Lightweight Modeling to Improve Requirements Engineering of Enterprise Software

点击翻译按钮获取中文摘要

The Standish Group's 1994 CHAOS Report cites requirements-related problems as the main reason for the failure of over half of the 8,380 surveyed projects, with a majority of them in the enterprise software domain. According to Boehm, correcting requirements errors in later development phases can cost up to 200 times as much as correcting them in the requirements phase. Thus, finding requirements errors early is critical. The overarching goal of this research is to improve the effectiveness (fewer errors) and efficiency (higher speed) of requirements engineering for enterprise software (REES) via lightweight modeling. We will develop a combination of a method, language, and tools to create precise models and designed to deliver maximum value while focusing on usability. Precise models reduce errors due miscommunication among project stakeholders, avoiding rework and improving software quality. Wide adoption of precise modeling has the potential to significantly improve the effectiveness and efficiency of engineering complex software-intensive systems. The proposed research will have immediate benefit for our industrial collaborators, improving their capabilities to determine requirements of enterprise systems. In the longer term, lightweight modeling will contribute to the wide adoption of precise modeling across different domains and the entire engineering cycle, reducing software faults-which impact negatively the society today-and allowing the software industry to build even more complex systems to address tomorrow's needs.