Applicability and Usability of Predefined Natural Language Boilerplates in Documenting Requirements

Abstract

Natural language is frequently applied to document the stakeholders’ statements during requirement elicitation activities. Nevertheless, the use of generic natural language has potential for the issues of unclear and inconsistent requirements. These issues may result from the diverse interpretations by the stakeholders or other various sources of documents and artefacts. The main objective of this paper was to discuss the definition and application of predefined boilerplates to specify the requirements in the form of natural language statements. The proposed boilerplates were defined and classified based on two main types of requirements, namely functional and non-functional (performance, constraints, and specific quality). Two methods have been applied to evaluate the research results; the applicability of the predefined boilerplates was demonstrated using two different case studies, and the usability aspect is evaluated through synthetic environment experimentation using human respondents. As a summary, the predefined boilerplates were found helpful, especially among novice requirement engineers to express and specify their requirements in a consistent manner and a standardized way, relatively able to improve the quality of the natural language statements.

Fagan Inspection: A Defects Finding Mechanism in Software Requirements Specification (SRS) Document

“…Owing to this reason most of the software projects failed considerably. So, it's necessary to define and discuss the application of predefined samples on the basis of functional and non-functional to specify the requirements in the form of NL [3]. Writing a good SRS is an important determinant of software quality [2].…”