Generating UML Class Diagram from Natural Language Requirements: A Survey of Approaches and Techniques

Abstract

In the last years, many methods and tools for generating Unified Modeling Language (UML) class diagrams from natural language (NL) software requirements. These methods and tools deal with the transformation of NL textual requirements to UML diagrams. The transformation process involves analyzing NL requirements and extracting relevant information from the text to generate UML class models. This paper aims to survey the existing works of transforming textual requirements into UML class models to indicate their strengths and limitations. The paper provides a comprehensive explanation and evaluation of the existing approaches and tools. The automation degree, efficiency, and completeness, as well as the used techniques, are studied and analyzed. The study demonstrated the necessity of automating the process, in addition to combining artificial intelligence with engineering requirements and using Natural Language Processing (NLP) techniques to extract class diagrams from NL requirements.