On 21 November at 13:00, Eric Müller will defend his Master's thesis on ‘Potentials and Limitations of Using Large language Models as a Proxy for Experts’ in room 242 (Konrad-Zuse-Haus). Eric is a Master's student on the Information Systems programme and was supervised by Benjamin Nast and Kurt Sandkuhl (both Chair of Information Systems).
Abstract
Enterprise modeling is crucial for the systematic representation and optimization of organizational processes and structures. Domain experts play a vital role in this context, providing essential knowledge and insights specific to the organization. This research investigates whether LLMs can alleviate the reliance on these experts for specific tasks within enterprise modeling. The main contributions of this study are threefold: (1) a comprehensive literature review on the application of LLMs in enterprise modeling, (2) the identification of potential applications and the effectiveness of LLMs in supporting various phases of enterprise modeling, and (3) empirical findings from quasi-experiments comparing the outputs of ChatGPT with those generated by domain experts across different enterprise modeling scenarios. The results highlight the situations where LLMs can be effectively integrated into enterprise modeling processes, while also addressing the inherent limitations and challenges in their use.