Research Scientist - Visual Analytics for Network Comparison

The Visual Data Analysis research group at KU Leuven (University of Leuven) is part of the department of Biosystems at the Faculty of Bioscience Engineering. The group focusses on complex data exploration in the domain of biological and agricultural sciences, through the use of visual analytics and topological data analysis. The group is currently looking for a dynamic and highly motivated PhD student.

Project

Networks and graphs - both static and dynamic - play an important role in data science. Not only can they represent primary relational data (e.g. gene interactions) but many algorithms generate networks either as intermediate or final output (e.g. DBSCAN clustering and topological data analysis).

We can quantitatively compare networks at different resolutions, and distinguish between changes in topology and changes in characteristics of the nodes and links themselves. Although many measures exist to do so (including degree, closeness, betweenness, etc), they mainly focus on the amount of difference but fall short in giving real insight in the quality of that difference. We want to focus on this qualitative understanding of networks rather than only a quantitative one.

We want to develop a human-in-the-loop visual analytics toolkit to support the user in exploring differences between two or more networks. Visual Analytics (VA) is often described as the science of analytical reasoning facilitated by the visual interface and combines interactive data visualisation and novel visual design on one hand with machine learning on the other.

The methodology will involve - among other things - requirement elicitation, custom visual design, definition and implementation of interestingness features, as well as definition of novel distance metrics for topological data analysis.

Keywords : graph, network, topological data analysis, visual analytics, data visualisation, multilayer networks

Profile

The candidate should have a Master's degree in Computer Science, Bioscience Engineering or similar. In addition, they should have a solid understanding of network analysis and a strong foundation in general data visualisation principles and techniques.

They must have good programming skills (incl Python, and if possible Javascript) and the ability to work independently as well as part of a collaborative cross-domain research team. The candidate should be proficient in oral and written English, possess excellent communication and multi-tasking skills, be team-oriented, proactive and result-driven.

Interested candidates should submit their curriculum vitae, contact information of 2 or 3 referees, and a motivation letter. This motivation letter should specify why the candidate applies to this position, and clearly illustrate possible past experience.

Offer

• A full-time PhD position for one year; after a positive evaluation, the contract can be extended to three additional years (four years in total)
• A working climate where trust, (international) collaboration, and commitment are essential
• An excellent young, stimulating and supportive international research environment
• High level scientific training at a top-ranked university; training in academic, thematic and soft skills.

Visual Data Analysis Lab

Biosystems Department

Interested?