Mitigating Challenges in the Elicitation and Analysis of Transparency Requirements

Software systems are getting more and more complex, with an increasing integration of machine-learning based decisions. The ubiquitous presence of these systems makes users more dependent on them and their correctness in many aspects of daily life. Thus, there is a rising need to make software systems and their decisions more comprehensible. This seems to call for more transparency in software-supported decisions. Therefore, transparency requirements have to be understood, elicited and translated to lower level requirements. However, there is a lack of understanding about the requirements engineering process for transparency and how the different roles, e.g., UX designers, data scientists, and other stakeholders, have to interact in this process. In order to fill this gap, the requirements engineering process for transparency requirements needs to be thoroughly investigated. For this purpose, I intend to conduct empirical studies with practitioners and other stakeholders to gain a deeper understanding of the process and its associated problems. Based on the results, additional research will be conducted to investigate and propose solutions with the purpose of supporting requirements engineering when transparency is required.