Scholarlink

<p>Requirements engineering focuses on eliciting, specifying, transforming and validating user and system requirements correctly and efficiently. Transforming user requirements to system requirements is a critical step in this process. It is a challenging because the high-level intentions of users often lack the detailed information to help specify system requirements. In practice, the successful transformation from user requirements to system requirements is labor-intensive, which requires the sophisticated human efforts of domain experts and developers for information processing and supplementation. It is desirable to have a method for automatically transforming user requirements into system requirements. In this paper, we propose an approach iStar2UML that can automatically transform the user requirements of iStar into the system requirements specified by UML models. iStar is a well-known goal-oriented model for eliciting and specifying user requirements; it concentrates on analyzing intentions and social dependencies of stakeholders. Unified Modeling Language (UML) is a de facto standard for object-oriented system requirements modeling and design. We evaluate the proposed approach through five case studies. The results indicate that 82.9% of UML models were successfully generated from iStar models and confirmed by domain experts. Additionally, the proposed transformation approach reduced transformation errors by an average of 11.7% and time costs by 21.4% compared to a fully manual approach. Overall, the results suggest the approach is effective, though further validation is necessary. The proposed approach can be extended and applied for the requirements engineering processes in the software industry.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

iStar2uml: toward automatic generation of UML model from iStar model

  • 期刊论文

  • 发表日期：

    2025年1月18日

查看全文
  • Yilong Yang,
  • Younggi Bok,
  • Hongyue Pan,
  • Yang Li,
  • Nan Niu,
  • Tong Li

摘要

Requirements engineering focuses on eliciting, specifying, transforming and validating user and system requirements correctly and efficiently. Transforming user requirements to system requirements is a critical step in this process. It is a challenging because the high-level intentions of users often lack the detailed information to help specify system requirements. In practice, the successful transformation from user requirements to system requirements is labor-intensive, which requires the sophisticated human efforts of domain experts and developers for information processing and supplementation. It is desirable to have a method for automatically transforming user requirements into system requirements. In this paper, we propose an approach iStar2UML that can automatically transform the user requirements of iStar into the system requirements specified by UML models. iStar is a well-known goal-oriented model for eliciting and specifying user requirements; it concentrates on analyzing intentions and social dependencies of stakeholders. Unified Modeling Language (UML) is a de facto standard for object-oriented system requirements modeling and design. We evaluate the proposed approach through five case studies. The results indicate that 82.9% of UML models were successfully generated from iStar models and confirmed by domain experts. Additionally, the proposed transformation approach reduced transformation errors by an average of 11.7% and time costs by 21.4% compared to a fully manual approach. Overall, the results suggest the approach is effective, though further validation is necessary. The proposed approach can be extended and applied for the requirements engineering processes in the software industry.