Best Paper Nominierungen

Power Imbalances in Society and AI: On the Need to Expand the Feminist Approach

Eva Johanna Gengler, Marco Wedel, Alexandra Wudel, Sven Laumer

Recent advances in Artificial Intelligence have brought novel opportunities for businesses, societies, and individuals alike, yet they also raise complex questions on inequitable power distribution. We see contemporary AI systems, that reinforce power imbalances and disadvantage marginalized, underrepresented, and underprivileged people. Current approaches to advancing AI, such as Ethical, Fair, or Trustworthy AI, have not included the effects of power in their considerations. As feminism has a long history of doing so, we introduce an intersectional and inclusive feminist approach to shape AI in a more equitable way. We approach this by building on recent Information Systems and interdisciplinary research as well as on evidence from expert interviews in focus groups, which we conducted in 2022 and 2023. Our study reveals that utilizing the feminist approach could be effective firstly, to shape AI systems and secondly, to change prevailing power structures in societal systems to become more equitable.

Track 1: Digital Responsibility: Social, Ethical & Ecological Implication of IS

Budget-Feasible Market Design for Biodiversity Conservation: Considering Incentives and Spatial Coordination

Eleni Batziou, Martin Bichler

How to best incentivize farmers to conserve biodiversity on private land is an important policy question. Conservation auctions provide a mechanism to elicit farmers’ opportunity costs, but their design is challenging and often suffer from low participation due to strategic complexity. Conservation auctions should ideally be incentive-compatible, address spatial synergies that maximize biodiversity gains, and respect the predefined budget of the government. Recent advances in mechanism design suggest budget-feasible auctions, but little is known about average-case efficiency. Based on this line of research, we introduce an incentive-compatible conservation auction mechanism that considers the bid taker’s spatial synergies and respects budget. The results are compared against the celebrated Vickrey-Clarke-Groves mechanism. Our numerical results estimate the efficiency loss that can be expected for different assumptions on the synergistic values of the government. They provide evidence that budget-feasible mechanisms provide a new tool for policymakers in this domain.

Track 1: Digital Responsibility: Social, Ethical & Ecological Implication of IS

Trust me, I’m an Intermediary! Exploring Data Intermediation Services 

Julia Christina Schweihoff, Ilka Jussen, Frederik Möller

Data ecosystems receive considerable attention in academia and practice, as indicated by a steadily growing body of research and large-scale (industry-driven) research projects. They can leverage so-called data intermediaries, which are mediating parties that facilitate data sharing between a data provider and a data consumer. Research has uncovered many types of data intermediaries, such as data marketplaces or data trusts. However, what is missing is a ‘big picture’ of data intermediaries and the functions they fulfill. We tackle this issue by extracting data intermediation services decoupled from specific instances to give a comprehensive overview of how they work. To achieve this, we report on a systematic literature review, contributing data intermediation services.

Track 3: Digital Markets, Platforms & Data Spaces

Improving the Efficiency of Human-in-the-Loop Systems: Adding Artificial to Human Experts

Johannes Jakubik, Daniel Weber, Patrick Hemmer, Michael Vössing, Gerhard Satzger

Information systems increasingly leverage artificial intelligence (AI) and machine learning (ML) to generate value from vast amounts of data. However, ML models are imperfect and can generate incorrect classifications. Hence, human-in-the-loop (HITL) extensions to ML models add a human review for instances that are difficult to classify. This study argues that continuously relying on human experts to handle difficult model classifications leads to a strong increase in human effort, which strains limited resources. To address this issue, we propose a hybrid system that creates artificial experts that learn to classify data instances from unknown classes previously reviewed by human experts. Our hybrid system assesses which artificial expert is suitable for classifying an instance from an unknown class and automatically assigns it. Over time, this reduces human effort and increases the efficiency of the system. Our experiments demonstrate that our approach outperforms traditional HITL systems for several benchmarks on image classification.

Track 5: Data Science & Business Analytics

Digital Service Innovation for Sustainable Development: A Systematic Literature Review

Daniel Heinz, Mingli Hu, Carina Benz, Gerhard Satzger

Creating and delivering products and services that promote sustainability is increasingly important in today’s economy. Novel services based on digital technologies and infrastructure can significantly contribute to sustainable development–as digitally enabled car-sharing services demonstrate: Increased asset utilization reduces production-related greenhouse gas emissions. However, there is still limited knowledge on how digital service innovation can purposefully be applied to promote sustainability. To address this gap, we conduct a systematic literature review and perform a qualitative inductive analysis of 50 articles covering digital service impact on social, environmental, and economic sustainability. We compile a comprehensive overview of real-world applications and identify five underlying mechanisms in which innovation with digital services can drive sustainable development. Thus, we aim to pave the way to purposefully conceive, design, and implement digital services for sustainability.

Track 8: Service Innovation, Engineering & Management

Towards Designing a NLU Model Improvement System for Customer Service Chatbots

Daniel Schloß, Ulrich Gnewuch, Alexander Maedche

Current customer service chatbots often struggle to meet customer expectations. One reason is that despite advances in artificial intelligence (AI), the natural language understanding (NLU) capabilities of chatbots are often far from perfect. In order to improve them, chatbot managers need to make informed decisions and continuously adapt the chatbot’s NLU model to the specific topics and expressions used by customers. Customer-chatbot interaction data is an excellent source of information for these adjustments because customer messages contain specific topics and linguistic expressions representing the domain of the customer service chatbot. However, extracting insights from such data to improve the chatbot’s NLU, its architecture, and ultimately the conversational experience requires appropriate systems and methods, which are currently lacking. Therefore, we conduct a design science research project to develop a novel artifact based on chatbot interaction data that supports NLU improvement.

Track 8: Service Innovation, Engineering & Management

Seizing the Potential of Algorithms: The Power of Personalized Persuasive Messages on the Use of Algorithmic Advice

Simon Asbach, Lorenz Graf-Vlachy, Andreas Fügener

Recommendations of algorithms are frequently superior to human judgment, but humans often fail to seize the full potential of algorithms. Research has made initial advances to increase the use of algorithmic advice, but achieved levels remain far from optimal. However, research has not yet considered user personality traits to develop interventions to increase the use of algorithms. We propose to communicate the algorithmic advice via a persuasive message that fits the user’s regulatory focus, i.e., the tendency towards promotion- versus prevention focus, to create regulatory fit. To test our hypotheses, we conduct an experiment with a forecasting task and 605 participants. Results support a positive effect of a persuasive message on the use of algorithmic advice. Further, results do not support a regulatory fit effect. Our article adds to theory and practice by developing and empirically testing novel interventions to increase the use of algorithms. Next research steps are discussed.

Track 9: Human Computer Interaction & Social Online Behavior

Towards a Taxonomy for Neighborhood Volunteering Management Platforms

Hans Betke, Michaelle Bosse, Fotini Tzavala-Reusch, Stephan Kuehnel, Stefan Sackmann

The management and organization of volunteering in the social sector have been strongly influenced by technological progress over the last two decades. New proposals for IT-based volunteering management platforms that draw on many elements of social media are appearing with increasing frequency. In this article, we analyzed the current state of the art and use a methodological approach to develop a taxonomy for classifying existing and emerging developments in the field. The taxonomy is intended to assist practitioners in selecting appropriate systems for their respective purposes as well as support researchers in identifying research gaps. The resulting research artifact has undergone an initial evaluation and can support maintaining a better overview in a growing subject area.

Track 9: Human Computer Interaction & Social Online Behavior

Executive AI Literacy: A Text-Mining Approach to Understand Existing and Demanded AI Skills of Leaders in Unicorn Firms

Marc Pinski, Thomas Hofmann, Alexander Benlian

Despite the growing relevance of artificial intelligence (AI) for busi-nesses, there is a lack of research on how top-level executives must be skilled in AI. Drawing on upper echelons theory, this paper explores executive AI literacy, defined as the combined AI skills of top-level executives, and its relevance for different executive roles. We conducted a text-mining analysis of 1,625 execu-tives’ online profiles and 1,033 executive job postings from unicorn firms re-trieved via web-scraping from an online professional social network. We find that AI skills are mostly required in product-related executive roles (vs. adminis-trative roles). Thus, we provide an AI-specific perspective complementing prior information systems research on executives, which asserts that (non-AI) IT is driven by administrative executive roles. Our paper contributes to AI literacy lit-erature by shedding light on the substance of executive AI literacy within firms. Lastly, we provide implications for AI-related information systems strategy.

Track 10: IT Strategy, Governance & Management

Toward Government as a Platform: An Analysis Method for Public Sector Infrastructure

Peter Kuhn, Liudmila Zavolokina, Dian Balta, Florian Matthes

Government as a Platform (GaaP) is a promising approach to the digital transformation of the public sector. In practice, GaaP is realized by platform-oriented infrastructures. However, despite successful examples, the transformation toward platform-oriented infrastructures remains challenging. A potential remedy is the analysis of existing public infrastructure regarding its platform orientation. Such an analysis can identify the gaps to an ideal platform-oriented infrastructure and, thus, support the transformation toward it. We follow the design science research methodology to develop a four-dimensional analysis method. We do so in three iterations, and, after each iteration, evaluate the method by its application to infrastructures in practice. With regard to theory, our results suggest extending GaaP conceptualizations with a specific emphasis on platform principles. With regard to practice, we contribute an analysis method that creates proposals for the improvement of infrastructures and, thus, supports the transformation toward GaaP.

Track 17: Smart Cities & Digital Government

Designing Unlearning Support Systems: A Requirements Catalog

Marco Di Maria, David Walter, Thorsten Schoormann, Ralf Knackstedt

As established routines and other forms of knowledge may prevent organizations to respond to changing situations, they need to learn how to break out of the old way of thinking and acting. While unlearning is among the most promising approaches to do so, there is only a limited understanding of the requirements for the effective design of unlearning support systems (USS). As part of a larger design research project, this research-in-progress paper reviews 41 articles on practical approaches for unlearning support and derives an initial catalog of 23 design requirements. Our work aims to guide designers in the implementation and operationalization of unlearning support.

Track 18: Digital Education & Learning

Addressing Learners’ Heterogeneity in Higher Education: An Explainable AI-based Feedback Artifact for Digital Learning Environments

Felix Haag, Sebastian A. Günther, Konstantin Hopf, Philipp Handschuh, Maria Klose, Thorsten Staake

Due to the advent of digital learning environments and the freedom they offer for learners, new challenges arise for students’ self-regulated learning. To overcome these challenges, the provision of feedback has led to excellent results, such as less procrastination and improved academic performance. Yet, current feedback artifacts neglect learners’ heterogeneity when it comes to prescriptive feedback that should meet personal characteristics and self-regulated learning skills. In this paper, we derive requirements from self-regulated learning theory for a feedback artifact that takes learners’ heterogeneity into account. Based on these requirements, we design, instantiate, and evaluate an Explainable AI-based approach. The results demonstrate that our artifact is able to detect promising patterns in data on learners’ behaviors and characteristics. Moreover, our evaluation suggests that learners perceive our feedback as valuable. Ultimately, our study informs Information Systems research in the design of future Explainable AI-based feedback artifacts that seek to address learners’ heterogeneity.

Track 18: Digital Education & Learning