Data Based Analysis of Requirements in Product Development Represented in Graph Based Semantic Requirement Nets
Michael Riesener, Viktor Slawik, T. Holtmann, S. Frölian, Maximilian Kühn, Günther Schuh · 2023
This paper systematically describes how the dependencies of natural language requirements can be implemented using a data based approach. A basis for further validation is conceptualized. In a short introduction and motivation, it is described that the need for a cross-system interconnection of individual requirements from different stakeholders can be met with data based methods. The methodology of automating the cross-linking as well as the representation of the interdependent requirements are described in the context of graph-based methods and generative large language models. Finally, an abstraction of the pipeline result is exemplified. This paper describes the basis for a methodology, which will be detailed in future work.