IntroductionAdvances in requirements engineering, driven by various paradigms and methodologies, have significantly influenced software development practices. The integration of agile methodologies and model-driven development (MDE) has become increasingly critical in modern software engineering. MDE emphasizes the use of models throughout the development process, necessitating structured approaches for handling requirements written in natural language.MethodsThis paper proposes an automated requirements engineering framework for agile model-driven development to enhance the formalization and analysis of textual requirements. The framework employs machine learning models to extract essential components from requirements specifications, focusing specifically on class diagrams. A comprehensive dataset of requirements specification problems was developed to train and validate the framework's effectiveness.ResultsThe framework was evaluated using comparative evaluation and two real-world experimental studies in the medical and information systems domains. The results demonstrated its applicability in diverse and complex software development environments, highlighting its ability to enhance requirements formalization.DiscussionThe findings contribute to the advancement of automated requirements engineering and agile model-driven development, reinforcing the role of machine learning in improving software requirements analysis. The framework's success underscores its potential for widespread adoption in software development practices.