Parallel sessions
 

11:30 - 12:30 hrs

Location: Mission 1 

• “Responsible Digital Twins for Local Regional Governance” (DiSa project)
  Principal investigator: Dr. Mirko Tobias Schaefer (Utrecht Data School, UU)
  Presenters: Drs. Sofie Wilde de Ligny (UU)
   
• “Co-designing a fair digital asylum procedure” (DiSa project)
  Principal investigator: Dr. Koen Leurs (Media and Culture, UU)
  Presenters: Dr. Koen Leurs, Dr. Rianne Dekker (School of Governance, UU) and Kinan Alajak  (Department of Media and Culture, UU / Refugee Wellbeing and Integration Initiative RWII)

Abstract: Migration management is increasingly digitized and datafied. This project evaluates the Athene Pilot, which was put to a stop in November 2021, as an entry point to explore challenges and opportunities newcomers experience in the digital society. The Athene Pilot, developed by the Dutch border police with stakeholders aimed to automate assessment of forensic copies of digital devices of asylum seekers to improve identification as well as to detect indicators of terrorism and human smuggling. In this presentation we reflect on the intricacies of researching black-boxed government AI and algorithms through the eyes of relevant stakeholders; present emerging findings of the technical, bureaucratic and everyday practice dimensions of the asylum procedure; and discuss knowledge transfer beyond academia.

Location: Mission 2

• “Co-Designing an Inclusive Collaborative Mapping Tool with People with Disabilities” (DiSa project)
  Principal investigator: Dr. Johannes Flacke (UT)
  Presenters: Prof. dr. Karin Pfeffer (UT) and Rob Wagemans (Toegankelijk Zwolle)

Abstract: Our project aims to engage people with disabilities in a co-design process to develop a prototype of an inclusive digital mapping tool for the collaborative design of public spaces. The main goals are to understand how people with disabilities can contribute to the design of inclusive public spaces and to learn about design requirements of people with disabilities for inclusive digital mapping tools. We do this with co-researchers from Toegankelijk Zwolle to create a prototype of an open-source tool that is contextualized for achieving inclusive public spaces in Zwolle. We also develop a set of generalized design requirements of inclusive digital mapping tools, and generate insights into the needs and contributions of people with disabilities to public spaces accessible for all.

• “Strengthening cyber-resilience by technological citizenship” (NWO project)
  Principal investigator: Dr. Lamber Royakkers (Eindhoven University of Technology)
  Presenters: Drs. Anne Marte Gardenier

Abstract: The digitization of society comes with opportunities and risks. In our research, we investigate what is needed to strengthen the cyber-resilience of citizens in the digital age, so the opportunities can be seized, and the risks averted. In this presentation, we present what we have learned from our collaboration with researchers and clinical staff affiliated with the FORSEE project in the Catharina hospital in Eindhoven. In this project, a machine learning driven monitoring camera that automatically detects patient deterioration is developed, and we aim to include insights and concerns from doctors, nurses and patients in the development process.

 

• “Informed citizenship for all” (NWO project)
  Principal investigator: Prof. dr. Marcel Broersma (RUG)
  Presenters: Denise Mensonides MA and Alex Smit MA

14:30 - 15:30 hrs

Location: Mission 1

• “The Power of Words: The Role of Mediators; Language in Increasing Intergenerational Empathy in Online Discussions” (NWO project)
  Principal investigator: Dr. Dong Nguyen (UU) and Dr. T. A. van den Broek (VU)
  Presenters: Dr. Dong Nguyen

Abstract: Many urgent societal challenges, like climate change, require communities to align their interests, but online discussions sometimes escalate. In this project, computer scientists and social scientists collaborate to analyze online conversations and to develop both data- and theory-driven interventions that will help bring communities together online. In this presentation, we reflect on one key challenge we have encountered so far: checking whether the text analysis models actually measure what we want them to measure. We will illustrate this with two case studies: detecting hate speech and measuring linguistic style similarity.

• “Constructive behavior online? "No idea, but we know what it is when we see it" (NWO project)
  Principal investigator: Prof. dr. Emiel Krahmer (UvT)
  Presenters: Dr. Cedric Waterschoot and Dr. Ernst van den Hemel

Abstract: In our talk we outline how news outlets struggle to define and promote 'constructive' behavior online. We argue that scholars can help moderators in their struggle by applying a mixed methods methodology: combining expertise on AI applications with representative panel data in order to form user profiles of commenters, ethnographic study of the lived reality of moderators and the cultural environments of news outlets.

Location: Mission 2

• “Including and Supporting Vulnerable Women through Conversation: Innovative Algorithm Design that Fosters Resilience in a Safe Space''
  Presenters: Dr. Vanessa Dirksen (OU) and Dr. Rachelle Bosua (OU, The University of Melbourne)

Abstract: Over the last decade there has been a significant rise in digital platforms that support many different aspects of remote work and business service delivery. Central to platforms are AI algorithms that mediate digital services to control and manage workers, their tasks and business transactions. This project focuses on social innovation through AI algorithm design embedded in digital platforms for the Common (Social) Good of marginal groups. Drawing on our collective experience related to limitations of Social Media platforms and corresponding algorithms in healthcare management, we reason for innovation in AI algorithm design to broaden platform support services for social impact. The first stage of the research will be a qualitative study that interweaves Design Science Research (DSR) and ethnographic narrative exploration to identify key elements of innovative AI algorithm design that are required for social support. We believe our insights might be valuable for social entrepreneurs who seek to build supportive platforms drawing on a new generation of AI algorithms that foster resilience of vulnerable women.

• “Causal Human-in-the-loop explainable AI: connecting philosophy and computer science on explainability”
  Presenters: Dr. Jie Yang and Dr. Stefan Buijsman (TUD)

Abstract: Explaining the behavior of machine learning models is an important aspect of the responsible use of these models, especially in high-risk contexts. However, clear standards for these explanations are lacking, as well as methods to generate good explanations. Our work uses the extensive work on explanations in the philosophy of science, where definitions and standards have been actively debated the last decades, to make progress in XAI. The philosophical frameworks are linked to human-in-the-loop explainability methods to develop new XAI methods that provide the right kinds of explanations. We will present this combination of perspectives, as well as the prototype XAI tool that resulted from it.

Location: Expedition

• “Microwork: The human labor necessary to make AI function".
  Presenter: Prof. Claartje Ter Hoeven (EUR)
   
• "Doing microwork: task quality, communities, and their impact on microwork market dynamics”Presenters: Prof. Alessandro Bozzon (TUD)