There is thus a need for scalable improvement approaches that leverage knowledge already residing in the organizations. This paper presents the Practice Selection Framework (PSF), an Experience Factory approach, enabling lightweight experience capture and use by utilizing postmortem reviews. Experiences gathered concern performance and applicability of practices used in the organization, gained from concluded projects. Project managers use these as decision support for selecting practices to use in future projects, enabling explicit knowledge transfer across projects and the development organization as a whole. Process managers use the experiences to determine if there is potential for improvement of practices used in the organization. This framework was developed and subsequently validated in industry to get feedback on usability and usefulness from practitioners. The validation consisted of tailoring and testing the framework using real data from the organization and comparing it to current practices used in the organization to ensure that the approach meets industry needs. The results from the validation are encouraging and the participants' assessment of PSF and particularly the tailoring developed was positive.

reviews

requirements abstraction model

industry

project

technology-transfer

success

knowledge

knowledge management

postmortem

software process improvement

software process improvement

experience

Postmortem review

software engineering

management

Författare

Martin Ivarsson

Chalmers, Data- och informationsteknik, Software Engineering

Tony Gorschek

Blekinge Tekniska Högskola, BTH

International Journal of Software Engineering and Knowledge Engineering

0218-1940 (ISSN)

Vol. 22 17-58

Ämneskategorier

Data- och informationsvetenskap

DOI

10.1142/s0218194012500027