Abstract
Requirement's analysis and design is a multifaceted and time-consuming process. The success of software projects critically relies on careful & timely analysis and modeling of system requirements. Mostly, the requirements gathered from the stakeholders are written in some language (probably English). In this regard, significant manual efforts are required for the formation of good class model which unfortunately results in time delays in the software industry. The problems associated with the requirement analysis and class modeling can be overcome by the appropriate employment of machine learning. In this paper, we propose a system, requirement engineering analysis & design (READ) to generate unified modeling language (UML) class diagram using natural language processing (NLP) and domain ontology techniques. We have implemented the READ system in Python and it successfully generates the UML class diagram i.e., class name, attributes methods, and relationships from the textual requirements written in English. To assess the performance of the proposed system, we have evaluated it on publicly available standards and the experimental results show that it outperforms the existing techniques for object-oriented based software designing.