Enhancement of Model Driven Software Development using AI
Bhakti Sanket Puranik, Arti Sonawane, Jithina Jose, Shubham Chavan, Yash Patil · 2024
The future of the MDSD in this emerging scenario of software engineering is that models are completely converted into Code i.e. program which are completely developed from high-level models, each one representing a various aspects of the parts of system. This work investigates the use of cutting-edge artificial intelligence (AI) methods to produce UML diagrams, which are essential for software engineering's system architecture visualization. Our methodology, which is based on machine learning algorithms such as natural language processing (NLP) and deep learning, is a reliable approach to automate the generation of UML diagrams from textual software requirements. Our methodology entails preprocessing textual input, using graph-based techniques to create cohesive UML diagrams, and using NLP models to extract important entities and relationships .The system's ability to handle several UML diagram types—such as use case, sequence, and class diagrams— ensures thorough coverage of software design requirements. Our AI-driven approach dramatically lowers manual labor, improves accuracy, and expedites the modeling process, as shown by many testing. The findings point to a viable path for incorporating AI into software development, which might revolutionize current procedures by giving engineers and developers access to an effective tool.