[Ieee_vis_open_positions] 3-year PhD position at Aarhus University on Progressive Visual Analytics

[Hans-Jörg Schulz]

Please distribute this call for a 3-year, fully funded PhD position at Aarhus University, Denmark starting February 2018 to prospective candidates.

Focused-PVA: Progressive Visual Analytics for Focus+Context Techniques

Progressive Visual Analytics (PVA) is currently revolutionizing big data analysis. Instead of crunching a whole dataset at once, PVA breaks a dataset down into smaller chunks and processes them in order of importance. This way, after each chunk, we can already output an in-progress result on which to base early analytical decisions long before the dataset is processed in its entirety. You probably know this concept from slowly transmitted images or maps online that are already shown while still being refined. PVA is like that, but for long-running data analysis and visualization operations.

This PhD project sets out to apply PVA to visual analysis techniques that follow the focus+context principle - i.e., by interacting with a visualization, users specify a point or region of interest (the focus) and its periphery of lesser interest (the context). Following the focus+context principle yields a natural, user-driven starting point for data partitioning and ordering strategies for PVA that is expected to outperform sampling-based methods, which partition the data without taking the user into account. It will be your task to explore this new combination of "Focused-PVA" to merge the swiftness and responsiveness of PVA with the interactive, user-driven nature of the focus+context principle.

Research questions to pursue in this project are:


·         What can we imply from the focus/context information in terms of suitable data partitioning schemes and processing strategies for PVA? How does the kind of visualization in which the focus is specified influence these strategies?

·         In which way can we leverage Focused-PVA to instantiate, for example, Progressive Lenses, Progressive Portals, or Progressive Probes on top of different visualizations?

·         How to generalize the above ideas from a binary distinction between focus and context regions to a continuous distance-based concept of importance? How to extend this concept to two or more focal regions?


Work on this PhD topic will be conducted in close collaboration with the DABAI project (https://dabai.dk), which provides the datasets and analysis scenarios on which the developed visualizations are tested. In addition, this research topic is embedded in a larger ongoing research effort to develop a visual analytics concept and system that will center on progressive visualizations using smart lenses or locally inserted "in situ visualizations".

For more information please refer to http://talent.au.dk/phd/scienceandtechnology/opencalls/calls-on-specific-projects/november-2018/focused-pva-progressive-visual-analytics-for-focus-context-techniques/

Application deadline is 1 November 2018 at 11.59 PM MET.

---
Assoc. Prof. Hans-Jörg Schulz
Department of Computer Science
Aarhus University
Åbogade 34
8200 Aarhus N, Denmark
eMail: hjschulz at cs.au.dk<mailto:hjschulz at cs.au.dk>
web: http://hjschulz.net
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://listserv.uni-tuebingen.de/pipermail/ieee_vis_open_positions/attachments/20180830/60463e20/attachment.html>