Which Requirements Artifact Quality Defects are Automatically Detectable? A Case Study

Abstract

Abstract-[Context:]The quality of requirements engineering artifacts, e.g. requirements specifications, is acknowledged to be an important success factor for projects. Therefore, many companies spend significant amounts of money to control the quality of their RE artifacts. To reduce spending and improve the RE artifact quality, methods were proposed that combine manual quality control, i.e. reviews, with automated approaches.

A Live Extensible Ontology of Quality Factors for Textual Requirements

“…Previous attempts at establishing a subject-based classification for requirements quality are to the best of our knowledge limited to an approach by Saavedra et al [14], which is, however, on a coarser granularity and elicits only high-level requirements quality aspects like correctness, completeness, and others. The work most comparable to our approach has been conducted by Femmer et al [15], where 129 industrial requirements writing rules were classified regarding their eligibility for automation. Our own work differs from theirs in that (1) we aim at integrating quality factors established in peer-reviewed literature instead of in industrial writing rules [15] into a holistic ontology, while (2) considering the eligibility of the individual factors for automation only as one of many sub-goals.…”

“…The work most comparable to our approach has been conducted by Femmer et al [15], where 129 industrial requirements writing rules were classified regarding their eligibility for automation. Our own work differs from theirs in that (1) we aim at integrating quality factors established in peer-reviewed literature instead of in industrial writing rules [15] into a holistic ontology, while (2) considering the eligibility of the individual factors for automation only as one of many sub-goals. Further (3), as our endeavour shall lay the groundwork for a long-term community initiative, one main contribution is to publicly disclose all of our results for an effective maintenance and evolution of the ontology by the community.…”

“…We chose the former approach for the initial iteration, as a significant understanding of the domain has already been established along previous engagement with requirements quality research. We distilled an initial structure of the ontology based on relevant literature [14], [15], [19].…”

“…Requirements quality factors are characterized by a name and scope, which is the dimension representing the granularity of input necessary in order to decide the quality factor. As an example, one publication by Femmer et al [15] contains multiple quality factors, one of which is named containing subflows. The scope of this factor is use case, as a full use case is necessary to decide whether at least one subflow is contained or not.…”

Crossing Disciplinary Borders to Improve Requirements Communication

“…Requirements artifacts are predominantly written in natural language, both in unstructured common language and in a structured form using templates and forms [13,14]. Other documentation types include conceptual models (like UML state diagrams or UML activity diagrams) or hybrid documentation combining natural language and conceptual models [5].…”