Answers are grounded in this paper and cite their sources.
Automatic generation improves requirements validation efficiency in software prototypes, indicating better accuracy and cost-effectiveness.
In the early stages of computer software information system development, requirements errors can lead to software system failure and performance degradation and even cause huge security incidents. Traditional requirements verification methods are inefficient and susceptible to human factors when dealing with complex software requirements. Rapid prototyping is an effective requirement validation method, but the generated prototype does not contain any data, and the traditional method requires domain experts to write the data manually, which is time consuming and complicated. In this study, an automatic software prototype data generation method, InitialGPT, is proposed, which automatically generates requirements-compliant prototype data by interacting with users through a requirements model to improve the efficiency and accuracy of requirements validation. We designed a framework containing a prompt generation template, a data generation model, a data evaluation model, and multiple prototype data tools, and validated it on four real-world software system cases. The results show that the approach improves the efficiency of requirements validation by a factor of 7.02, and generates data of similar quality to those written manually, but at a more advantageous cost and efficiency, demonstrating its potential for application in the computer software industry.
No takes yet. Share an insight, caveat, or question.
Chang et al. (2025) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: