Logo des Repositoriums

Auflistung nach Autor:in "Heuer, Hendrik"

1 - 10 von 11
  • Konferenzbeitrag
    Accessible Text Tools for People with Cognitive Impairments and Non-Native Readers: Challenges and Opportunities
    (Mensch und Computer 2023 - Tagungsband, 2023) Heuer, Hendrik; Glassman, Elena Leah
    Many people have problems with reading, which limits their ability to participate in society. This paper explores tools that make text more accessible. For this, we interviewed experts, who proposed tools for different stakeholders and scenarios. Important stakeholders of such tools are people with cognitive impairments and non-native readers. Frequently mentioned scenarios are public administration, the medical domain, and everyday life. The tools proposed by experts support stakeholders by improving how text is compressed, expanded, reviewed, and experienced. In a survey of stakeholders, we confirm that the scenarios are relevant and that the proposed tools appear helpful to them. We provide the Accessible Text Framework to help researchers understand how the different tools can be combined and discuss how individual tools can be implemented. The investigation shows that accessible text tools are an important HCI+AI challenge that a large number of people can benefit from.
  • Workshopbeitrag
    Audit, Don’t Explain – Recommendations Based on a Socio-Technical Understanding of ML-Based Systems
    (Mensch und Computer 2021 - Workshopband, 2021) Heuer, Hendrik
    In this position paper, I provide a socio-technical perspective on machine learning-based systems. I also explain why systematic audits may be preferable to explainable AI systems. I make concrete recommendations for how institutions governed by public law akin to the German TÜV and Stiftung Wartentest can ensure that ML systems operate in the interest of the public.
  • Konferenzbeitrag
    Auditing the Biases Enacted by YouTube for Political Topics in Germany
    (Mensch und Computer 2021 - Tagungsband, 2021) Heuer, Hendrik; Hoch, Hendrik; Breiter, Andreas; Theocharis, Yannis
    With YouTube’s growing importance as a news platform, its recommendation system came under increased scrutiny. Recognizing YouTube’s recommendation system as a broadcaster of media, we explore the applicability of laws that require broadcasters to give important political, ideological, and social groups adequate opportunity to express themselves in the broadcasted program of the service. We present audits as an important tool to enforce such laws and to ensure that a system operates in the public’s interest. To examine whether YouTube is enacting certain biases, we collected video recommendations about political topics by following chains of ten recommendations per video. Our findings suggest that YouTube’s recommendation system is enacting important biases. We find that YouTube is recommending increasingly popular but topically unrelated videos. The sadness evoked by the recommended videos decreases while the happiness increases.We discuss the strong popularity bias we identified and analyze the link between the popularity of content and emotions.We also discuss how audits empower researchers and civic hackers to monitor complex machine learning (ML)-based systems like YouTube’s recommendation system.
  • Konferenzbeitrag
    Data Leakage Through Click Data in Virtual Learning Environments
    (20. Fachtagung Bildungstechnologien (DELFI), 2022) Hartmann, Johanna; Heuer, Hendrik; Breiter, Andreas
    Unsupervised machine learning techniques are increasingly used to cluster students based on their activity in virtual learning environments. It is commonly assumed that clusters formed by click data merely represent the actions of users and do not allow inferring personal information about individual users. Based on an analysis of 18,660 students and 5.56 million data points from the Open University Learning Analytics Dataset, we show that clusters trained on "raw" click data are highly correlated with personal information like student success, course specifics, and student demographics. Our analysis demonstrates that these clusters allow conclusions about demographic variables like the previous education and the affluence of the residential area. Our investigation shows that apparently, objective click data can leak private attributes. The paper discusses the implications of this for the design of virtual learning environments, especially considering the legal requirements posed by the principle of data minimization of the EU GDPR.
  • Konferenzbeitrag
    Design Frictions on Social Media: Balancing Reduced Mindless Scrolling and User Satisfaction
    (Proceedings of Mensch und Computer 2024, 2024) Ruiz, Nicolas; León, Gabriela Molina; Heuer, Hendrik
    Design features of social media platforms, such as infinite scroll, increase users’ likelihood of experiencing normative dissociation — a mental state of absorption that diminishes self-awareness and disrupts memory. This paper investigates how adding design frictions into the interface of a social media platform reduce mindless scrolling and user satisfaction. We conducted a study with 30 participants and compared their memory recognition of posts in two scenarios: one where participants had to react to each post to access further content and another using an infinite scroll design. Participants who used the design frictions interface exhibited significantly better content recall, although a majority of participants found the interface frustrating. We discuss design recommendations and scenarios where adding design frictions to social media platforms can be beneficial.
  • Konferenzbeitrag
    Student Success Prediction and the Trade-Off between Big Data and Data Minimization
    (DeLFI 2018 - Die 16. E-Learning Fachtagung Informatik, 2018) Heuer, Hendrik; Breiter, Andreas
    This paper explores student’s daily activity in a virtual learning environment in the anonymized Open University Learning Analytics Dataset (OULAD). We show that the daily activity of students can be used to predict their success, i.e. whether they pass or fail a course, with high accuracy. This is important since daily activity can be easily obtained and anonymized. To support this, we show that the binary information whether a student was active on a given day has similar predictive power as a combination of the exact number of clicks on the given day and sensitive private data like gender, disability, and highest educational level. We further show that the anonymized activity data can be used to group students. We identify different student types based on their daily binarized activity and outline how educators and system developers can utilize this to address different learning types. Our primary stakeholders are designers and developers of learning analytics systems as well as those who commission such systems. We discuss the privacy and design implications of our findings for data mining in educational contexts against the background of the principle of data minimization and the General Data Protection Regulation (GDPR) of the European Union.
  • Workshopbeitrag
    UCAI 2022: Workshop on User-Centered Artificial Intelligence
    (Mensch und Computer 2022 - Workshopband, 2022) Buschek, Daniel; Hauptmann, Hanna; Heuer, Hendrik; Loepp, Benedikt; Riener, Andreas; Yigitbas, Enes
    The proliferation of AI-based techniques poses a range of new challenges for the design and engineering of intelligent and adaptive systems since they tend to act as black boxes and do not offer the user sufficient transparency, control, and interaction opportunities, which are considered major goals of user-centered design in the HCI field. This workshop aims at sharing and discussing recent developments at the intersection of HCI and AI, and at exploring novel methodological, technical, and interaction approaches. Researchers and practitioners with diverse disciplinary backgrounds can and should contribute to advancing the research agenda in this emerging field of human-centered artificial intelligence.
  • Workshopbeitrag
    UCAI 2023: Workshop on User-Centered Artificial Intelligence
    (Mensch und Computer 2023 - Workshopband, 2023) Buschek, Daniel; Frommel, Julian; Hauptmann, Hanna; Heuer, Hendrik; Loepp, Benedikt
    The proliferation of AI-based techniques poses a range of new challenges for the design and engineering of intelligent and adaptive systems since they tend to act as black boxes and do not offer the user sufficient transparency, control, and interaction opportunities, which are considered major goals of user-centered design in the HCI field. This workshop aims at sharing and discussing recent developments at the intersection of HCI and AI, and at exploring novel methodological, technical, and interaction approaches. Researchers and practitioners with diverse disciplinary backgrounds can and should contribute to addressing the challenges in this emerging field of human-centered artificial intelligence.
  • Konferenzbeitrag
    Unpacking a model: An interactive visualization of a text similarity algorithm for legal documents
    (Mensch und Computer 2019 - Tagungsband, 2019) Soroko, Daria; Ndöge, Nina; Al-Shafeei, Ahmed; Heuer, Hendrik
    This paper presents a functional prototype for an interactive web-based interface i_sift developed to foreground the decision-making process of an algorithm that detects similarities in legal texts through word embeddings. Using this as a case study in Computational Social Science, our goal is, first, to highlight the importance of making computational tools and methods transparent to social scientists. Secondly, we suggest an approach that accomplishes this using methods and principles from Interactive Machine Learning and the Algorithmic Experience framework.
  • Konferenzbeitrag
    Visualization Needs in Computational Social Sciences
    (Mensch und Computer 2019 - Tagungsband, 2019) Heuer, Hendrik; Polizzotto, Anna; Marx, Franziska; Breiter, Andreas
    With the advent of digital humanities and computational social sciences, machine learning techniques like topic modeling are increasingly employed by social scientists and humanities scholars. This poses the question what visualization needs these researchers have when confronted with such complex systems. In this paper, we investigate visualization needs in the context of the topic modeling algorithm Latent Dirichlet Allocation and the 950,000 articles of the New York Times corpus. We presented visualizations of how the topics in the newspaper changed over time to seven participants, who fulfilled three tasks with three visualization types. Qualitative interviews with the participants supported our assumptions that visualizations for these tasks need to be visually appealing, intuitively interpretable, and minimizing mental effort.