Steffen Koch

Case studies are central to evaluating visualization research as they provide evidence for how the developed approaches support real users, real analytical work, and real data. Conducting such studies can be challenging, since target users with relevant domain expertise are often scarce or even entirely unavailable, while the visualization researchers may lack the required expertise to perform the evaluation themselves. Current research on Large Language Models (LLMs) has shown their strong reasoning ability and usefulness in domain‐specific downstream tasks, provoking the question: to what degree and with what capacity can LLMs fill the role of the target users in visualization case studies? We investigate this question by evaluating how LLMs can participate across different phases of visualization case studies. We propose a conceptual framework that explores the different levels of LLM involvement and potential roles when simulating target users. To sketch where the LLM substitution is most plausible, we embed our suggested integration of LLMs within the nested model for visualization theory and design. For evaluation, we replicate case studies from published and unpublished visualization research using multiple state‐of‐the‐art LLMs and compare the model‐generated insights to those reported in the papers. Our results show that LLMs can often generate plausible and sometimes even novel interpretations of visual patterns when used for result analysis and validation, but can also struggle when highly contextual, domain‐specific knowledge is required.

Large transparent touch screens (LTTS) have recently become commercially available. These displays have the potential for engaging Augmented Reality (AR) applications, especially in public and shared spaces. However, the interaction with objects in the real environment behind the display remains challenging: Users must combine pointing and touch input if they want to select objects at varying distances. There is a lot of work on wearable or mobile AR displays, but little on how users interact with LTTS. Our goal is to contribute to a better understanding of natural user interaction for these AR displays. To this end, we developed a prototype and evaluated different pointing techniques for selecting 12 physical targets behind an LTTS, with distances ranging from 6 to 401 cm. We conducted a user study with 16 participants and measured user preferences, performance, and behavior. We analyzed the change in accuracy depending on the target position and the selection technique used. Our fndings include: (a) Users naturally align the touch point with their line of sight for targets farther than 36 cm behind the LTTS. (b) This technique provides the lowest angular deviation compared to other techniques. (c) Some user close one eye to improve their performance. Our results help to improve future AR scenarios using LTTS systems.

Sports visualization has developed into an active research field over the last decades. Many approaches focus on analyzing movement data recorded from unstructured situations, such as soccer. For the analysis of choreographed activities like formation dancing, however, the goal differs, as dancers follow specific formations in coordinated movement trajectories. To date, little work exists on how visual analytics methods can support such choreographed performances. To fill this gap, we introduce a new visual approach for planning and assessing dance choreographies. In terms of planning choreographies, we contribute a web application with interactive authoring tools and views for the dancers' positions and orientations, movement trajectories, poses, dance floor utilization, and movement distances. For assessing dancers' real‐world movement trajectories, extracted by manual bounding box annotations, we developed a timeline showing aggregated trajectory deviations and a dance floor view for detailed trajectory comparison. Our approach was developed and evaluated in collaboration with dance instructors, showing that introducing visual analytics into this domain promises improvements in training efficiency for the future.

Temporal action localization aims to identify the boundaries and categories of actions in videos, such as scoring a goal in a football match. Single-frame supervision has emerged as a labor-efficient way to train action localizers as it requires only one annotated frame per action. However, it often suffers from poor performance due to the lack of precise boundary annotations. To address this issue, we propose a visual analysis method that aligns similar actions and then propagates a few user-provided annotations (e.g., boundaries, category labels) to similar actions via the generated alignments. Our method models the alignment between actions as a heaviest path problem and the annotation propagation as a quadratic optimization problem. As the automatically generated alignments may not accurately match the associated actions and could produce inaccurate localization results, we develop a storyline visualization to explain the localization results of actions and their alignments. This visualization facilitates users in correcting wrong localization results and misalignments. The corrections are then used to improve the localization results of other actions. The effectiveness of our method in improving localization performance is demonstrated through quantitative evaluation and a case study.

Sports visualization has developed into an active research field over the last decades. Many approaches focus on analyzing movement data recorded from unstructured situations, such as soccer. For the analysis of choreographed activities like formation dancing, however, the goal differs, as dancers follow specific formations in coordinated movement trajectories. To date, little work exists on how visual analytics methods can support such choreographed performances. To fill this gap, we introduce a new visual approach for planning and assessing dance choreographies. In terms of planning choreographies, we contribute a web application with interactive authoring tools and views for the dancers' positions and orientations, movement trajectories, poses, dance floor utilization, and movement distances. For assessing dancers' real-world movement trajectories, extracted by manual bounding box annotations, we developed a timeline showing aggregated trajectory deviations and a dance floor view for detailed trajectory comparison. Our approach was developed and evaluated in collaboration with dance instructors, showing that introducing visual analytics into this domain promises improvements in training efficiency for the future.

The number of social media posts and news articles that are being published every day is high. This makes them an attractive source of human-generated information for different domain experts such as journalists and business analysts but also emergency responders, particularly if posts contain references to geolocations. Visual analytics approaches can help to gain insights into such datasets and inform decision-makers. However, the high volume and the veracity of the data, as well as the velocity in the case of streaming data, pose challenges when supporting explorative analysis with interactive visualization. Based on four exemplary approaches, we outline recently proposed strategies to tackle these challenges. We describe how geo-aware filtering and anomaly detection methods can help to inform stakeholders based on geolocated tweets. We show that data-aware tag maps can provide analysts with an overview-first, details-on-demand visual summary of large amounts of text content over time. With space-filling curves, we can visualize the temporal evolution of geolocations in a two-dimensional plot without relying on animations that would impede comparative analyses. Additionally, we discuss the use of an efficient dynamic clustering algorithm for enabling large-scale visual analyses of streaming posts.