Four Contributions from the University Presented at ECIS 2025
Four Contributions from the University Presented at ECIS 2025
At this year’s European Conference on Information Systems (ECIS) – the most prominent European academic conference in the field of Information Systems – the University of Liechtenstein was represented with four scientific contributions. Two of these were developed within Erasmus+-funded projects.
The presented papers address current challenges in the areas of artificial intelligence (AI), micro-credentials, and design science research.
Making Machine Learning Models More Understandable Through LLM-Based Chatbots
The paper “Enhancing ML Model Interpretability: Leveraging Fine-Tuned Large Language Models for Better Understanding of AI” introduces a novel explainable AI (XAI) reference architecture. By integrating a fine-tuned chatbot based on large language models (LLMs), it improves the accessibility of machine learning (ML) predictions for end users. The architecture was validated in the context of battery State of Health (SoH) prediction and demonstrated improved interpretability—especially for less experienced users.
Bokstaller, Jonas; Altheimer, Julia; Dormehl, Julian Armin; Buss, Alina; Wiltfang, Jasper I.; Schneider, Johannes; and Röglinger, Maximilian, "Enhancing ML Model Interpretability: Leveraging Fine-Tuned Large Language Models for Better Understanding of AI" (2025). ECIS 2025 Proceedings. 3. https://aisel.aisnet.org/ecis2025/human_ai/human_ai/3
AI-Powered Access to Design Knowledge
The paper “Facilitating Design Knowledge Access with a Generative AI-Based Conversational Assistant” addresses challenges in systematically retrieving knowledge in design science research (DSR). A generative AI-based chatbot was developed to assist researchers in identifying and reusing existing design knowledge. The contribution is part of the Erasmus+ project Design Science Research Academy and is based on an initial design cycle, outlining specific requirements, design principles, and a prototype implementation.
Gau, Michael and Gnewuch, Ulrich, "Facilitating Design Knowledge Access With a Generative AI-Based Conversational Assistant" (2025). ECIS 2025 Proceedings. 2. https://aisel.aisnet.org/ecis2025/des_research/des_research/2
Micro-Credentials for Standardized Skills Recognition
In the Erasmus+ project “Broadening the Recognition Ecosystem in VET with Micro Credentials”, a European-wide concept was developed for recognising vocational competencies. The paper “Design of a Micro-Credentialing System for Skills Recognition in Education and Training” presents an innovative system for awarding micro-credentials, enabling standardized terminology and categorization of skills across national borders. The results include the first design cycle, with defined principles, a prototype, and its evaluation.
Kipping, Gregor; Gau, Michael; Terzis, Joseph; and Zourou, Katerina, "Design of a Micro-Credentialing System for Skills Recognition in Education and Training" (2025). ECIS 2025 Proceedings. 6. https://aisel.aisnet.org/ecis2025/education/education/6
Strategic Management of AI Initiatives
The paper “How Organizations Design Portfolio Management to Govern AI: A Taxonomy Approach” examines how organizations strategically govern AI initiatives. Based on five case studies, the authors developed a multi-layered taxonomy of key design decisions in AI portfolio management. The taxonomy was validated through expert interviews and practical interventions. The study offers valuable insights for both academia and practice on the governance of AI projects.
Sturm, Simon and van Giffen, Benjamin, "How Organizations Design Portfolio Management to Govern AI: A Taxonomy Approach" (2025). ECIS 2025 Proceedings. 9. https://aisel.aisnet.org/ecis2025/ai_org/ai_org/9
Academic Visibility on the European Stage
The acceptance and presentation of four contributions at ECIS 2025 highlight the relevance and quality of the University’s research in the field of Information Systems. The University of Liechtenstein warmly congratulates all participating researchers on this achievement.
