Semi-automatic generation of UML models from natural language requirements

Authors:

Abstract
Peer-reviewedGoing from requirements analysis to design phase is considered as\ud one of the most complex and difficult activities in software\ud development. Errors caused during this activity can be quite\ud expensive to fix in later phases of software development. One\ud main reason for such potential problems is due to the specification\ud of software requirements in Natural Language format. To\ud overcome some of these defects we have proposed a technique,\ud which aims to provide semi- automated assistance for developers\ud to generate UML models from normalized natural language\ud requirements using Natural Language Processing techniques. This\ud technique initially focuses on generating use-case diagram and\ud analysis class model (conceptual model) followed by\ud collaboration model generation for each use-case. Then it\ud generates a consolidated design class model from which code\ud model can also be generated. It also provides requirement\ud traceability both at design and code levels by using Key-Word-In-\ud Context and Concept Location techniques respectively to identify\ud inconsistencies in requirements. Finally, this technique generates\ud XML Metadata Interchange (XMI) files for visualizing generated\ud models in any UML modeling tool having XMI import feature.\ud This paper is an extension to our existing work by enhancing its\ud complete usage with the help of Qualification Verification System\ud as a case study
Funding Information
  • Science Foundation Ireland (03/CE2/I303_1)

This publication has 10 references indexed in Scilit: