%0 Journal Article %1 journals/ese/LindvallTC03 %A Lindvall, Mikael %A Tvedt, Roseanne Tesoriero %A Costa, Patricia %D 2003 %J Empirical Software Engineering %K architecture caseStudy implementation process software %N 1 %P 83-108 %T An Empirically-Based Process for Software Architecture Evaluation. %U http://dblp.uni-trier.de/db/journals/ese/ese8.html#LindvallTC03 %V 8

@article{journals/ese/LindvallTC03, added-at = {2009-01-18T15:37:35.000+0100}, author = {Lindvall, Mikael and Tvedt, Roseanne Tesoriero and Costa, Patricia}, biburl = {https://www.bibsonomy.org/bibtex/2f6b1a3a9a6cebe33020499b04bc9acd2/ericbouwers}, date = {2003-04-30}, description = {This paper describes a case study in which the authors have compared the implemented architecture of two versions of a software system with each other. First, the authors describe a process for their case study: 1) select a perspective for evaluation 2) define / select metrics and establish guidelines to be used in the evaluation 3) collect metrics 4) Evaluate/compare the architecture The perspective for this case study is defined as maintainability. The authors define a new set of metrics, based on the concepts of coupling between objects, called coupling between modules (CBM). To support analysis at a finer level the authors also take into account the coupling between module classes (CBMC). Also, a distinction is made between library and non-library modules. After the definition of the metrics an overview is given of the expected values of the metrics. After this, the modules in the system are extracted and the calls between modules are visualized. A comparison is made between both implemented architectures and between the new implemented architecture and its design. A few violations of the desired architecture are found and discussed. The conclusion is that this kind of evaluation can serve as a proof that a newly implemented architecture is an improvement over an old architecture. Furthermore, the analysis of an implemented architecture can show violations in the implementation that would not be found otherwise.}, interhash = {d50fee4ecff33520742ada1494234729}, intrahash = {f6b1a3a9a6cebe33020499b04bc9acd2}, journal = {Empirical Software Engineering}, keywords = {architecture caseStudy implementation process software}, number = 1, pages = {83-108}, timestamp = {2009-04-21T13:13:59.000+0200}, title = {An Empirically-Based Process for Software Architecture Evaluation.}, url = {http://dblp.uni-trier.de/db/journals/ese/ese8.html#LindvallTC03}, volume = 8, year = 2003 }