Tagungsband MuC 2024
See proceedings in the digital library of the ACM at https://dl.acm.org/doi/proceedings/10.1145/3670653
Autor*innen mit den meisten Dokumenten
Auflistung nach:
Neueste Veröffentlichungen
- KonferenzbeitragFrom Text to Treatment: How Medical Discharge Letters Are Used as a Key Artifact for Managing Patient Care(Proceedings of Mensch und Computer 2024, 2024) Zakreuskaya, Anastasiya; Buschek, Daniel; Mackay, Wendy E.; Avellino, Ignacio; Dove, Graham; Eskofier, Bjoern MHospital physicians must navigate through vast quantities of patient information represented in text-based reports. Although intended to improve patient care, their effectiveness hinges on each physician’s ability to successfully handle and interpret fragmented information from diverse sources. The increasing automation of text interactions are a potential support but are still at the early phase of implementation in real-world scenarios. We observed 144 hours of clinical shifts in a German internal medicine hospital and collected structured field notes on physicians’ current practices with text-based reports to enrich existing understanding of the requirements for including automation to clinical text. We identified medical discharge letters as most frequently consulted text document and a qualitative analysis of the field notes revealed that this document acts as a key artifact that serves different roles and purposes in the hospitalization of a patient. Based on our findings we discuss possible loss of these nuanced uses through automation and propose design implications for medical text reports.
- KonferenzbeitragUnderstanding the Adoption of ChatGPT in Higher Education: A Comparative Study with Insights from STEM and Business Students(Proceedings of Mensch und Computer 2024, 2024) Kubullek, Ann-Kathrin; Kumaç, Nadire; Dogangün, AysegülSince ChatGPT’s introduction, generative artificial intelligence (AI) has significantly influenced the media, technological innovation, and educational discourse. Its increasing importance, especially in academia, necessitates a detailed examination of the impact of AI on higher education, particularly on how it changes teaching and learning processes. This study therefore looks at the factors affecting students’ attitudes towards AI technologies in the university setting, with a particular focus on the differences between business and STEM programmes. Using a mixed methods approach, the study combines surveys and interviews to collect data on students’ perceptions, attitudes and experiences with generative AI technology in academia. The data collected is analysed both quantitatively and qualitatively to reveal significant trends and insights into the adoption and use of generative AI tools in the university environment. The main objective of the study is to shed light on the determinants that determine the varying degrees of AI adoption in different academic disciplines. The findings have the potential to inform the implementation of educational technology and assist in the development of strategies for the effective integration of generative AI tools to meet the different needs and preferences of students in a range of academic contexts.
- KonferenzbeitragCoLoTiMa: A Cognitive-Load Based Time Management Tool(Proceedings of Mensch und Computer 2024, 2024) Maleck, Moritz; Gross, TomTime management has the potential to maintain the learning and working productivity. Prominent techniques—such as Pomodoro—typically suggest to alternate productive periods and breaks. They are mostly time-based and lack adaptability to individual preferences and cognitive workloads. In the context of learning, this leads to suboptimal learning experiences, with rigid time structures hindering productivity and reducing efficiency. We introduce CoLoTiMa, a novel approach that dynamically adjusts learning period durations. It integrates real-time cognitive-load measurements and user self-assessment to tailor learning experiences. Through the use of eye-tracking, CoLoTiMa optimises the duration of learning blocks in accordance with individual learning preferences and thus fosters personalised and efficient outcomes.
- KonferenzbeitragAI-based Tools in Higher Education: A Comparative Analysis of University Guidelines(Proceedings of Mensch und Computer 2024, 2024) Hofmann, Paula; Brand, Alexa; Späthe, Eva; Lins, Sebastian; Sunyaev, AliIm Bildungswesen werden verstärkt KI-basierte Tools wie ChatGPT eingesetzt. Allerdings sind viele Studierende und Lehrende unsicher, ob, wie und in welchem Maß sie diese Tools im Hochschulkontext einsetzen dürfen. Insgesamt mangelt es in Deutschland an Universitäten an Richtlinien zum Umgang mit KI-basierten Tools. Aus diesem Grund führt diese Studie eine vergleichende Analyse von bereits existierenden Richtlinien durch, um die wichtigsten Empfehlungen für den Umgang zu extrahieren und zu aggregieren. Die Ergebnisse zeigen, dass die Relevanz von Richtlinien hoch ist und dabei insbesondere geklärt werden sollte, unter welchen Bedingungen KI-basierte Tools als Hilfsmittel gelten, welche Verantwortlichkeiten bei den Akteuren liegen und wie Risiken und Herausforderungen begegnet werden können, um u.a. die akademische Integrität sicherzustellen. Die Ergebnisse der Arbeit unterstützen bei der Ableitung und Synthese von Richtlinien im Hochschulkontext.
- KonferenzbeitragLooking Through the Deep Glasses: How Large Language Models Enhance Explainability of Deep Learning Models(Proceedings of Mensch und Computer 2024, 2024) Spitzer, Philipp; Celis, Sebastian; Martin, Dominik; Kühl, Niklas; Satzger, GerhardAs AI becomes more powerful, it also becomes more complex. Traditionally, eXplainable AI (XAI) is used to make these models more transparent and interpretable to decision-makers. However, research shows that decision-makers can lack the ability to properly interpret XAI techniques. Large language models (LLMs) offer a solution to this challenge by providing natural language text in combination with XAI techniques to provide more understandable explanations. However, previous work has only explored this approach for inherently interpretable models–an understanding of how LLMs can assist decision-makers when using deep learning models is lacking. To fill this gap, we investigate how different augmentation strategies of LLMs assist decision-makers in interacting with deep learning models. We evaluate the satisfaction and preferences of decision-makers through a user study. Overall, our results provide first insights into how LLMs support decision-makers in interacting with deep learning models and open future avenues to continue this endeavor.
- KonferenzbeitragKey persons in the selection and implementation of UX processes for the development of digital products: Schlüsselpersonen bei der Auswahl und Implementierung von UX-Prozessen zur Entwicklung digitaler Produkte(Proceedings of Mensch und Computer 2024, 2024) Niels, Adelka; Fortmann, Jutta; Trageiser, EileenMany companies face the challenge of identifying and effectively implementing a suitable UX process that meets their individual requirements and organisational contexts. This article analyses which stakeholders in companies significantly influence the selection and implementation of UX processes in the development of digital products. Based on an online survey, the degree of involvement of various stakeholders is analysed. The study shows that UX and design specialists in particular, as well as product management, play a central role. It also shows that many companies favour flexible UX processes and prefer internal expertise. The research emphasises the need for practice-oriented frameworks and tools to support the decision-making and implementation of UX processes. The insights gained are to be incorporated into the development of a solution that is customised to the specific needs and UX maturity levels of companies in order to enable more effective UX integration.
- KonferenzbeitragDo You Dare to Ask? Influence of Question Recipient and Information Medium on Prompting Preparatory Questions for MR Imaging(Proceedings of Mensch und Computer 2024, 2024) Kamps, Jonas Carl Christian; Birk, Max V.; Vriens, Chantal; Wallner, Günter; Bernhaupt, ReginaPatient portals empower patients to schedule appointments, ask questions, and provide information for upcoming procedures. While patient portals provide advantages for cooperation between medical institutions and patients, provided information can be unengaging. Interactive alternatives include question prompting, which can elicit helpful questions patients would otherwise not ask; and simulations, which have been used to engage patients when informing them about medical procedures. To investigate barriers of simulations and question prompting to engage patients when informing them about a medical procedure, we developed an MRI (Magnetic Resonance Imaging) virtual walkthrough and compared task load of this procedural simulation with a standard information leaflet. After receiving information either through the virtual walkthrough or leaflet, we asked 105 participants recruited through Amazon Mechanical Turk to ask questions about the procedure either to an expert or to a peer. Comparing virtual walkthrough and leaflet, we find positive effects of the walkthrough on mental load. Investigating the linguistic properties of questions, we find experts received more technical questions; peer questions focused on the procedure’s emotional content. Our results give insight into linguistic differences in questions, contribute practical knowledge to prompt questions, and discuss the potential of virtual walkthroughs and peer questions for preparing medical procedures.
- KonferenzbeitragNachhaltigkeit gestalten: Aktivierung der Generation Z für ökologisches Engagement durch Persuasive Computing(Proceedings of Mensch und Computer 2024, 2024) Hermann, Julia; Mäder, Aiden Danny; Dogangün, AysegülAngesichts des dringenden Bedarfs, nachhaltiges Verhalten in der Gesellschaft zu fördern, gewinnen Technologien zur Unterstützung positiver Verhaltensänderungen an Bedeutung. Generation Z, bekannt für ihre technologische Affinität und ihr Umweltbewusstsein, steht im Mittelpunkt unserer Betrachtung. Unser Paper berichtete den Einsatz persuasiver Computing-Strategien zur Förderung ökologischer Nachhaltigkeit innerhalb dieser Generation. Im Rahmen einer Lehrveranstaltung zu User Experience für mobile Endgeräte entwickelten Studierende der Generation Z App- und haptische Prototypen anhand von Anforderungen ihrer Altersgenoss:innen. Interviews und Evaluationen zeigten Gründe für Zurückhaltung gegenüber umweltfreundlichen Lebensstilen und lieferten Implikationen für das Design persuasiver Technologien. Die Ergebnisse bieten erste Einblicke in die Gestaltung von Technologien, die Generation Z für Umwelt- und Nachhaltigkeitsthemen mobilisieren können.
- KonferenzbeitragEstablishing Foundations on Both Sides of the Bridge: Exploring Contextual Barriers to Cardiac Rehabilitation Uptake to Inform Digital Health Technology Design(Proceedings of Mensch und Computer 2024, 2024) Höppchen, Isabel; Kulnik, Stefan Tino; PhD., Bernhard Reich; Niebauer, Josef; Smeddinck, Jan; Meschtscherjakov, Alexander; Wurhofer, DanielaCardiac rehabilitation (CR) is crucial for secondary prevention of cardiovascular diseases. Yet, many patients with a clear indication of CR do not participate in such programs. To build a more nuanced foundation for developing digital health technologies supporting patients’ CR uptake, we explored the practiced CR referral process to identify context-specific barriers in Austria. We conducted a contextual inquiry including four observations in both out- and inpatient settings and 13 stakeholder interviews. The results indicate a lack of a defined referral process, limited intersectoral communication, doctors’ unawareness and knowledge gaps, and patients missing clear information about CR. Thus, digital health technologies that aim to support patient-centered cardiovascular care pathways should be designed for timely patient information and engagement, automated referral support, and enhanced intersectoral communication.
- KonferenzbeitragSearch User Interaction in Multi-Theme Map-based Applications: A Preliminary Assessment(Proceedings of Mensch und Computer 2024, 2024) Degbelo, Auriol; Schmidt, Benno; Vuong, Johnni; Henzen, Christin; Zander, Franziska; Lechler, Sarah; Lier, BernadetteMap-based web applications are often used to display spatial information in the context of (open) data from the environmental and urban infrastructure sectors. As the number of thematic layers displayed on these applications grows, maintaining user interface intuitiveness becomes increasingly challenging. To provide an initial account of design practices in the area, this article examines 19 map-based applications with multiple thematic layers, along four dimensions: thematic content complexity, functionalities for layer search, legend design and functionalities for the search of location information. We then report on findings from a brainwriting workshop involving 11 participants, during which three applications were used to identify design challenges and opportunities. The observations and design issues presented are relevant to researchers and designers of applications displaying multi-theme spatial information.