A Problem-Oriented Research And Visualization Of Requirements Decomposition And Complexity Evaluation

With the rapid development of information technology today,information technology has become an important cornerstone of the development of modern society.As an integral part of software system development,requirements analysis is directly related to the success or failure of information technology.Requirements analysis plays a crucial role in the process of software system development.It not only determines the functionality,performance and quality of the software system,but also has a direct impact on the satisfaction and efficiency of the system users.Without adequate requirements analysis at the early stage of software system development,it is easy to cause deviations in development direction and unnecessary duplication of effort,resulting in development schedule delays and cost increases.Problem Frames(PF),as a mainstream software requirements analysis method,has gained the attention and recognition of the requirements engineering community during its continuous research and development in recent decades.The approach defines software requirements as a collection of behavioral descriptions of software systems interacting with real-world problem domains,and treats software requirements as a problem around which the software system corresponding to the requirements is solved.The problem framework approach describes the software requirements problem by constructing a problem graph,reasoning about and solving the software requirements problem through problem asymptotic and chains of causal phenomena.Although much research has been done on problem framing,the problem of how to effectively split requirements and evaluate the complexity of requirements is still a problem that has yet to be solved.Therefore,the research on problem-oriented requirements splitting and complexity evaluation is of great importance.The research in this dissertation is as follows:This dissertation proposes a problem-oriented requirement decomposition method,which is a problem graph-based requirement tracking method aimed at better understanding requirements and specifying their boundaries.The method incorporates shared phenomena from the problem framework approach based on semi-structured requirement descriptions,and extracts causal phenomena from requirements,combining the three methods of proposed requirement tracing to traverse the causal relationships of requirements,and to find associated requirements.This dissertation proposes a problem-oriented requirement complexity evaluation method,aiming to provide a holistic complexity evaluation method for relevant stakeholders in the early stages of software development to enable scientific allocation of manpower and resources for subsequent software development work.The method is mainly based on requirement decomposition and clarifying the inclusion relationship of requirements.By mapping the problem graph to a use case diagram and combining the global view of requirements in the use case diagram,the initiator effect and trigger effect are proposed on the use case diagram,while the complexity of the requirements is evaluated by integrating the relationships between use cases and the actual number of scenarios.This dissertation develops Trace4 PF,a requirement modelling visualization tool.Trace4 PF aims to visualize the above research methodology and provide a visualization solution for requirements modelling.Trace4 PF allows for requirements modelling,requirements decomposition,use case diagram mapping and visualization of requirements complexity assessment without the need for a complex environment and with a good user experience compared to other related modelling tools.In the final chapter,we also perform requirements modelling,requirements decomposition,use case diagram mapping and requirements complexity assessment based on a case blood glucose detection system to demonstrate the usefulness of our approach and visualization platform in a real case.