In diesem Semester bieten wir wieder spannende Studienprojekte in vier Themenbereichen an. Im Bereich Plattformökosysteme könnt Ihr digitale Plattformstrukturen in 3D visualisieren, Governance-Entscheidungen interaktiv explorieren oder eine empirisch fundierte Governance-Taxonomie für GenAI-Plattformen aufbauen. Im Schwerpunkt Agentic IS erforscht Ihr, wie KI-Agenten als Teammitglieder in der agilen Softwareentwicklung mitwirken, wie ein agentenbasierter Mietrechtsassistent rechtliche Beratung zugänglicher macht oder wie Multi-Agenten-Systeme komplexe Aufgaben koordiniert bewältigen. Im Forschungsfeld Embodied Generative AI arbeitet Ihr mit einem Roboterhund an Such- und Rettungsszenarien oder an einem Assistenzsystem für sehbehinderte Menschen – oder entwickelt ein Interaktionskonzept für einen humanoiden Roboter im Haushaltskontext. Im Bereich Generative KI erwarten Euch Projekte zu virtuellen Patienten für die medizinische Ausbildung, GenAI-basiertem Funktraining im Katastrophenschutz, simulierten Kommunikationspartnern in der Flugverkehrskontrolle, KI-gestützter Übersetzung behördlicher Dokumente in leichte Sprache sowie die Weiterentwicklung unseres digitalen Lernassistenten ProfessorGPT.

Wir freuen uns auf Eure Bewerbungen und die Zusammenarbeit mit Euch in unseren Studienprojekten!

Hier geht es zu unseren Projekten!

Euer SOFTEC-Team

In der renommierten Zeitschrift Electronic Markets (Jg. 36, 2026) veröffentlichen Florian Holldack, Leonardo Banh und Gero Strobel einen konzeptionellen Grundlagenartikel zum Thema Agentic Information Systems. Der Beitrag beschreibt einen Paradigmenwechsel in der Informationssystemforschung: Während klassische IS primär reaktiv agierten, übernehmen agentische IS zunehmend autonome Entscheidungsfindung, handeln in unstrukturierten Umgebungen und delegieren sogar Aufgaben an Menschen. Die Autoren systematisieren zentrale Merkmale agentischer IS, identifizieren archetypische Ausprägungen und diskutieren neue Formen der Mensch-Maschine-Interaktion sowie deren sozio-technische Chancen und Risiken. Dies ist insbesondere im Zeitalter (generativer) künstlicher Intelligenz und KI-Agenten von Relevanz für Forschung und Praxis.

Auch auf internationaler Konferenzebene war der Lehrstuhl präsent: Auf der 46th International Conference on Information Systems (ICIS 2025) in Nashville, Tennessee, USA stellten Florian Holldack, Leonardo Banh und Gero Strobel ihr Paper “Synthetic Conversations, Real Insights: Towards Realistic User Simulations with Generative Agents” vor. Die Studie zeigt, wie generative Agenten auf Basis großer Sprachmodelle realistische Nutzersimulationen für die Evaluation dialogbasierter Empfehlungssysteme ermöglichen. Die Ergebnisse verdeutlichen, dass solche Agenten menschliche Bewertungen auf aggregierter Ebene gut approximieren können und damit eine skalierbare Alternative zu klassischen, aufwendigen Nutzerstudien darstellen.

Zuletzt folgte ein weiterer Forschungserfolg zum Jahresbeginn auf der 59th Hawaii International Conference on System Sciences (HICSS 2026). Das Paper “Hiring Tomorrow’s Talents: How Generative Artificial Intelligence Transforms Human Resources Recruitment” von Leonardo Banh, Alexander Rex, Gero Strobel und Nils Urbach untersucht den Einsatz generativer KI in der Personalrekrutierung. Basierend auf qualitativen Interviews entwickeln die Autoren ein theoretisches Modell, das aufzeigt, wie generative KI Recruiter im Arbeitsalltag unterstützt, Prozesse transformiert und zugleich neue organisatorische und ethische Herausforderungen mit sich bringt. Der Beitrag wurde am 6. Januar 2026 von unseren Forschungspartnern und Co-Autoren der Frankfurt UAS/Fraunhofer FIT in Hawaii präsentiert.

Wir freuen uns über die Erfolge unserer Mitarbeiter, welche die internationale Sichtbarkeit des SOFTEC-Lehrstuhls stärkt und bedanken uns herzlich bei allen Forschungspartnern, Editoren und Begutachtern und Konferenzveranstaltern.

The International Conference on Information Systems (ICIS 2025) is the world’s leading conference in the field of Information Systems and will take place this year in Nashville, Tennessee, USA. The paper was published in the IS Design, Development, and Project Management track.

The study addresses a central challenge in research on Conversational Recommender Systems (CRS): traditional user studies are costly and time-consuming, while existing simulations fail to adequately capture realistic user interactions. The paper demonstrates how generative agents based on Large Language Models can be used to conduct user simulations. In an experiment, 200 generative agents and 50 human participants interacted with a prototypical CRS and were subsequently evaluated using standardized questionnaires.

The results show that, on a macro level, generative agents behave similarly to human users and provide subjective evaluations that can serve as an alternative to labor-intensive studies. Thus, the work makes an important contribution to advancing research methods in the field of Information Systems—particularly for user-centered design and evaluation of socio-technical systems in the age of generative AI.

You can find the paper here.

Abstract

The rise of generative AI and large language models (LLMs) has sparked interest in applying generative agents to simulate users in conversational recommender system (CRS) evaluations. While CRS rely on user perception for success, traditional user studies are costly and time-intensive. Existing simulation approaches lack interaction depth and user-centric evaluation. This study addresses these gaps by leveraging LLM-based generative agents to conduct scalable, subjective CRS assessments. We present a comparative study where 200 generative agents and 50 human participants interact with a prototypical agentic CRS and evaluate their experience using a structured questionnaire. Results indicate that generative agents approximate human-like behavior and subjective assessments at the macro-level, despite granular precision limitations, offering an alternative to traditional user studies. Our findings advance research on CRS evaluation by demonstrating how agentic simulations can support human-aligned assessments of socio-technical systems and open new avenues for applying Generative AI in user-centered decision support.

We warmly congratulate the authors on this successful publication and are delighted about the strong presence of our chair at ICIS 2025!

The Chair of Information Systems and Software Engineering (SOFTEC) also took part in the event. Jan Laufer, Leonardo Banh, and Gero Strobel presented their current research titled “Towards True Feelings: A Design Theory for Compassionate AI.”

While today’s AI systems are often optimized for efficiency and objectivity, the concept of Compassionate AI takes a different approach: it aims to recognize suffering and actively contribute to its alleviation. The proposed design theory argues that moving from efficiency-driven AI to truly compassionate AI is not merely a technical advancement, but rather a fundamental shift in system design thinking.

The research focuses on the field of assistive technologies for blind and visually impaired people. This group experiences various forms of suffering, from perceptual and navigational challenges to loss of independence and social or infrastructural barriers. Moreover, individual needs vary widely depending on residual vision, the timing of vision loss, and environmental factors such as accessibility in the local context.

Generative AI opens new opportunities to enhance existing assistive technologies such as apps (e.g., Seeing AI) or wearables (e.g., Ray-Ban Meta Glasses). However, such systems are often designed for mass-market applications without fundamentally addressing the specific needs and forms of suffering experienced by blind and visually impaired individuals. Compassionate AI may therefore represent the necessary paradigm shift toward truly human-centered AI systems.

As part of the project, 50 interviews with blind and visually impaired participants have already been conducted and are currently being analyzed. The insights gained will inform the development of prototypes for future applications—potentially including our robot dog as a robotic guide dog in the future.

We look forward to seeing where this research will lead and would like to thank the organizers of the Day of Computer Science 2025 for an inspiring and successful event!

In diesem Semester bieten wir wieder spannende Studienprojekte im Bereich der Künstlichen Intelligenz (KI), insbesondere generative AI und Large Language Models (LLMs) an. Unser vielfältiges Projektangebot umfasst Themen wie die Entwicklung von LLM-basierten digitale Assistenten als Lernunterstützung, zur Modellierung von Plattformökosystemen oder als Mietrechtsexperte. Ein besonderer Fokus liegt dieses Jahr auf dem Bereich der generativen Agenten und wie mit Hilfe dieser Echtweltszenarien wie Krankenhäuser, agilen Teams in der Softwareentwicklung oder virtuelle Patienten für die medizinischen Ausbildung abgebildet werden können. Darüber hinaus bieten wir zwei Projekte zum Thema Embodied Generative AI im spannenden Forschungsfeld zwischen KI und Robotik.

Wir freuen uns auf Eure Bewerbung und die Zusammenarbeit mit Euch in unseren Studienprojekten.

Euer SOFTEC-Team

First and foremost, we warmly congratulate our former master’s student Ben Möllmann on a very special recognition. His paper  „Trust Me, I'm a Tax Advisor: Influencing Factors for Adopting Generative AI Assistants in Tax Law“, developed together with Leonardo Banh, Jan Laufer and Gero Strobel received the Best Student Paper Award at WI 2025! The award was presented during the festive conference dinner under the motto “The Great GatsWI”.

Generative AI is increasingly being used in the field of tax law, where frequent regulatory changes demand accurate and up-to-date information. The paper explores how trust acts as a critical factor in successful human-AI collaboration. Using a mixed-methods approach, the study shows that factors such as transparency, anthropomorphism, and adherence to social norms enhance users’ trust. These findings highlight the importance of trust-building measures when designing GenAI applications for highly regulated domains.

Two additional papers from our chair were also accepted and presented at the conference:
Leo presented his conference paper “Developing a Generative AI Maturity Model for Supporting the Organizational Adoption Journey”.

Generative AI can only reach its full potential when organizations possess the necessary capabilities and structures to implement and scale such systems effectively. Leo’s paper addresses this challenge by introducing the Generative AI Maturity Model (GAI-MM) – a comprehensive framework that helps organizations assess and develop their GenAI capabilities across key dimensions such as people, processes, technology, data, and organization.

Jan presented his contribution “Bridging Mind and Matter: A Taxonomy of Embodied Generative AI”.

While generative AI has largely been considered a purely digital technology, Jan’s work shifts the focus toward Embodied Generative AI – systems that can physically act, perceive, or operate in real-world environments. His paper introduces a well-founded taxonomy of embodied GenAI systems, organizing core characteristics such as sensors, actuators, autonomy, interaction types, and application areas. The taxonomy provides a conceptual basis for better understanding and advancing embodied GenAI – for example, in robotics.

In addition, Gero Strobel served as Track Chair and Session Chair for the track “Generative AI: Shaping Human Collaboration, Organizations, and Societal Impact”, moderating two well-attended sessions on this timely and highly relevant topic.

We are proud of this strong representation at one of the leading conferences in the field of Information Systems!

Beyond the academic highlights, WI 2025 was also a place for exchange, inspiration, and new perspectives.

On the one hand, Leo had the opportunity to present his dissertation project at the Doctoral Consortium and received valuable feedback from experienced mentors. On the other hand, we were especially pleased that our chairholder, Prof. Dr. Stefan Eicker, joined us in Münster – returning to his alma mater.

With many new ideas, engaging discussions, and a healthy dose of pride, we look back on a thoroughly successful WI 2025 and thank the organizers and our (academic) partners for the fruitful collaboration and exchange!

Our booth attracted a diverse audience: students, research assistants from other departments, and many high school students who were looking for the right degree program. WDR Lokalzeit also paid us a visit: we were able to provide a practical insight into the content and opportunities offered by information system research at the University of Duisburg-Essen. The WDR Lokalzeit Duisburg television report can be found online (from minute 17:30).

The focus was on our robot dog Ben, who impressively demonstrated how we conduct research at the SOFTEC Chair at the interface of technology, humans, and society. Using the robot, we presented our current project on embodied generative AI (embodied GenAI) – i.e., the integration of generative AI into physical systems such as robots, drones, or vehicles. The goal is to develop intelligent, autonomous, and interactive agents that are not only powerful but also empathetic, context-sensitive, and interact with humans in an everyday manner. The aim is to contribute to improving the quality of life of individuals and communities.

We are particularly pleased about the great cooperation within the Faculty of Computer Science. Visitors to our joint booth were able to win small prizes on the wheel of fortune and learn about the content of our courses and everyday student life in information systems research from our computer science student council. Information materials on studying, career paths, and research rounded out the offerings.

We would like to thank everyone who stopped by, asked questions, and participated. We look forward to seeing you again next time!

Further information about the UDE Summer Festival can be found at:
https://www.uni-due.de/de/ude_sommerfest/

In their two contributions, Leonardo Banh, Fabian Stangl, Gero Strobel and René Riedl explore the use of Generative Artificial Intelligence (GenAI) in the context of NeuroIS research:

Exploring the NeuroIS Potential for Generative Artificial Intelligence: Findings from a Literature Review.

Generative Artificial Intelligence (GenAI) is transforming human-computer interaction, shaping behavior as well as cognitive and emotional processes. This paper explores how Neuro-Information Systems (NeuroIS) measurements can be applied to the study of GenAI, addressing their role in human-AI interaction. Through a literature review, we identify 21 papers using neurophysiological measurements, including autonomic nervous system (ANS) markers (e.g., eye-tracking), brain imaging (e.g., EEG), and multimodal approaches such as combining eye tracking and EEG. Our findings highlight main research themes, including cognitive offloading, trust, and decision-making biases in human interaction with GenAI. While research on this topic is becoming more prominent, neurophysiological investigations remain limited. We anticipate that measures of brain and ANS system activity, as well as hormone measures, will play an increasing role in future empirical research on GenAI. This study contributes to the advancement of NeuroIS by providing a structured foundation for better understanding the role of GenAI in this research field.

The Role of Generative Artificial Intelligence in the NeuroIS Research Process: Applications and Opportunities

Research in Neuro-Information Systems (NeuroIS) integrates neuroscience methods with Information Systems (IS) research to advance our understanding of human-technology interactions. As neurophysiological data collection methods evolve, the analysis of large and complex data sets remains a significant challenge. Generative Artificial Intelligence (GenAI) offers new opportunities for NeuroIS by improving data analysis, experimental design, and interpretation of neural patterns. This paper systematically reviews 56 studies applying GenAI in NeuroIS and identifies five main research themes: (1) GenAI for Autonomic Nervous System Measurements, (2) GenAI for Brain Research, (3) GenAI for General Applications, (4) GenAI for Genetics, and (5) GenAI for Multimodal Approaches. Our findings highlight how GenAI improves data interpretation, integration, and processing while streamlining the use of research methods. Overall, this review underscores the transformative potential of GenAI in NeuroIS, paving the way for scalable, precise, and dynamic research methodologies in the field.

The publications will be presented during the conferences in front of the scientific community.

In this publication, the authors explore the transformative potential of generative artificial intelligence (GenAI) tools—such as ChatGPT and GitHub Copilot—in the context of Software Engineering. Based on a qualitative study involving 18 industry experts, the paper analyzes both the opportunities and challenges associated with integrating GenAI into existing development processes.

The article is available at the following link: https://doi.org/10.1016/j.infsof.2025.107751 

Leonardo Banh führte gemeinsam mit Florian Holldack und Jan Laufer durch die Veranstaltung und gab den Teilnehmerinnen einen ersten Einblick in das Studienfach Wirtschaftsinformatik an der Universität Duisburg-Essen, sowie eine Einführung in Generative KI. Anschließend konnten die Schülerinnen ihre kreative Seite entfalten und spielerisch in einem “Prompt-A-Thon” die Chancen und Herausforderungen der KI-Technologie eigenständig erleben. Dabei präsentierten sie engagiert eindrucksvoll ihre kreativen Schöpfungen, von verzauberten Kurzgeschichten bis zu spannenden Produktideen. Ein abschließender Diskurs über die Learnings, Best-Practices aber auch Limitationen aktueller Generativer KI-Tools verdeutlichte die Relevanz des kritischen Auseinandersetzens mit Technologie für die Schülerinnen, insbesondere den bedachten Einsatz von KI und die Kontrolle der Ergebnisse. 

Zum Schluss gab es als überraschenden “Gastauftritt” eine kurze Demonstration unseres Roboterhundes Ben, welcher nicht nur die neugierigen Blicke der Schülerinnen auf dem Campus auf sich zog, und als Teil unserer Forschungsvorhaben die spannenden Facetten soziotechnischer Studien in der Zukunft der KI-Technologien und Robotik skizzierte.

Wir bedanken uns herzlich bei Yasmin Taounza vom IIS-Lehrstuhl, welche uns als aktive Studentin beim Workshop mit ihren persönlichen Erfahrungen im WI-Studium unterstütze, sowie dem Team des zentralen Gleichstellungsbüros der UDE, das den Tag organisiert hat!

Mehr Informationen zum UDE Girls'Day finden Sie unter: https://www.uni-due.de/schuelerinnenprogramme/girlsday/