Size, Coupling and Cohesion Metrics for Product-Line Architecture Evaluation: Proposal and Experimental Validation
Andre F.R. Cordeiro and Edson OliveiraJr
DOI : 10.3844/jcssp.2018.408.422
Journal of Computer Science
Volume 14, Issue 3
Software Product Line (SPL) is an approach for de facto reusing of software artifacts for a given domain. Amongst the artifacts developed in the context of SPL is the Product-Line Architecture (PLA). It is one of the central artifacts of the SPL core assets responsible for abstracting a common architecture for specific products. Organizations should continuously evaluate the quality of their products by managing their PLA evolution and variabilities. Thus, the PLA evaluation should be taken into consideration as one of the most important activities throughout a SPL life cycle. Existing literature presents different PLA evaluation methods, from which metrics are most used. Thus, metrics allow different PLA quality attributes to be prioritized when deriving existing products, most of the time by trading-off them. Size, Coupling and Cohesion are examples of important attributes related to Maintainability from ISO/IEC 25010. Metrics for size, coupling and cohesion have successfully been applied in SPL, however, not taking into account commonalities and variabilities. Therefore, this paper proposes metrics for measuring Size, Coupling and Cohesion of PLA, as well as their experimental validations. Such validations are performed correlating the values collected for each metric to participants rating for each quality attribute. Results indicate a weak positive correlation for the size and coupling metrics and a weak negative correlation for the Cohesion metric. In view of the presented evidence, new studies and/or investigations must be carried out to make results generalizable.
© 2018 Andre F.R. Cordeiro and Edson OliveiraJr. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.