Sorry, this content is only available in English. For your convenience, it is shown below in this language.

With the increasing diversity as well as sheer number of available devices, the potential of using them in combination in device ecologies is big – particularly for facilitating more flexible and more natural visual data analyses.

The Thesis in a Nutshell

Device ecologies can enable more flexible workflows for visual data analysis by incorporating the devices that are most suited for the current situation and handle the overall interface in a more natural way. Further, device ecologies can be available and set up anywhere, anytime. The work presented in this thesis contributes to different aspects relevant for supporting visual data analysis in device ecologies: Responsive visualization design can provide more flexible data representations that can be used on a wide range of devices and contexts. Incorporating device roles for providing interaction and visualization concepts allows to better support specific analysis workflows within a used device ensemble. And, novel interaction concepts in combination with automated optimizations allow to handle and simplify the dynamic aspects of device ecologies for analysts.

Quick Access

  1. Video Wall
  2. Main Contributions
  3. Further Publications
  4. Related Student Theses

Defense

The successful defense of the PhD thesis took take place on July 7, 2021.

The Thesis Document

  • Visual Data Analysis in Device Ecologies

    Visual Data Analysis in Device Ecologies

    Horak, T.

    Qucosa - Technische Universität Dresden,2021.

    @phdthesis{fixme,
       author = {Tom Horak},
       title = {Visual Data Analysis in Device Ecologies},
       year = {2021},
       month = {7},
       location = {Dresden},
       numpages = {283},
       url = {https://nbn-resolving.org/urn:nbn:de:bsz:14-qucosa2-758986},
       publisher = {Qucosa - Technische Universit\"{a}t Dresden}
    }

    List of additional material

    Dissertation

The thesis document is openly accesible at QUCOSA.

Fast-Forward: Video Wall

Main Contributions

Adapting Visualizations for Heterogeneous Devices

Before placing visualizations on different devices within device ecologies, it must be considered how visualizations can be adapted for all the different devices in general. In order to inform an improved understanding, my dissertation project included an analytical exploration of the notion of Responsive Visualization, which is yet underexplored in the literature. Following these generalized considerations, the aspects of responsiveness have then been explored specifically for multivariate graph visualization. Here, as the major contribution, the Responsive Matrix Cells was introduced, which is a focus+context approach for matrix visualization that allows embedding responsive detail visualizations in local focus areas. In sum, this part of my dissertation project provides an improved understanding of how visualizations can adapt to different contexts and support data analysis within a compact space.

Analysis Workflows in Dedicated Device Ensembles

With the strategies for bringing visualizations to heterogeneous devices in mind, the focus can now be shifted to using specific device ensembles for data analysis. For this, I considered two specific ensembles. In the first, the novel display combination of smartwatch and display-equipped watchstraps and its usage for personal visualization was proposed. With this self-contained ensemble called Watch+Strap, concepts for using the different displays in combination were developed and tested. For the second ensemble, it was investigated how the contrary device types of smartwatch and large display—like David and Goliath—can be used in a complementing synergy for visual data analysis. Following the design of a conceptual framework, this device setup was studied more extensively with a focus on which workflows analysts adopt while working in this environment. Both explorations provide novel concepts for mobile visualization and visualization beyond the desktop, as well as a better understanding of how device roles can aid and support common exploration workflows when investigating multivariate data sets in such ensembles.


Analysis Interfaces for Dynamic Device Ecologies

Building upon the considerations from the previous two parts, I have then investigated how visual data analysis can be supported in fully dynamic device ecologies. For that, I worked on an interaction design that allows for applying device coordination and view combinations in an ad-hoc fashion. Specifically, the conceptual framework called VisTiles focuses on exploiting changing physical device arrangements to resemble similar device roles as explored in the previous part of the dissertation topic. The flexibility of dynamic device ecologies also means that the present device setup can change often and that the interface has to be adapted accordingly. Here, I worked on the Vistribute framework, in which multiple heuristics are proposed that can support an automatic distribution of visualization views across available devices by considering
both view specifications and device properties. The accompanying implementation showed that the quality of such automatic distributions is rated by users similar to manually created ones, while notably reducing the setup effort. In sum, the works presented in this part provide concepts and approaches that can allow to actually enable visual data analysis in dynamic device ecologies and support a more natural way of conducting these analyses.

  • VisTiles: Coordinating and Combining Co-located Mobile Devices for Visual Data Exploration

    VisTiles: Coordinating and Combining Co-located Mobile Devices for Visual Data Exploration

    Langner, R.; Horak, T.; Dachselt, R.

    In IEEE Transactions on Visualization and Computer Graphics (Volume 24, Issue 1).InfoVis '17, Phoenix, Arizona USA.IEEE,page 626-636,2018.10.1109/TVCG.2017.2744019 Online publication date: 29 August 2017

    @article{Langner:2018aa,
       author = {Ricardo Langner and Tom Horak and Raimund Dachselt},
       title = {VisTiles: Coordinating and Combining Co-located Mobile Devices for Visual Data Exploration},
       journal = {IEEE Transactions on Visualization and Computer Graphics},
       volume = {24},
       issue = {1},
       year = {2018},
       month = {1},
       location = {Phoenix, Arizona USA},
       pages = {626--636},
       numpages = {11},
       doi = {10.1109/TVCG.2017.2744019},
       url = {http://dx.doi.org/10.1109/TVCG.2017.2744019},
       publisher = {IEEE},
       keywords = {collaboration, data visualization, mobile communication, prototypes, smart phones, visualization, mobile devices, coordinated \& multiple views, cross-device interaction, multi-display environment}
    }

    List of additional material

    Slides, Video


Further Publication

Related Student Theses

  • Tobias Knothe

    Responsive Graphs: Anpassung von Node-Link Diagrammen für Mobilgeräte

    Tobias Knothe Juni 19th, 2020 until November 20th, 2020

    Supervision: Tom Horak, Eva Goebel, Raimund Dachselt

  • Eva Goebel

    Responsive Visualizations: Systematische Betrachtung und Charakterisierung

    Eva Goebel Juli 19th, 2019 until Dezember 20th, 2019

    Supervision: Tom Horak, Tamara Flemisch, Raimund Dachselt

  • Lars Arne Beck

    Strategien zur Anpassung von Visualisierungen in Dashboards für die Nutzung auf Mobilgeräten

    Lars Arne Beck Mai 24th, 2019 until Dezember 8th, 2019

    Supervision: Tom Horak, Tamara Flemisch, Raimund Dachselt

  • Elisabeth Baudisch

    Visuelle Datenexploration von Graphen mit geräteübergreifenden Workflows zwischen einem großen Display und Mobilgeräten

    Elisabeth Baudisch Februar 23rd, 2018 until September 22nd, 2018

    Supervision: Tom Horak, Ricardo Langner, Raimund Dachselt

  • Javid Abbasov

    Pointing Interaction with Mobile Devices Using Internal Sensors

    Javid Abbasov Oktober 2nd, 2017 until März 12th, 2018

    Supervision: Tom Horak, Raimund Dachselt

  • Paul Hoffmann

    Eingebettete Mikro-Visualisierungen in interaktiven Matrix-Visualisierungen von multivariaten Graphen

    Paul Hoffmann Mai 24th, 2019 until Oktober 11th, 2019

    Supervision: Tom Horak, Raimund Dachselt