IVAPP 2025 Abstracts

Area 1 - Information Visualization

Full Papers

Paper Nr:	63
Title:	Evaluating Transformers Learning by Representing Self-Attention Weights as a Graph
Authors:	Rebecca Leygonie, Sylvain Lobry and Laurent Wendling
Abstract:	Transformers architectures have established themselves as the state of the art for sequential data processing, with applications ranging from machine translation to the processing of Electronic Health Records (EHR). These complex data present a particular challenge in terms of explainability, which is a crucial aspect for their adoption in the healthcare field, subject to strict ethical and legal requirements. To address this challenge, we propose an approach to represent learning through graphs by exposing the self-attention links between tokens. We introduce a metric to assess the relevance of the connections learned by the model, in comparison with medical expertise. We apply our approach to the Behrt model, designed to predict future hospital visits based on sequences of previous visits, trained on data from the French National Health Data System. Our experiments show that our method facilitates understanding of model learning, and enables a better appreciation of the influence of diagnoses on each other, as well as of the biases present in the data, than global model evaluation measures.
Download

Paper Nr:	231
Title:	RapViz: Rhyme Detection and Visualization of Rap Music
Authors:	Paul Müller, Lukas Panzer and Fabian Beck
Abstract:	RapViz transforms lyrics and audio of rap music into interactive visualizations, highlighting assonance rhymes and rhyme schemes. To accomplish this task, we have built a custom rhyme detector and extract respective timestamps from the audio file. The visualization integrates dynamic, time-based components to present insights from the rhyme analysis. Two linked views provide textual and temporal perspectives on a song. They can be viewed as an animation while the song plays and explored interactively afterwards. We demonstrate how our approach helps analyzing different songs covering different styles of rap.
Download

Paper Nr:	387
Title:	GLIMPSE-Med: Single-Screen Visualization of Multivariate Time Series for a Single Individual
Authors:	Hugo Le Baher, Jérôme Azé, Sandra Bringay, Pascal Poncelet, Arnaud Sallaberry and Caroline Dunoyer
Abstract:	The widespread digitization of hospital information systems is paving the way for the integration of interactive visualization methods into decision support systems. This progress enhances the ability to anticipate critical risks in monitored patients and alleviates the workload of healthcare providers. However, Electronic Health Records (EHRs) encompass large, heterogeneous, and temporal records, making it a significant challenge to develop tools that enable effective understanding trajectories embedded in these complex data. We introduce GLIMPSE-Med, an interactive timeline-based visualization interface for temporal and heterogeneous events in the EHR, incorporating a score generated by a predictive model. The evaluation of this interface, conducted with healthcare professionals, confirmed that it meets two essential needs: (1) Assess the quality of data collected in an EHR ; (2) Estimate the patient’s condition over time.
Download

Paper Nr:	394
Title:	AlertSets: Supporting Exploratory Analysis of Cybersecurity Alerts Through Set Interactions
Authors:	Franziska Becker, Christoph Müller, David Karpuk, Tanja Blascheck and Thomas Ertl
Abstract:	Security providers typically deal with large numbers of alerts based on heterogeneous data from many endpoint sensors. While the number of alerts is generally much smaller than the volume of raw data, most alerts are false positives that do not reflect genuinely malicious activity. All types of experts work on such alerts, be it to determine whether they are indeed false positives, to build machine learning models to support their analysis or to keep an eye on the current threat landscape. We conducted a design study to support a diverse group of experts whose working environments are connected to the same alert data. Based on an ongoing industry project that clusters alerts, we designed and evaluated a visual analytics system which enables exploration via powerful, easy-to-understand filtering mechanisms framed through set operations. In this article, we describe our system, give a detailed breakdown of the design process and the lessons we learned. We discuss the results from expert interviews, which showed the set-based framing to align with experts’ intuitive approach to data analysis and helped users uncover improvement opportunities for the clustering and detection pipelines.
Download

Short Papers

Paper Nr:	18
Title:	Evaluating LLMs for Visualization Tasks
Authors:	Saadiq Rauf Khan, Vinit Chandak and Sougata Mukherjea
Abstract:	Information Visualization has been utilized to gain insights from complex data. In recent times, Large Language Models (LLMs) have performed very well in many tasks. In this paper, we showcase the capabilities of different popular LLMs to generate code for visualization based on simple prompts. We also analyze the power of LLMs to understand some common visualizations by answering simple questions. Our study shows that LLMs could generate code for some visualizations as well as answer questions about them. However, LLMs also have several limitations. We believe that our insights can be used to improve both LLMs and Information Visualization systems.
Download

Paper Nr:	40
Title:	MillefioriAnalyzer: Machine Learning, Computer Vision and Visual Analytics for Provenance Research of Ancient Roman Artefacts
Authors:	Alexander Wiebel, Oliver Gloger and Hella Eckardt
Abstract:	In this position paper, we explore ways to digitally support provenance research of ancient Roman artefacts decorated with millefiori. In particular, we discuss experiments applying visual analytics, computer vision and machine learning approaches to analyze the relations between images of individual millefiori slices called florets. We start by applying automatic image analysis approaches to the florets and discover that image quality and the small overall number of images pose serious challenges to these approaches. To address these challenges, we bring human intuition and pattern recognition abilities back into the analysis loop by developing and employing visual analytics techniques. We achieve a convenient analysis workflow for the archaeologists by integrating all approaches into a single interactive software tool which we call MillefioriAnalyzer. The software is tailored to fit the needs of the archaeological application case and links the automatic image analysis approaches with the interactive visual analytics views. As appropriate for a research software, MillefioriAnalyzer is open-source and publicly available. First results include an automatic approximate ordering of florets and a visual analytics module improving upon the current manual image layout for further analytic reasoning.
Download

Paper Nr:	53
Title:	Interactive Wind Simulation in Settlement Areas
Authors:	Michael Burch, Raphael Brunold and Ralf-Peter Mundani
Abstract:	We investigate the research problem of simulating interactive wind flows in settlement areas containing buildings of arbitrary shapes. To reach this goal we generate two-dimensional wind flow simulations based on geographic data from areas in Switzerland. In modern cities it is crucial to explore wind flows that might have effects on the fresh air circulation, urban heat islands, or transport and flow directions of polluted or contaminated air. In our work, we create a pipeline to define and implement the steps and techniques to generate a wind flow simulation with which we can monitor the flow around buildings while also allowing user interactions during and after the wind flow computation. To achieve our results we focus on data accessed from public geographic information systems (GIS) in Switzerland that are available in different geo-spatial granularities. The visualizations can combine several wind flow metrics like wind directions, wind intensities and velocities, as well as air pressure, either in separate visual depictions or as overlays in geographic maps. Finally, we discuss limitations and scalability issues and provide an outlook based on future directions.
Download

Paper Nr:	54
Title:	Designing Explainable and Counterfactual-Based AI Interfaces for Operators in Process Industries
Authors:	Yanqing Zhang, Leila Methnani, Emmanuel Brorsson, Elmira Zohrevandi, Andreas Darnell and Kostiantyn Kucher
Abstract:	Industrial applications of Artificial Intelligence (AI) can be hindered by the issues of explainability and trust from end users. Human-computer interaction and eXplainable AI (XAI) concerns become imperative in such scenarios. However, the prior evidence of applying more general principles and techniques in specialized industrial scenarios is often limited. In this case study, we focus on designing interactive interfaces of XAI solutions for operators in the pulp and paper industry. The explanation techniques supported and compared include counterfactual and feature importance explanations. We applied the user-centered design methodology, including the analysis of requirements elicited from operators during site visits and interactive interface prototype evaluation eventually conducted on site with five operators. Our results indicate that the operators preferred the combination of counterfactual and feature importance explanations. The study also provides lessons learned for researchers and practitioners.
Download

Paper Nr:	62
Title:	Why Does It Look like this? Introducing a Preliminary Framework for Explainable Graph Drawing (XGD)
Authors:	Keisi Çela, Stef van den Elzen, Jarke van Wijk and Alessio Arleo
Abstract:	The discipline of Explainable Artificial Intelligence (XAI) enhances the transparency and trustworthiness of AI models by providing human-readable interpretable explanations for AI-driven decisions. The recent introduction of AI-accelerated techniques to the graph drawing community brings the challenge of comprehending the black-box ML and AI outputs when suggesting a layout for a specific graph - a problem we dub Explainable Graph Drawing (XGD). As a first step in addressing this challenge, this paper introduces a preliminary framework to match existing XAI methods to present and future AI approaches in graph drawing. This supports researchers in framing the used AI algorithm in XAI literature and helps in selecting the appropriate explanation method. We apply our approach on a chosen AI technique for graph drawing and present our findings. Finally, we discuss future perspectives and opportunities for explainable graph drawing.
Download

Paper Nr:	73
Title:	HoloGraphs: An Interactive Physicalization for Dynamic Graphs
Authors:	Daniel Pahr, Henry Ehlers and Velitchko Filipov
Abstract:	We present HoloGraphs, a novel approach for physically representing, explaining, exploring, and interacting with dynamic networks. HoloGraphs addresses the challenges of visualizing and understanding evolving network structures by providing an engaging method of interacting and exploring dynamic network structures using physicalization techniques. In contrast to traditional digital interfaces, our approach leverages tangible artifacts made from transparent materials to provide an intuitive way for people with low visualization literacy to explore network data. The process involves printing network embeddings on transparent media and assembling them to create a 3D representation of dynamic networks, maintaining spatial perception and allowing the examination of each timeslice individually. Interactivity is envisioned using optional Focus+Context layers and overlays for node trajectories and labels. Focus layers highlight nodes of interest, context layers provide an overview of the network structure, and global overlays show node trajectories over time. In this paper, we outline the design principles and implementation of HoloGraphs and present how elementary digital interactions can be mapped to physical interactions to manipulate the elements of a network and temporal dimension in an engaging matter. We demonstrate the capabilities of our concept in a case study. Using a dynamic network of character interactions from a popular book series, we showcase how it represents and supports understanding complex concepts such as dynamic networks.
Download

Paper Nr:	102
Title:	A Subspace Projection Based Technique for Visualizing Machine Learning Models
Authors:	Ziqian Bi, Raymond Gao and Shiaofen Fang
Abstract:	As Artificial Intelligence (AI) technology, particularly Machine Learning (ML) algorithms, becomes increasingly ubiquitous, our abilities to understand and interpret AI and ML algorithms become increasingly desirable. Visualization is a common tool to help users understand individual ML decision-making processes, but its use in demonstrating the global patterns and trends of a ML model has not been sufficiently explored. In this paper, we present a visualization technique using subspace projection to visualize ML models as scalar valued multi-dimensional functions to help users understand the global behaviors of the models in different 2D viewing spaces. A formal definition of the visualization problem will be given. The visualization technique is developed using an interpolation-based subspace morphing algorithm and a subspace sampling method to generate various renderings through projections and cross-sections of the model space as 3D surfaces or heatmap images. Compared to existing ML visualization methods, our work provides better global views and allows the users to select viewing spaces to provide user-specified perspectives. This method will be applied to two real-world datasets and applications: the diagnosis of Alzheimer's Disease (AD) using a human brain networks dataset and a real-world benchmark dataset for predicting home credit default risks.
Download

Paper Nr:	108
Title:	Immersive in Situ Visualizations for Monitoring Architectural-Scale Multiuser MR Experiences
Authors:	Zhongyuan Yu, Daniel Zeidler, Krishnan Chandran, Lars Engeln, Kelsang Mende and Matthew McGinity
Abstract:	Mixed reality (MR) environments provide great value in displaying 3D virtual content. Systems facilitating co-located multiuser MR (Co-MUMR) experiences allow multiple users to co-present in a shared immersive virtual environment with natural locomotion. They can be used to support a broad spectrum of applications such as immersive presentations, public exhibitions, psychological experiments, etc. However, based on our experiences in delivering Co-MUMR experiences in large architectures and our reflections, we noticed that the crucial challenge for hosts to ensure the visitors’ quality of experience is their lack of insight into the real-time information regarding visitor engagement, device performance, and system events. This work facilitates the display of such information by introducing immersive in situ visualizations.
Download

Paper Nr:	260
Title:	Penta: Towards Visualizing Compound Graphs as Set-Typed Data
Authors:	Henry Ehlers, Mario Kerndler and Renata G. Raidou
Abstract:	Compound graphs are graphs whose nodes, in addition to topological connections, share group-level relationships. The need to incorporate both topological and group-level relationships makes them inherently challenging to visualize, especially for large data. We present Penta, a prototypical dashboard that, by combining elements of compound graph and set visualization, provides a complete view of both types of relationships. To this end, we employ five linked views that provide insight into a compound graph’s i) global and set-local topology using both hypernode and traditional node-link diagrams, respectively, ii) set and entity-level relationship and identity using similarity matrices linked by a bipartite node-link diagram, as well as iii) node-centric topology across sets visualized as a layered node-link diagram. We demonstrate the workflow and advantages of Penta in three small-scale case studies, using character co-occurrence networks as well as biochemical pathway data. While still a prototype, the proposed dashboard shows promise in facilitating a complete visual exploration of the topology and group-level relationships present in compound graphs, simultaneously.
Download

Paper Nr:	369
Title:	SPViz: A DSL-Driven Approach for Software Project Visualization Tooling
Authors:	Niklas Rentz and Reinhard von Hanxleden
Abstract:	For most service architectures, such as OSGi and Spring, architecture-specific tools allow software developers and architects to visualize configurations that are usually spread through project files. Such visualization tools are used for documentation purposes and help to understand programs. However, such tools often do not address project-specific peculiarities, or do not exist at all for less common architectures. We propose a DSL-driven approach that allows software architects to define and adapt their own project visualization tool. The approach, which we refer to as Software Project Visualization (SPViz), uses two DSLs, one to describe architectural elements and their relationships, and one to describe how these should be visualized. We demonstrate how SPViz can then automatically synthesize a customized, project-specific visualization tool that can adapt to changes in the underlying project automatically. We implemented our approach in an open-source library and discuss and analyze three different tools that follow this concept, including open-source projects and projects from an industrial partner in the railway domain.
Download

Paper Nr:	399
Title:	Studying Parallel Coordinates Under Varying Aspect Ratios
Authors:	Leon Meka, Mirjam Kronsteiner and Johanna Schmidt
Abstract:	In constraint layout environments like dashboards and multi-view applications, designers have less freedom in selecting the correct aspect ratios for plots. Especially for web-based, responsive dashboards, designers have little to no control over the layout and size of the presented plots. The effect of aspect ratios on the readability of line charts and scatter plots has already been studied. However, more evidence is needed for parallel coordinates, where line slopes indicate correlations between variables. This paper presents a first step towards understanding the effect of aspect ratios on the readability of parallel coordinates. We present a statistical analysis of aspect ratio effects and summarize the results of a quantitative user study on user literacy under different aspect ratios. The statistical analysis revealed that angle parameters stay more homogeneous when changing the plot size in case landscape orientation is used. The user study showed that human observers perform well when judging correlation based on the angles under differences between plot width and height.
Download

Paper Nr:	402
Title:	ScaleVis: Interactive Exploration of Measurement Instrument Verification Data
Authors:	Haris Memić, Sanjin Radoš, Almira Softić and Krešimir Matković
Abstract:	Legal metrology ensures consumer protection from inaccurate measurements by regulating numerous instruments, some under EU harmonized legislation and others governed by national decisions based on the International Organization for Legal Metrology (OIML) recommendations. Verification laboratories produce measurement reports, often in unstructured PDF formats. Exploring and analyzing these reports remains inherently tedious and error-prone due to their format as numerous unstructured PDF files. To address these challenges, we introduce ScaleVis, a system combining standard and specialized visualizations to facilitate the exploration and analysis of measurement data including spatial information relevant to eccentricity measurements. The system incorporates data cleaning to resolve inconsistencies from manual entry and provides insights into measurement trends and deviations. Focusing on non-automatic weighing instruments, we analyze verification results to identify significant deviations in linearity and eccentricity. This study focuses on the analysis of non-automatic weighing instruments from various manufacturers and application domains. Using verification results from competent laboratories, we examine the metrological behavior of these instruments, identifying the ranges of linearity and eccentricities with the largest deviations from prescribed errors. A use case with domain experts underscores ScaleVis’s potential to streamline data analysis in legal metrology, with initial feedback indicating strong utility and effectiveness.
Download

Paper Nr:	77
Title:	Delphi: A Natural Language Interface for 2.5D Treemap Visualization of Source Code
Authors:	Adrian Jobst, Daniel Atzberger, Willy Scheibel, Jürgen Döllner and Tobias Schreck
Abstract:	Modern software development projects are characterized by large teams of developers, diverse technology stacks, and systematic workflows. This inherent complexity makes it difficult for stakeholders to maintain an overview of the project and its evolution. Software Visualization concerns generating data-driven geometric representations of specific aspects of software systems to provide insights and enable exploration. However, effective utilization of these specialized visualizations requires expertise in visualization theory and software development. This paper presents Delphi, the first system that combines a Natural Language Interface backed by a Large Language Model with a 2.5D treemap as software visualization technique. Delphi modifies the visual mapping to answer questions related to the software project, highlights objects, and provides explanations for the user. We demonstrate our system’s workflow through a use case study involving a mid-sized TypeScript project, showing how Delphi facilitates exploration. Our findings indicate that Delphi enhances the exploration process’s efficiency and broadens accessibility for a wider range of users. We release our source code as open source at https://github.com/hpicgs/llm-treemaps, with our prototype hosted on https://hpicgs.github.io/llm- treemaps.
Download

Paper Nr:	157
Title:	Active Physicalization of Temporal Bone
Authors:	Fatemeh Yazdanbakhsh and Faramarz Samavati
Abstract:	This paper explores the integration of conductivity, electricity, controllers, and 3D printing technologies to develop an active physicalization model. The model is interactive and utilizes conductive 3D filaments as sensors to trigger a feedback system when activated. The resulting interactive model can be applied to various physicalized models where the internal structure is crucial. As a case study, we have 3D printed a temporal bone model with its inner organs, using conductive material to sense the proximity of a drill around the inner organs. When a surgical drill comes into contact with these conductive-material-printed inner organs, it triggers the feedback system, producing feedback in the form of a buzzer or blinking LED. Our adaptable feedback system extends beyond surgery rehearsal, with the case study serving as a representative example.
Download

Paper Nr:	288
Title:	PyNarrative: A Python Library for Data Storytelling
Authors:	Angelica Lo Duca and Roberto Olinto Barsotti
Abstract:	Data storytelling is an emerging approach combining data visualization with narrative techniques to enhance data insights’ interpretability and emotional impact. Traditional Python libraries for data visualization, such as Matplotlib, Seaborn, and Plotly, offer powerful tools for creating static and interactive graphs. However, they lack specialized features that allow users to effectively structure and convey data-driven narratives. This paper introduces PyNarrative, an innovative Python library designed to fill this gap by integrating storytelling elements—such as annotations, context, and next steps boxes—into data visualizations. PyNarrative enables users to craft data stories that are informative but also engaging and memorable, making complex data accessible to a broader audience. This paper details the design and functionality of PyNarrative and shows a practical use case. Through PyNarrative, we aim to empower developers and data storytellers to transform raw data into meaningful narratives, advancing the field of data storytelling and contributing to more effective data communication.
Download

Paper Nr:	385
Title:	Multivariate Time Series Visualization for a Single Individual: A Scoping Review Using PRISMA-ScR
Authors:	Hugo Le Baher, Jérôme Azé, Sandra Bringay, Pascal Poncelet, Arnaud Sallaberry and Caroline Dunoyer
Abstract:	The digitization of hospital information systems is becoming widespread, enabling the increasing integration of interactive visualization methods into decision support systems. This development facilitates the anticipation of critical risks in monitored patients and helps reduce the workload of healthcare providers. However, Electronic Health Records (EHRs) contain large, heterogeneous, and temporal data. Then, providing tools to understand these complex data is a challenge. Using PubMed and Google Scholar, we conducted a search for articles using keywords related to time, visualization, and data. Out of 3,197 retrieved articles, we identified 111 relevant ones through clustering. Applying exclusion criteria to focus on implemented prototypes, we manually annotated 21 articles for our review. This exploratory literature analysis reveals that while this research area has garnered recent interest, it demonstrates limitations in the proposed solutions. Few approaches employ temporal axis distortion, and no approach in the medical domain visually integrates model predictions. The study highlights preferred functionalities for the visual representation of multivariate temporal data, such as parallel time series and hierarchical views.
Download

Area 2 - Scientific Visualization

Full Papers

Paper Nr:	279
Title:	Investigating the Propagation of CT Acquisition Artifacts along the Medical Imaging Pipeline
Authors:	Jakob Peischl and Renata G. Raidou
Abstract:	We propose a framework to support the simulation, exploration, and analysis of uncertainty propagation in the medical imaging pipeline—exemplified with artifacts arising during CT acquisition. Uncertainty in the acquired data can affect multiple subsequent stages of the medical imaging pipeline, as artifacts propagate and accumulate along the latter, influencing the diagnostic power of CT and potentially introducing biases in eventual decision-making processes. We designed and developed an interactive visual analytics framework that simulates real-world CT artifacts using mathematical models, and empowers users to manipulate parameters and observe their effects on segmentation outcomes. By extracting radiomics features from artifact-affected segmented images and analyzing them using dimensionality reduction, we uncover distinct patterns related to individual artifacts or combinations thereof. We demonstrate our proposed framework on use cases simulating the effects of individual and combined artifacts on segmentation outcomes. Our application supports the effective and flexible exploration and analysis of the impact of uncertainties on the outcomes of the medical imaging pipeline. Initial insights into the nature and patterns of the simulated artifacts could also be derived.
Download

Short Papers

Paper Nr:	21
Title:	Pulsating Uncertainties: Visualization and Highlighting of Uncertainty in 3D Data Using Animated 2D Transfer Functions
Authors:	Viktor Leonhardt, Tobias Neeb, Christoph Garth and Alexander Wiebel
Abstract:	While data with uncertainties arises in many scientific domains and engineering applications, the visualization of such data remains challenging as uncertainty information must be included in an accessible and compre-hensible manner.In this paper we present pulsating uncertainties as a novel way to highlight uncertainties by animated two-dimensional transfer functions (2DTF) for uncertain scalar data sets. It allows for a flexible classification by 2DTFs and an effective and pre-attentive highlighting of uncertainty by animating the 2DTFs while enabling users to simultaneously explore the 3D scene.In addition, we present the isosurface variability widget to highlight the variability of isosurfaces for data with uncertainty.We demonstrate the characteristics of the new approach by experiments using climate simulation and medical data.
Download

Paper Nr:	101
Title:	LIC-R: Line Integral Convolution Revisited
Authors:	Khatereh Mohammadi, Marco Agus, Ahmad Abushaikha and Jens Schneider
Abstract:	We present a novel formulation of Line Integral Convolution (LIC), a fundamental method for visualizing vector fields in flow visualization. Our approach reinterprets the traditional LIC technique by leveraging a regularized, directional curvature flow along streamlines, utilizing material derivatives to achieve the desired convolution. By adopting an entirely Eulerian framework, our method eliminates the need for complex numerical integration and high-order interpolation schemes that are typically required in classical LIC algorithms. This shift not only simplifies the implementation of LIC, making it more accessible for both CPU and GPU architectures, but also significantly reduces the computational overhead. Despite these simplifications, our method maintains visual quality comparable to that of more traditional and computationally expensive approaches. Moreover, the discrete nature of our formulation makes it particularly well-suited for irregular grids and sparse data, broadening its applicability in practical settings. Through various experiments, we demonstrate that our algorithm delivers efficient and visually coherent results, offering an attractive alternative for dense flow visualization with reduced complexity.
Download

Area 3 - Visualization Techniques

Full Papers

Paper Nr:	109
Title:	Inconspicuous Guides: A Qualitative Study to Understand the Degree of Intrusiveness in Visualization Onboarding
Authors:	Margit Pohl, Benjamin Potzmann, Christina Stoiber and Wolfgang Aigner
Abstract:	Current visualizations are becoming more and more complex. Visualization onboarding is a possibility to assist users in understanding such visualizations. Nevertheless, previous research indicates that users tend to ignore this kind of assistance, even if it is evident that they need help to interact with the system efficiently. This investigation aims to collect prototypes of visualization onboarding systems taking into account the degree of intrusiveness through a user inquiry using sketching and a questionnaire. We conducted a study with 65 participants. We asked them to what extent intrusion on the part of the system (e.g., compulsory tutorials, automatic pop-up messages) would be acceptable. In addition, we asked them to create solutions to this problem. Most found less intrusive visualization onboarding forms (e.g., tooltips, optional and concise tutorials) helpful. Compulsory visualization onboarding forms were considered less helpful. Participants also suggested chatbots or ML algorithms as convenient solutions for visualization onboarding.
Download

Paper Nr:	230
Title:	Interactive Platform for Surveys in Information Visualisation
Authors:	Barbara Nascimento, João M. Cunha and Evgheni Polisciuc
Abstract:	Conventional literature review platforms typically present data in static formats, restricting users from fully exploring complex, multivariate information and hindering their ability to stay up-to-date with evolving research. A way to address this involves the development of interactive platforms that enable users to dynamically explore data and the addition of new information. In this paper, we analyse existing platforms, identifying key features that improve data visualisation and interaction. Our analysis is used to support the proposal of a novel interactive survey platform aimed at improving the analysis and exploration of bibliographic data using various visualisation and interactivity techniques. The proposed platform integrates data representation approaches, such as grids, distribution charts, and tree diagram, and interaction techniques like filtering, selecting, and reconfiguring data layouts. To assess the usability of the platform, we applied it to a specific scenario – a survey on data glyphs – and conducted a study with users. The results of the user study indicate that the platform’s functionalities, which include various representation approaches and interactive techniques, enable deep exploration and effectively assist in conducting data analysis tasks. The platform is a promising step towards creating more dynamic and interactive survey tools and it serves as a starting point for further development.
Download

Paper Nr:	257
Title:	SnakeTrees: A Visualization Solution for Discovery and Exploration of Audiovisual Features
Authors:	Xiao Tan, Ünsal Satan, Jonas Zellweger, Gaudenz Halter, Barbara Flückiger, Renato Pajarola and Alexandra Diehl
Abstract:	Digital archives, especially audiovisual archives, often contain a large number of features of interest to digital humanities scholars, including video, audio, metadata, and annotation data. These large and complex datasets pose numerous challenges, such as how to get an overview of the overall data structure, how to identify associations between relevant data features, and how to formulate hypotheses based on observations or elicit new conceptualizations. To address these challenges, we propose a visualization tool SnakeTrees that allows digital humanities scholars to explore audiovisual archives in a novel interactive way based on computational grouping and similarity analysis provided by dimensionality reduction methods and clustering techniques. The main goal of visualizing and exploring these abstract representations is to encourage the finding of new concepts, discover new unexpected connections between different audiovisual elements, and engage users in exploratory analysis. Our approach uses interactive visualization and computational hierarchical structures to provide pre-configured groupings and categorizations that users can use as a basis for exploration and analysis.
Download

Short Papers

Paper Nr:	50
Title:	Empowering Facility Managers: Multimodal Location-Based Visualization for Smart Sanitizer Dispensers Management
Authors:	Victor-Adriel De-Jesus-Oliveira, Thomas Bigler, Florian Grassinger, Michael Zauchinger, Albert Treytl and Wolfgang Aigner
Abstract:	Facility management employees are in charge of maintaining a safe and operational infrastructure. Even a seemingly simple task, such as maintaining disinfectant dispensers that need to be operational and constantly refilled, can become costly as the number of units and the distance between them increases. In this context, the deployment of wireless sensor networks allows not only scaling data collection but also for optimizing otherwise mundane tasks and improving response times. Location-based and situated data representations are then presented as powerful tools to visualize and assess data in relation to its physical referent. In this paper, we report a proof-of-concept study on the implementation of sensor-powered disinfectant dispensers deployed on a university campus. We present the design of different location-based visualization dashboards to explore and monitor dispenser status including a traditional web dashboard with list and map views, as well as an augmented reality (AR) app displaying embedded data representations. Interviews with facility management employees are conducted to review their impressions, discuss how such dashboards could be incorporated into their workflow, and explore potential improvements to better support their work.
Download

Paper Nr:	58
Title:	Reflections on the Uses and Available Choices of Categorical Colorschemes
Authors:	Sara Di Bartolomeo, Raphael Buchmüller, Alexander Frings, Johannes Fuchs and Daniel Keim
Abstract:	Categorical colorschemes must respect a number of criteria — mainly, they need to incorporate a number of easily distinguishable colors, and they need to avoid giving to the reader the impression that the colors in the visualization have particular relationships. Crafting these palettes requires careful attention to the distribution of colors; thus, for a long time, visualization designers have been relying on a limited choice of readily available palettes. Although such palettes have been proven practical and functional, our own experience with designing visualizations had us struggle repeatedly with the limited choice, the feeling of repetitiveness in seeing the same colors in visualization papers, and a number of other limitations that we discuss in the paper. In this document, we discuss some properties of the most common categorical colorschemes, and propose a method to generate new palettes that are comparable in properties to the existing ones.
Download

Paper Nr:	301
Title:	Posterizing Diagrams with Top-Down Layout
Authors:	Maximilian Kasperowski and Reinhard von Hanxleden
Abstract:	Diagrams serve as a communication tool and automatic graph drawing enables the creation of diagramming tools that allow seamless graph modeling and visualization. These tools typically have the goal of optimizing the produced diagrams for computer screens, which support zooming in and out of diagrams and further interaction techniques. In this paper, we consider what automatic graph drawing needs to provide to produce diagrams suitable for printing instead, where we can no longer zoom and are restricted to a fixed area. We identify key differences in the requirements between printed static diagrams and interactive on-screen diagrams, including layout speed and layout area. Our goal is to automatically draw a complex diagram to be printed on a large poster, with maximal readability. We propose a collection of refinements to the well-established Sugiyama algorithm and illustrate our approach with a large SCChart that implements a railway controller.
Download

Paper Nr:	322
Title:	Generative Artificial Intelligence for Immersive Analytics
Authors:	Chaoming Wang, Veronica Sundstedt and Valeria Garro
Abstract:	Generative artificial intelligence (GenAI) models have advanced various applications with their ability to generate diverse forms of information, including text, images, audio, video, and 3D models. In visual computing, their primary applications have focused on creating graphic content and enabling data visualization on traditional desktop interfaces, which help automate visual analytics (VA) processes. With the rise of affordable immersive technologies, such as virtual reality (VR), augmented reality (AR), and mixed reality (MR), immersive analytics (IA) has been an emerging field offering unique opportunities for deeper engagement and understanding of complex data in immersive environments (IEs). However, IA system development remains resource-intensive and requires significant expertise, while integrating GenAI capabilities into IA is still under early exploration. Therefore, based on an analysis of recent publications in these fields, this position paper investigates how GenAI can support future IA systems for more effective data exploration with immersive experiences. Specifically, we discuss potential directions and key issues concerning future GenAI-supported IA applications.
Download

Paper Nr:	328
Title:	Colorimetric Compensation in Video Mapping for Luggage Inspection
Authors:	Maëlan Poyer and Christophe Hurter
Abstract:	Airport luggage inspection agents work under time pressure to identify and localize dangerous items using three-dimensional scans. Video mapping can significantly enhance speed and efficiency by projecting real-time information helping to localize threats to be removed. However, maintaining color fidelity is crucial, as accurate color representation provides key information for decision-making. Previous research has explored color correction techniques for complex surfaces, but these often require extensive calibration, limiting their real-time applicability. Our approach addresses this limitation by using a pre-recorded database to maintain color compensation without the need for frequent recalibration. We built this colorimetric database that records how surfaces with similar textures reflect colors. Using Shepard’s interpolation, our algorithm generalizes the color correction to new surfaces with similar textures, allowing for real-time adjustments without interrupting workflow. This paper aims to lay the foundation for large-scale studies. The results show good performance for hues such as orange but the method’s effectiveness varies across the color spectrum, with limited improvements on blue hues due to predictable losses in luminance and saturation. This highlights the need for new techniques to overcome the physical limitations of projectors.
Download

Paper Nr:	351
Title:	C4D3: A View-Level Abstraction and Coordination Library for Building Coordinated Multiple Views with D3
Authors:	Omkar S. Chekuri and Chris Weaver
Abstract:	D3 is a popular and effective library for the development and deployment of visualizations in web pages. Numerous applications testify to its accessibility and expressiveness for representing and manipulating page content. By eschewing toolkit-specific abstraction, D3 gains representational transparency, but at a cost of modularity. Relatively few D3 applications compose multiple views or support coordinated interaction beyond basic navigation and brushing. Rather than relegate view composition to custom integration code, we overlay D3 with a view-level abstraction that utilizes a general parameter sharing model to offer simple yet flexible composition of coordinated multiple views (CMV) while preserving the expressiveness of individual D3 components. Coupling of event handling to shared parameters recasts modeling of interactive state and simplifies the declarative specification of interactive dependencies between views. We present an example of an extensively coordinated visualization, illustrative code to show CMV construction in C4D3, and demonstrate how view-level abstraction can reduce the code needed to compose complex D3 visualizations.
Download

Paper Nr:	393
Title:	Visual Analytics for the Analysis of Sleep Quality
Authors:	Maria Tsiobra, Georgios Nikolis, Christos Diamantakis, Matthew Salanitro, Ilias Kalamaras, Vasilis Lwlis, Thomas Penzel, Konstantinos Votis and Dimitrios Tzovaras
Abstract:	Monitoring the quality of sleep in patients of sleep disorders is often a time-consuming process, where the clinician manually navigates through large volumes of recorded polysomnography data in an effort to visually detect sleep patterns, such as sleep spindles, sleep stages and hints of disorders. We propose an application that provides healthcare professionals with advanced tools for sleep analysis and spindle detection through visual analytics for pattern detection, AI-based sleep scoring, and an interactive user interface. The system processes multiple physiological signals and provides both raw data visualization, advanced feature analysis capabilities, and a two-dimensional embedding of sleep intervals. By combining signal processing, spindle detection, sleep stage identification and interactive visualization tools, this work helps researchers to efficiently identify, validate, and analyze sleep and spindle characteristics with higher precision than traditional methods.
Download