Machine learning (ML) applications have surged in popularity in the industry, however, the lack of transparency of ML-models often impedes the usability of ML in practice. Especially in the corporate performance management (CPM) domain, transparency is crucial to support corporate decision-making processes. To address this challenge, approaches of explainable artificial intelligence (XAI) provide solutions to reduce the opacity of ML-based systems. This design science study further builds on prior user experience (UX) and user interface (UI) focused XAI-research, to develop a user-centric approach to XAI for the CPM field. As key results, we identify design principles in three decomposition layers, including ten explainability UI-elements that we developed and evaluated through seven interviews. These results complement prior research by focusing it on the CPM domain and provide practitioners with concrete guidelines to foster ML adoption in the CPM field.


Paper Number 1349; Track AI; Complete Paper



When commenting on articles, please be friendly, welcoming, respectful and abide by the AIS eLibrary Discussion Thread Code of Conduct posted here.