NFRNet: A Deep Neural Network for Automatic Classification of Non-Functional Requirements
Bing Li, Zhi Li, Yilong Yang
Abstract
Non-functional requirements specify those qualities that software products must have in order to meet the user’s business requirements. The elicitation of these non-functional requirements requires expertise, experience, and domain knowledge, which is challenging and time-consuming for requirements engineers and developers. It would be very beneficial if the nonfunctional requirements can be automatically extracted from the requirements documentation to reduce the human efforts, time, and avoid the mental fatigue. In this paper, we present a novel deep neural network model called NFRNet to automatically extract non-functional requirements from software requirements documentation.
Topics & Concepts
Non-functional requirementFunctional requirementDocumentationRequirements elicitationComputer scienceRequirements analysisSoftware requirements specificationNon-functional testingBusiness requirementsSoftware engineeringDomain (mathematical analysis)User requirements documentRequirements managementFunctional specificationSoftware requirementsArtificial neural networkSoftwareRequirements engineeringArtificial intelligenceSoftware systemSoftware developmentSoftware designEngineeringBusiness processSoftware constructionProgramming languageMathematical analysisChemical engineeringCompatibility (geochemistry)MathematicsSoftware Engineering ResearchSoftware Engineering Techniques and PracticesAdvanced Software Engineering Methodologies