EuroVA2025

Permanent URI for this collection

https://diglib.eg.org/handle/10.2312/3607188

Browse

Now showing 1 - 12 of 12

Evaluating Autoencoders for Parametric and Invertible Multidimensional Projections
(The Eurographics Association, 2025) Dennig, Frederik L.; Geyer, Nina; Blumberg, Daniela; Metz, Yannick; Keim, Daniel A.; Schulz, Hans-Jörg; Villanova, Anna
Recently, neural networks have gained attention for creating parametric and invertible multidimensional data projections. Parametric projections allow for embedding previously unseen data without recomputing the projection as a whole, while invertible projections enable the generation of new data points. However, these properties have never been explored simultaneously for arbitrary projection methods. We evaluate three autoencoder (AE) architectures for creating parametric and invertible projections. Based on a given projection, we train AEs to learn a mapping into 2D space and an inverse mapping into the original space. We perform a quantitative and qualitative comparison on four datasets of varying dimensionality and pattern complexity using t-SNE. Our results indicate that AEs with a customized loss function can create smoother parametric and inverse projections than feed-forward neural networks while giving users control over the strength of the smoothing effect.
Integrating Layer-Wise Relevance Propagation with Stable Diffusion for Enhanced Interpretability
(The Eurographics Association, 2025) Auman, Christian; Bhati, Deepshikha; Arquilla, Kyle; Neha, Fnu; Guercio, Angela; Schulz, Hans-Jörg; Villanova, Anna
Diffusion-based generative models, such as Stable Diffusion and DALL-E, have revolutionized artificial intelligence by enabling high-quality image generation from textual descriptions. Despite their success, these models raise ethical concerns, such as style appropriation and misuse, closely tied to the interpretability and transparency of the underlying mechanisms. This paper introduces a framework integrating Layer-wise Relevance Propagation (LRP) into the Stable Diffusion model to enhance interpretability. LRP assigns relevance scores to specific elements of textual prompts, allowing users to understand and visualize how input text influences image generation. We also present an interactive web-based visualization tool that supports intuitive exploration of diffusion processes. By improving interpretability, this approach fosters responsible use of generative AI technologies. A user study involving 35 participants demonstrates the tool's accessibility and effectiveness.
Towards Integrating Visual Analytics in Multi-Perspective Conformance Checking: A Call to Action
(The Eurographics Association, 2025) Linden, Sanne van der; Filipov, Velitchko Andreev; Pufahl, Luise; Miksch, Silvia; Elzen, Stef van den; Schulz, Hans-Jörg; Villanova, Anna
The research fields of Process Mining (PM) and Visual Analytics (VA) can mutually benefit from each other by combining their strengths. PM tasks include process discovery, enhancement, and conformance checking. This paper focuses on conformance checking, where the event log is compared against a reference model to identify potential deviations in process behavior. Conformance checking is often limited to analyzing the control flow (i.e., sequences of activities), while other relevant perspectives present in the data, such as resources and time, are frequently overlooked. These additional perspectives are crucial to form a holistic understanding of deviations and their underlying causes. To address these limitations, we propose a conceptual framework and explore future opportunities for integrating VA with PM to support conformance checking from multiple perspectives. Our contribution emphasizes interactive visualization and analysis for a more flexible and iterative conformance checking process by, for example, allowing to dynamically refine and define additional constraints based on insights from multiple perspectives and making all deviations explainable and understandable.
Visually Exploring Team Communication and Gameplay Events in League of Legends
(The Eurographics Association, 2025) Glombik, David; Beck, Fabian; Agarwal, Shivam; Wallner, Günter; Schulz, Hans-Jörg; Villanova, Anna
Popular team-based esport games such as League of Legends rely on effective verbal team communication. Analyzing and reflecting on the own communication behavior is hence a relevant optimization strategy for a team, but needs to be sufficiently contextualized through game events. In this work, we present an analysis approach to investigate both communication and game events in an integrated visual analytics system. Aside to providing overview statistics, this novel blend builds on timeline and word cloud representations to show the rich data from League of Legends game sessions, visually comparing the opposing teams. We demonstrate the approach through analyzing relevant communication patterns in an application example.
Cutting Through the Clutter: The Potential of LLMs for Efficient Filtration in Systematic Literature Reviews
(The Eurographics Association, 2025) Joos, Lucas; Keim, Daniel A.; Fischer, Maximilian T.; Schulz, Hans-Jörg; Villanova, Anna
Systematic literature reviews (SLRs) are essential but labor-intensive due to high publication volumes and inefficient keywordbased filtering. To streamline this process, we evaluate Large Language Models (LLMs) for enhancing efficiency and accuracy in corpus filtration while minimizing manual effort. Our open-source tool LLMSurver presents a visual interface to utilize LLMs for literature filtration, evaluate the results, and refine queries in an interactive way. We assess the real-world performance of our approach in filtering over 8.3k articles during a recent survey construction, comparing results with human efforts. The findings show that recent LLM models can reduce filtering time from weeks to minutes. A consensus scheme ensures recall rates >98.8%, surpassing typical human error thresholds and improving selection accuracy. This work advances literature review methodologies and highlights the potential of responsible human-AI collaboration in academic research.
Scalable Force Scheme: a fast method for projecting large datasets
(The Eurographics Association, 2025) Ros, Jaume; Arleo, Alessio; Paulovich, Fernando V.; Schulz, Hans-Jörg; Villanova, Anna
Global dimensionality reduction (DR) methods are widely used to project high-dimensional data into a low-dimensional representation, preserving the overall structure of the dataset. Global nonlinear DR techniques allow one to capture complex features of the data but are limited by their high computational cost, making them an unfeasible choice to process large datasets. Force scheme (FS) is one of the most popular of such examples, being adopted in a wide variety of domains, but limiting its application to small datasets. In this paper, we extend FS to improve its convergence quality and speed, by introducing several concepts from gradient descent (GD) theory and lowering its algorithmic complexity. Our new proposed method is less prone to generate distorted projections due to the presence of artifacts, while significantly improving the running times, allowing for nonlinear global DR projections of large datasets.
We Should Change How We Measure User Experience in Visual Analytics Systems
(The Eurographics Association, 2025) Victorelli, Eliane Zambon; Cabouat, Anne-Flore; Santos, Emanuele; Cabric, Florent; Isenberg, Petra; Schulz, Hans-Jörg; Villanova, Anna
The evaluation of user experience in visual analytics systems is a complex and multifaceted challenge that demands a specialized approach. Traditional user experience measurement tools, including standardized questionnaires, often fail to capture the unique interactions and cognitive demands of visual analytics environments. In this position paper, we argue that it is necessary to develop a new measurement tool specifically tailored to user experience in visual analytics systems. We explore the relationship between user experience and usability and review current methods for collecting, measuring, and analyzing user experience data. We discuss the limitations of existing methods for user experience evaluations of visual analytics systems with a focus on the problematic use of current user experience questionnaires. Our research agenda outlines the key challenges for developing a dedicated evaluation instrument and proposes strategies for exploring the dimensions of user experience relevant to the visual analytics systems context. In particular, future research on a framework for developing robust and context-appropriate measurement tools that target the many important dimensions of user experience with visual analytics systems is needed.
EuroVa 2025: Frontmatter
(The Eurographics Association, 2025) Schulz, Hans-Jörg; Villanova, Anna; Schulz, Hans-Jörg; Villanova, Anna
Investigating Control Areas in Table Tennis
(The Eurographics Association, 2025) ERADES, Aymeric; Peuch, L.; Vuillemot, Romain; Schulz, Hans-Jörg; Villanova, Anna
Control areas are models designed to determine which portions of space can be reached by a moving entity. Such models have powerful applications in various domains where spatio-temporal information is key, ranging from urban analysis to sports spatial analysis. In this article, we explore the use of these models in table tennis to understand player strategies. We build upon existing models, originally designed for large-field or team sports, and adapt them to the adversarial context of table tennis-where the goal is to determine which regions a player can effectively return the ball to. In particular, we account for player reachability using a peripheral model that captures arm and racket positions. We report on an early evaluation of our model using TV broadcast videos and discuss potential improvements for our models.
The Human-Data-Model Interaction Canvas for Visual Analytics
(The Eurographics Association, 2025) Bernard, Jürgen; Schulz, Hans-Jörg; Villanova, Anna
Visual Analytics (VA) integrates humans, data, and models as key actors in insight generation and data-driven decision-making. This position paper values and reflects on 16 VA process models and frameworks and makes nine high-level observations that motivate a fresh perspective on VA. The contribution is the HDMI Canvas, a perspective to VA that complements the strengths of existing VA process models and frameworks. It systematically characterizes diverse roles of humans, data, and models, and how these actors benefit from and contribute to VA processes. The descriptive power of the HDMI Canvas eases the differentiation between a series of VA building blocks, rather than describing general VA principles only. The canvas includes modern humancentered methodologies, including human knowledge externalization and forms of feedback loops, while interpretable and explainable AI highlight model contributions beyond their conventional outputs. The HDMI Canvas has generative power, guiding the design of new VA processes and is optimized for external stakeholders, improving VA outreach, interdisciplinary collaboration, and user-centered design. The utility of the HDMI Canvas is demonstrated through two preliminary case studies.
A Design Space for the Critical Validation of LLM-Generated Tabular Data
(The Eurographics Association, 2025) Sachdeva, Madhav; Narayanan, Christopher; Wiedenkeller, Marvin; Sedlakova, Jana; Bernard, Jürgen; Schulz, Hans-Jörg; Villanova, Anna
LLM-generated tabular data is creating new opportunities for data-driven applications in academia, business, and society. To leverage benefits like missing value imputation, labeling, and enrichment with context-aware attributes, LLM-generated data needs a critical validation process. The number of pioneering approaches is increasing fast, opening a promising validation space that, so far, remains unstructured. We present a design space for the critical validation of LLM-generated tabular data with two dimensions: First, the Analysis Granularity dimension-from within-attribute (single-item and multi-item) to acrossattribute perspectives (1×1, 1×m, and n×n). Second, the Data Source dimension-differentiating between LLM-generated values, ground truth values, explanations, and their combinations. We discuss analysis tasks for each dimension cross-cut, map 19 existing validation approaches, and discuss the characteristics of two approaches in detail, demonstrating descriptive power.
Guided Visual Analysis of Time Series Data with Spiral Views and View Quality Measures
(The Eurographics Association, 2025) Stoppacher, Stefanie; Rakuschek, Julian; Schreck, Tobias; Schulz, Hans-Jörg; Villanova, Anna
Seasonal variations in energy consumption and temperature, like many other time series, exhibit periodically repeating patterns. Identifying and analyzing these cyclic patterns is crucial for understanding underlying trends and predicting future behavior. Spiral visualizations are commonly used to highlight periodicity, as they intuitively arrange seasonal data in spirals. We introduce encompassing user-guided enhancements to spiral visualizations, supporting the search and analysis of patterns in cyclic time series. A key element is a parameter space visualization by an interactive heat map, which highlights important quality measures, such as similarity and monotonicity, across different segments of the spiral. This approach helps users efficiently locate areas of interest that meet specific criteria, thereby streamlining the discovery of significant patterns. To further support analysis, the system offers a linked stacked area or line chart representation of selected segments, providing a clearer understanding of the quality measures. The effectiveness of the quality measures is demonstrated by use cases on several datasets.

Browse

Browsing EuroVA2025 by Issue Date

Results Per Page

Sort Options