The key to software development is the transformation of requirements into test information. Our understanding of this transformation is limited. To help fill this gap, the authors conducted an empirical study of linkages between requirements and tests, for the purpose of characterizing, analyzing, and assessing methods for aligning them.

This well-organized paper starts with the importance of alignment, continues with a bottom-up approach for developing a taxonomy (requirements engineering and software test (REST)) for alignment methods, follows with threats to validity, and finishes with limitations and lessons learned. The authors adapt the results to define an alignment assessment process, and conclude with a concise summary and considerations for future research.

The central element of the REST taxonomy is the information dyad, represented by two nodes and a link between them. The node represents a primary information object used in software development. The link represents the alignment of the two information objects via transformations found in the literature, including semantic, syntactic, and logical connections. Methods are classified by information dyads together with context information. The collection of information dyads comprises the structure of the method, which is classified by properties of the graph.

A limitation of the study is the small scope of relevant publications. A way to broaden the study is suggested by its use of a graph structure, which is fundamentally a software architecture approach. The literature on software architecture consists of numerous publications on software structures that map to requirements via software patterns.

I highly recommended this paper for software developers, process engineers, and educators.