Conformance to templates Sample Clauses

Conformance to templates. Templates or boilerplates are frequently used to structure requirements. This allows the authors of the requirement documents to focus on eliciting the specific information requested by the template. Moreover, such structure makes automatic requirements analysis easier. Therefore, it is important, from a quality assurance perspective that the requirements conform to the templates. Arora (2013, 2015) proposes a template conformance checking technique based on text chunking — a NLP approach which identifies sentence segments (i.e., chunks). The templates under investigation, namely Xxxx and EARS, are transformed into xxxxxx-xxxx form grammars. This allows to define a series of pattern-matching rules for checking conformance based on the identified chunks. The tool XXXX (REquirement Template Analyzer) is developed to demonstrate the proposed approach. The evaluation consisted of four case studies (two using Xxxx and two using EARS) in the context of safety critical systems (e.g., satellite control, nuclear energy safety control) covering more than 1700 requirements. The results show that, from a practical standpoint, the tool is accurate for automatically checking requirements conformance to templates. Specifically, the number of false positives and false negatives is small across all cases in absolute numbers (38 in total) and as percentage of total number of non-conformant requirements (from 5% to 18%). Lucassen (2015) presents an NLP-based approach to assist stakeholders in producing high quality requirements by enforcing the conformance to user stories templates. The work first devises a conceptual model of high-quality user stories based on three high-level criteria — syntactic quality, semantic quality, and pragmatic quality— which are further broken down into 14 criteria (e.g., independent, uniform, unique, and complete). A tool, AQUSA, enforces that the guidelines for writing user stories, based on these criteria, are correctly followed. The tool supports syntactic criteria and, partially, pragmatic criteria, whereas semantic ones are not covered. The evaluation of the tool is done on three sets of user stories collected from industry. The tool shows 71% precision in detecting the user stories that violated quality criteria. The results show that one of the most common violations is minimality —i.e., the syntactic quality property for which a user story should not contain more than role (i.e., a relevant role should always be defined), means (i.e., de...