Generative AI is rapidly employed by software developers to generate code or other software artifacts.However,the analysis and assessment of generative AI with respect to requirements analysis and modeling tasks,espec...Generative AI is rapidly employed by software developers to generate code or other software artifacts.However,the analysis and assessment of generative AI with respect to requirements analysis and modeling tasks,especially with UML,has received little attention.This paper investigates the capabilities of generative AI to aid in the creation of three types of UML models:UML use case models,class diagrams,and sequence diagrams.For this purpose,we designed an AI-aided UML modeling task in our course on software requirements modeling.50 undergraduates who majored in Software Engineering at Wuhan University completed the modeling task and the corresponding online survey.Our findings show that generative AI can help create these three types of UML models,but its performance is limited to identifying essential modeling elements of these UML models.展开更多
System analysis and design (SAD) is a crucial process in the development of software systems. The impact of modeling techniques and software engineering practices on SAD has been the focus of research for many years. ...System analysis and design (SAD) is a crucial process in the development of software systems. The impact of modeling techniques and software engineering practices on SAD has been the focus of research for many years. Two such techniques that have had a significant impact on SAD are Unified Modeling Language (UML) and machine learning. UML has been used to model the structure and behavior of software systems, while machine learning has been used to automatically learn patterns in data and make predictions. The purpose of this paper is to review the literature on the impact of UML and machine learning on SAD. We summarize the findings from several studies and highlight the key insights related to the benefits and limitations of these techniques for SAD. Our review shows that both UML and machine learning have had a positive impact on SAD, with UML improving communication and documentation, and machine learning improving the accuracy of predictions. However, there are also challenges associated with their use, such as the need for expertise and the difficulty of interpreting machine learning models. Our findings suggest that a combination of UML and machine learning can enhance SAD by leveraging the strengths of each technique.展开更多
文摘Generative AI is rapidly employed by software developers to generate code or other software artifacts.However,the analysis and assessment of generative AI with respect to requirements analysis and modeling tasks,especially with UML,has received little attention.This paper investigates the capabilities of generative AI to aid in the creation of three types of UML models:UML use case models,class diagrams,and sequence diagrams.For this purpose,we designed an AI-aided UML modeling task in our course on software requirements modeling.50 undergraduates who majored in Software Engineering at Wuhan University completed the modeling task and the corresponding online survey.Our findings show that generative AI can help create these three types of UML models,but its performance is limited to identifying essential modeling elements of these UML models.
文摘System analysis and design (SAD) is a crucial process in the development of software systems. The impact of modeling techniques and software engineering practices on SAD has been the focus of research for many years. Two such techniques that have had a significant impact on SAD are Unified Modeling Language (UML) and machine learning. UML has been used to model the structure and behavior of software systems, while machine learning has been used to automatically learn patterns in data and make predictions. The purpose of this paper is to review the literature on the impact of UML and machine learning on SAD. We summarize the findings from several studies and highlight the key insights related to the benefits and limitations of these techniques for SAD. Our review shows that both UML and machine learning have had a positive impact on SAD, with UML improving communication and documentation, and machine learning improving the accuracy of predictions. However, there are also challenges associated with their use, such as the need for expertise and the difficulty of interpreting machine learning models. Our findings suggest that a combination of UML and machine learning can enhance SAD by leveraging the strengths of each technique.
