School of Computing

Staff Profiles

Dr Sara Fernstad

Lecturer

Background

I am a Lecturer in Data Science with ten years' experience of Information Visualization research. Before joining the School of Computing Science at Newcastle University, I was a Senior Lecturer in Computer Science at University of Northumbria at Newcastle. Prior to that I held post-doctoral research positions at Cambridge University, UK, and at Unilever R&D, UK. I completed my PhD on algorithmically guided visualization for analysis of high dimensional and heterogeneous data in 2011, at Linköping University in Sweden.

My main research interests are in the area of Information Visualization and particularly in challenges relating to incomplete (missing), high dimensional and heterogeneous data, visualization of biomedical and ‘omics-type’ data, and the combination of visualization and data mining methods.

I am currently supervising three PhD projects as main supervisor:

  • Mining the Biological Data Deluge: Highly Interactive Exploration of 'Omics Type Data Through Novel Visualisation Methods. Alexander Macquisten
  • Visualisation of Heterogeneous Multi-source Data. Michael Adele
  • Designing an Interactive Visualisation Framework that Supports Understanding of Deep Neural Networks. Kenan Koc

Google Scholar: Click here.

Research

My main research interests are in the field of Information Visualization (Visual Analytics/Data Visualization) and I am particularly interested in challenges relating to high dimensional data, heterogeneous data and uncertainty in data. These challenges are all highly relevant in a range of application domains, not least in biomedical domains and life sciences, which are becoming more and more data driven.

From a methodological point of view, my interests are focused around the combination of visualization and data mining methods and user-centred design approaches. Visualization and data mining both address the same type of challenges, with similar goals but with slightly different approaches and strengths. By combining the two we may be able to find better solutions to major data analysis challenges.

My research to date can broadly be separated into four major themes, which ever so often crosses each other’s paths: Visualization of Missing Data, Visual Exploration of Microbial Populations, Interactive Visual Exploration of High Dimensional Data, and Visualization of Heterogeneous Data. As an overarching theme to this lies the concept of 'interestingness'. How do we bring out the most interesting and useful information in a dataset? How do we define what is interesting? How can we use 'interestingness' to support knowledge discovery and gaining of insights in the most useful way?

Google scholar profile: https://scholar.google.com/citations?user=DKcE7HwAAAAJ&hl=en&oi=ao

Invited talks:

  • Designing Interactive Visual Analytics Tools for Exploration of Microbial Communities. Centre for Bioinformatics, Hamburg University, Germany. Jan 2018.
  • To See What Isn't There - Visualisation of Missing Data. Unilever R&D Colworth, UK. Oct 2017.
  • To See What Isn't There - Visualisation of Missing Data. School of Computing Science, University of Glasgow, UK. April 2017.
  • To See What Isn't There - Visualisation of Missing Data. School of Computing Science, University of Leeds, UK. Feb 2015.
  • Designing Visual Analytics Tools for Exploration of Microbial Ecology. Oxford e-Research Centre, University of Oxford, UK. May 2014.

Publications