Abstract
Requirements include the core functionalities of the software, however, due to their typical representations in natural language, professionals invest considerable effort in modeling to achieve more precise, coherent, and verifiable requirements analysis. This paper proposes a method called NL2PD for Problem Frames approach to extract requirements entities and relationships from requirements texts described in natural language and generate a problem diagrams to assist software requirements analysis. Given a requirements document in natural language, we can automatically extract required information and generate problem diagrams, on top of which users can edit, import and export. The method was tested with common requirements formats, demonstrating its ability to streamline the process of constructing problem diagrams from requirements. A demo video of this tool is available at https://youtu.be/SxoTikU1Mek.