EuroVisShort2017

Permanent URI for this collection

https://diglib.eg.org/handle/10.2312/2631438

Browse

Now showing 1 - 20 of 31

Reflections on an Experiment, Evaluating the Impact of Spatialisation on Exploration
(The Eurographics Association, 2017) Roux, Clement; McAuley, John; Barbora Kozlikova and Tobias Schreck and Thomas Wischgoll
This paper reports on an experiment designed to evaluate whether visualising a digital library (using a spatialisation technique) can influence exploratory search behaviour. In the experiment we asked participants to complete a set of novel tasks using one of two interfaces - a visualisation interface, ExploViz, and its search-based equivalent, LibSearch. A set of measures were used to capture sensemaking and exploratory behaviour and to analyse cognitive load. As results were non-significant, we reflect upon the design of the experiment, consider possible issues and suggest how these could be addressed in future iterations.
PubViz: Lightweight Visual Presentation of Publication Data
(The Eurographics Association, 2017) Rind, Alexander; Haberson, Andrea; Blumenstein, Kerstin; Niederer, Christina; Wagner, Markus; Aigner, Wolfgang; Barbora Kozlikova and Tobias Schreck and Thomas Wischgoll
Publications play a central role in presenting the outcome of scientific research but are typically presented as textual lists, whereas related work in visualization of publication focuses on exploration - not presentation. To bridge this gap, we conducted a design study of an interactive visual representation of publication data in a BibTeX file. This paper reports our domain and problem characterization as well as our visualization design decisions in light of our user-centered design process including interviews, two user studies with a paper prototype and a d3.js prototype, and practical application at our group's website.
Visualizing Large Time-series Data on Very Small Screens
(The Eurographics Association, 2017) Chen, Yang; Barbora Kozlikova and Tobias Schreck and Thomas Wischgoll
In this paper, we present a space-efficient visualization and an interaction technique for exploring time-series data on very small screens. The visualization is based on a simultaneous display of contextual visualization in the display border and focused interfaces in the center region of displays. The interaction technique utilizes fisheye distortion to facilitate data selection and navigation interactions in the contextual visualization. The proposed techniques could benefit a wide range of analysis applications on wearable devices. Two applications are highlighted to demonstrate the usefulness of the techniques.
TexTile: A Pixel-Based Focus+Context Tool For Analyzing Variants Across Multiple Text Scales
(The Eurographics Association, 2017) Asokarajan, Bharathi; Etemadpour, Ronak; Abbas, June; Huskey, Sam; Weaver, Chris; Barbora Kozlikova and Tobias Schreck and Thomas Wischgoll
Before the advent of printed texts, text duplication was done primarily by hand. Errors, alterations, and erasures were common and varied widely across different copies of the same text. Classics scholars seek to reconstruct an ''original'' text by analyzing and merging variations across copies as ''witnesses'' to a conjectured original. Many scholars continue to use spreadsheets, sometimes as large sheets of actual paper, to visually collate variations across known versions. These approaches are generally well suited for collection of data about variations, a process that can take decades. However, they are poorly suited for analysis of variation above the level of individual words. Visualization techniques are needed to reveal patterns of variation at the level of lines, pages, and entire texts. We present TexTile, a new tool that integrates pixel-based and focus+context visualization techniques for analysis of reconstructed classical Latin texts. TexTile provides a comprehensive yet compact representation of variation at multiple levels over an entire text. The tool helps scholars validate the accuracy of textual variants and analyze similarities between different contributing copies. The integrated visualization design allows exploration of variation across textual scales while preserving continuity of browsing, much like when examining a physical manuscript. We conducted a mixed quantitative-qualitative user study to assess the usability of the integrated design.
Marching Pentatopes for Continuous Morphing of Isosurfaces From Four Dimensional Data in HTML5/WebGL
(The Eurographics Association, 2017) Watters, Aaron R.; Barbora Kozlikova and Tobias Schreck and Thomas Wischgoll
Animations which show three dimensional volumes continuously changing over time facilitate the exploration and analysis of complex data sets such as calcium image data of neural activity and phase contrast magnetic resonance imaging of blood flows. This paper explains the marching pentatopes method for representing the iso-surfaces of a four dimensional data set as a triangulated surface smoothly deforming as time progresses. The morphing triangulations generated by the this method may be rendered using the morph geometry capabilities provided by the three.js javascript library for cross platform HTML5/WebGL presentation in standard web browsers [Cab17].
TExVis: An Interactive Visual Tool to Explore Twitter Data
(The Eurographics Association, 2017) Humayoun, Shah Rukh; Ardalan, Saman; AlTarawneh, Ragaad; Ebert, Achim; Barbora Kozlikova and Tobias Schreck and Thomas Wischgoll
Exploring tweets enables us understanding people's reaction and feedback regarding any particular event or product. Many tools have been developed to visualize Twitter data based on some criteria, e.g., keyword frequency or evolution of topics. Visualizing the relations between the keywords of the underlying Twitter data opens another window to analyze the people's reaction towards a particular event/product. Targeting this concern, our developed tool, called TExVis (Tweets Explorer and Visualizer), visualizes important keywords (e.g., hashtags, nouns, verbs) from the underlying tweets based on their frequency and shows the relations between them based on some criteria (e.g., the common tweets), using an extended Chord diagram. TExVis also visualizes the sentimental polarity for a better understanding of the keywords associated tweets. Further, the provided interaction, multi-level navigation, and filtering options help the users in better exploration of the underlying tweets. A user study with 16 participants shows a high acceptance towards the tool and our approach in general.
On Establishing Visualization Requirements: A Case Study in Product Costing
(The Eurographics Association, 2017) Vosough, Zana; Groh, Rainer; Schulz, Hans-Jörg; Barbora Kozlikova and Tobias Schreck and Thomas Wischgoll
The process of identifying visualization requirements is an important part of every visualization researcher's and practitioner's job. Nevertheless, the scientific literature is rather sparse on this topic, usually resorting to some form of user-centered design that is rarely further detailed. In this paper, we give an account of our procedure, our results, our problems and solutions for gathering visualization requirements in an ongoing business project to introduce visualization to the field of product costing. By providing insight in our experiences and extracting general points of advice from them, we aim to give some practical guidance for establishing requirements in real-world visualization projects.
Natural Language Interfaces for Data Analysis with Visualization: Considering What Has and Could Be Asked
(The Eurographics Association, 2017) Srinivasan, Arjun; Stasko, John T.; Barbora Kozlikova and Tobias Schreck and Thomas Wischgoll
Natural language is emerging as a promising interaction paradigm for data analysis with visualization. Designing and implementing Natural Language Interfaces (NLIs) is a challenging task, however. In addition to being able to process and understand natural language expressions, NLIs for data visuailzation must consider other factors including input modalities, providing input affordances, and explaining system results, among others. In this article, we examine existing NLIs for data analysis with visualization, and compare and contrast them based on the tasks they allow people to perform. We discuss open research opportunities and themes for emerging NLIs in the visualization community. We also provide examples from the existing literature in the broader HCI community that may help explore some of the highlighted themes for future work. Our goal is to assist readers to understand the subtleties and challenges in designing NLIs and encourage the community to think further about NLIs for data analysis with visualization.
Trajectory Mapper: Interactive Widgets and Artist-Designed Encodings for Visualizing Multivariate Trajectory Data
(The Eurographics Association, 2017) Lange, Devin; Samsel, Francesca; Karamouzas, Ioannis; Guy, S. J.; Dockter, Rodney; Kowalewski, Timothy; Keefe, Daniel F.; Barbora Kozlikova and Tobias Schreck and Thomas Wischgoll
We present TrajectoryMapper, a system of novel interactive widgets and artist-designed visual encodings to support exploratory multivariate visualization of spatial trajectories. Trajectories are rendered using a three-way multi-texturing algorithm so that the color, texture, and shape of each mark can be manipulated separately in response to data. Visual encodings designed by artists and arranged in categories (e.g., divergent, linear, structured) are utilized as strong starting points for visual exploration. Interactive widgets including linked parallel coordinates plots, 3D camera controls, and projection to arbitrary 3D planes facilitate data exploration. An innovative visual mapper menu enables rapid experimentation with alternative data mappings using the artist-designed or custom encodings that can be created with no programming using image editing software. In addition to system design details and insights, two applications with collaborating domain science users are presented. The first requires analyzing 2D crowd simulations and the second 3D tool traces from laparoscopic surgery training exercises.
ETK: An Evaluation Toolkit for Visualization User Studies
(The Eurographics Association, 2017) Turton, Terece L.; Berres, Anne S.; Rogers, David H.; Ahrens, James; Barbora Kozlikova and Tobias Schreck and Thomas Wischgoll
This paper describes the design and features of the Evaluation Toolkit (ETK), a set of JavaScript/HTML/CSS modules leveraging the Qualtrics JavaScript API that can be used to automate image-based perceptual user evaluation studies. Automating the presentation of the images can greatly decrease the time to build and implement an evaluation study while minimizing the length and complexity of a study built within Qualtrics, along with decreasing the possibility of error in image presentation. The ETK modules each focus on automating a specific psychophysical or experimental approach. Because each module is an extension or plug-in to a Qualtrics question, the resultant study can be easily used in a laboratory setting or in a crowdsourced approach. We present the open source repository of ETK with the six modules that currently make up the toolkit and invite the community to explore, utilize, and contribute to the toolkit.
MVN-Reduce: Dimensionality Reduction for the Visual Analysis of Multivariate Networks
(The Eurographics Association, 2017) Martins, Rafael M.; Kruiger, J. F.; Minghim, Rosane; Telea, Alexandru C.; Kerren, Andreas; Barbora Kozlikova and Tobias Schreck and Thomas Wischgoll
The analysis of Multivariate Networks (MVNs) can be approached from two different perspectives: a multidimensional one, consisting of the nodes and their multiple attributes, or a relational one, consisting of the network's topology of edges. In order to be comprehensive, a visual representation of an MVN must be able to accommodate both. In this paper, we propose a novel approach for the visualization of MVNs that works by combining these two perspectives into a single unified model, which is used as input to a dimensionality reduction method. The resulting 2D embedding takes into consideration both attribute- and edge-based similarities, with a user-controlled trade-off. We demonstrate our approach by exploring two real-world data sets: a co-authorship network and an open-source software development project. The results point out that our method is able to bring forward features of MVNs that could not be easily perceived from the investigation of the individual perspectives only.
Readability and Precision in Pictorial Bar Charts
(The Eurographics Association, 2017) Skau, Drew; Kosara, Robert; Barbora Kozlikova and Tobias Schreck and Thomas Wischgoll
Bar charts embellished with unique artistic styles, or made to look like real objects, are common in information graphics. Embellishments are typically considered detrimental to readability and accuracy, since they add clutter and noise. Previous work has found that some of the shapes used, like rounded tops, triangles, etc., decreased accuracy when judging relative and absolute sizes, while T-shaped bars even showed a slight increase relative to the basic bar chart. In this paper, we report on a study that adds pictorial elements to bar charts of four different shapes tested previously, thus also including the elements of color and texture. We find that pictorial bar charts reduce accuracy, but not beyond the effect already observed for their shape. They also do not significantly increase response time. Embellished bar charts may not be as problematic as commonly assumed.
An Argument Structure for Data Stories
(The Eurographics Association, 2017) Kosara, Robert; Barbora Kozlikova and Tobias Schreck and Thomas Wischgoll
Many data stories in journalism do not have a story arc, but rather present facts without much structure. This mirrors the popular inverted pyramid style of writing that presents the most important information up front, to be followed by evidence. We have found a subset of stories that follow a more structured approach, however. These stories begin with a claim or question, but do not immediately present that as the conclusion. Instead, they then present pieces of evidence that are only tied together, and back to the initial claim, at the end. In this paper, we formalize and discuss this structure, and present a few examples. We believe that this is a viable and practical model for data stories more generally, and certainly a stronger arc than most existing stories today.
Nodes, Paths, and Edges: Using Mental Maps to Augment Crime Data Analysis in Urban Spaces
(The Eurographics Association, 2017) Godwin, Alex; Stasko, John T.; Barbora Kozlikova and Tobias Schreck and Thomas Wischgoll
Citizen participation can provide valuable insight into data that is not captured by official sources. In this paper, we propose a technique for using mental maps consisting of three fundamental elements: nodes, paths, and edges. These elements can be used to augment crime data analysis in urban spaces by incorporating the values and knowledge of citizens. We apply this technique to an analysis of property crime in three US cities: Baltimore, Atlanta, and Chicago. Subsequently, we find these cities have neighborhoods where the crime could be substantially higher-or perceived by citizens as higher-than is accounted for in the official public crime data. This analysis can be a vital first step for identifying hidden hotspots or better understanding public perceptions of high crime.
A Stream Ribbon Seeding Strategy
(The Eurographics Association, 2017) Rees, Dylan; Laramee, Robert S.; Nguyen, Duong; Zhang, Lei; Chen, Guoning; Yeh, Harry; Zhang, Eugene; Barbora Kozlikova and Tobias Schreck and Thomas Wischgoll
Streamline seeding algorithms have a long and rich history dating back over two decades. And in recent years algorithms for stream surface placement have been developed. However, stream ribbons have been generally overlooked. We present, to our knowledge, the first stream ribbon seeding strategy. Stream ribbons are a tool for visualizing vector fields and are a common extension of streamlines with the added benefit of conveying a fluid's twisting motion along the direction of flow. Presented in this short paper is a novel strategy for seeding stream ribbons in vector fields. The strategy exploits the flow's local helicity, an important property of flow identified over 40 years ago, to guide ribbon seeding. Seed points are prioritised based on a derived helicity field. A selection of user options including adjusting ribbon width, separating distance, and ribbon length filtering are applied to support visualization and cater to the users interests. A filtering method is also presented whereby the number of stream ribbons can be reduced in order to highlight the most helical flow features. We demonstrate the technique on various flow fields and report feedback from a domain expert in fluid mechanics.
Molecular Visualization of Computational Biology Data: A Survey of Surveys
(The Eurographics Association, 2017) Alharbi, Naif; Alharbi, Mohammad; Martinez, Xavier; Krone, Michael; Rose, Alexander S.; Baaden, Marc; Laramee, Robert S.; Chavent, Matthieu; Barbora Kozlikova and Tobias Schreck and Thomas Wischgoll
Visualizations for computational biology have been developing for over 50 years. With recent advances in both computational biology and computer graphics techniques, these fields have witnessed rapid technological advances in the last decade. Thus, coping with the large number of scientific articles from both fields is a challenging task. Furthermore, there remains a gap between the two communities of visualization and computational biology, resulting in additional challenges to bridge the divide. A team of computational biology and visualization scientists attempts to address these challenges by presenting unified state-of-the-art reviews from both communities. We apply a variety of data-driven analysis to highlight links or differences between studies from both communities. This approach facilitates the identification of present and future challenges in visualizing and analyzing computational biology data. It offers a distinctive step forward in managing the literature on visualization of molecular dynamics and related simulation approaches.
Spatial Interaction for the Post-Processing of 3D CFD Datasets
(The Eurographics Association, 2017) Niebling, Florian; Barbora Kozlikova and Tobias Schreck and Thomas Wischgoll
Virtual Reality visualizations are often used for the evaluation of three-dimensional datasets generated by Computational Fluid Dynamic (CFD) simulations. Several different tracked devices, such as pointers or data gloves, have been employed for spatial interactive post-processing and navigation. However, users utilizing these devices often achieve inaccurate results performing interactive slicing of datasets, a common technique during the assessment of CFD datasets. In our approach, we propose spatial interaction using tablets, which more closely resemble cutting planes, for planar surface extraction in VR environments. In contrast to traditional tracked devices, inside-out tracking of the environment can be achieved using the rear camera of the tablet, helping to make expensive external tracking systems expendable. A user study among novice and expert VR users supports the notion that tablets can be a fast and accurate alternative to traditional spatial input devices in VR environments.
Scale-Adaptive Placement of Hierarchical Map Labels
(The Eurographics Association, 2017) Wu, Hsiang-Yun; Takahashi, Shigeo; Poon, Sheung-Hung; Arikawa, Masatoshi; Barbora Kozlikova and Tobias Schreck and Thomas Wischgoll
Nowadays, digital map services provide a large amount of spatial data and thus facilitate users to dynamically navigate map contents across multiple scales on small mobile devices. In this context, consistently placing map labels in interactive navigation is important but still technically challenging, especially when the labels are associated with multiple layers, which are inherent in map contents. In this paper, we introduce a genetic-based approach to optimize the placement of annotation labels with different ranges of map scales by maximizing label visibility of the existing scale while avoiding unwanted mutual overlaps and sudden popping effects. This is accomplished by grouping the label IDs into multiple chromosomes according to their importance and then forming composite chromosomes, each of which is reordered to optimize the overall visibility of the labels. Our formulation also allows the individual labels to move across the scale adaptively in order to further improve label placement on the respective scales. We show several experimental results to present the effectiveness of the proposed approach.
The Situation Universe: Visualizing the Semantics of Integrated Data Structures
(The Eurographics Association, 2017) Karer, Benjamin; Fernández-Prieto, Diana; Hagen, Hans; Barbora Kozlikova and Tobias Schreck and Thomas Wischgoll
The efficient extraction and communication of information in heterogeneous data domains is a major challenge in the context of the ongoing digitalization efforts in industry and in the public sector. The heterogeneity of the data itself and the diverse interests of the users addressing it demand the integration of structural and semantic information about data aggregated from multiple sources into a single model and unified visualization. In this paper, we present an approach to visualize the possible interpretations of data integrated from heterogeneous environments, including the sequences of operations applied to filter, transform, and reinterpret the data, such that the result supports these interpretations. Users can thereby access and explore integrated data from the perspective specific to their respective fields of experience.
Dynamic Label Placement for Forensic Case Visualization
(The Eurographics Association, 2017) Sirk, Christof; Kalkofen, Denis; Schmalstieg, Dieter; Bornik, Alexander; Barbora Kozlikova and Tobias Schreck and Thomas Wischgoll
Forensic case analysis and in court presentation requires comprehensible illustrations and animations of findings and their relations to the course of events. Often this can only be achieved by adding textual descriptions. From a systems point of view, this requires automated label placement functionality for scenes composed of translucent polyhedral models and volumes, which we achieve through tight integration of the automated label placement algorithm and the hybrid volume/surface rendering system. Our method exploits transparency in order to place labels close to their anchors, either inside the scene, on-top or outside the occupied screen region. Inside placement makes it possible to zoom into the dataset, leads to more temporal coherency, and improves layout quality, especially for large numbers of labels. New measures for scene content importance and label occlusion prevent masking of important scene details by labels and vice versa.

Browse

Browsing EuroVisShort2017 by Issue Date

Results Per Page

Sort Options