Using Problem Frames Approach for Key Information Extraction from Natural Language Requirements
Shangzhi Tang, Xuan Chen, Hongbin Xiao, Jiahao Wei, Zhi Li · 2023
The correct capturing and understanding of requirements play a crucial role in the software development process. However, the ambiguity in requirements described using natural language (NL) poses a challenge. Existing methods often require the use of specific requirements templates to rewrite system requirements, which increases efforts in analyzing requirements. Additionally, frequent requirement changes during the development process can lead to extra costs. In this paper, we propose a visualization method that allows user to extract key information from NL requirements, aiming to mitigate the difficulty and efforts in analyzing NL requirements. We also argue that more attention should be paid to the problems that need to be solved in the development process, especially the security problems. Application of our method to practical requirements sentences demonstrates its effectiveness in extracting the problem concerns from the requirements, leading to the stepwise derivation of solutions. We envision that our approach may have the potential to support iterative updates and analysis of NL requirements automatically.