THE AU: MUSIC ARTS ON THE AU THE IRIS FOOD & LIFESTYLE THE IRIS: GAMES & TECHNOLOGY HELLO ASIA! AU APPROVED AU ABROAD

the AU Interview: Professor Tamara Munzner & the CSIRO's Christian Stolte discuss how to create great visualisations


Photo credit: Janis Franklin

Visualisations have saved lives. In an increasingly complex, data-driven world they are becoming especially useful at helping people understand complex things. Vivid Sydney is playing host to a visualisation masterclass which will see the likes of Tamara Munzner, a professor at the University of British Columbia and the CSIRO’s Christian Stolte, come together to teach students about the principles and methods for creating great visualisations. The AU Review sat down with the Senior Bioinformatics Visualization Specialist, Christian Stolte and Professor Munzner to learn more about their work...

Professor Munzner has worked across a range of different domains including: genomics, evolutionary biology, geometric topology, computational linguistics, large-scale system administration, web log analysis, and journalism, to name a few. During this interview we were able to learn more about visualisations and their power to unlock and crystallise important information.

Can you tell us a little bit about yourself and your background?

Tamara Munzner (TM): I got started in visualization in the early 1990s as technical staff at a mathematical visualization research group called The Geometry Center. Our big software project was Geomview, an interactive 3D visualization system that supported non-Euclidean geometry and 4D projections. The two major videos that I helped to create were Outside In, about turning a sphere inside out, and The Shape of Space, about spaces that are finite but have no boundary. I decided that research was the One True Way and that I should get a PhD to be able to keep doing it, and ended up at Stanford in Pat Hanrahan's group from 1995 to 2000. I changed my focus a bit to information visualization, where the big question is how to visually represent non-spatial data. My timing was very good: 1995 was my first year there and also the year that the InfoVis conference started, so I've attended every one. After a few years in a research lab, I emigrated to Canada to become a professor at the University of British Columbia. I've been in Vancouver since 2002.

What can audience members expect to learn at the Data Visualisation masterclass?

Christian Stolte (CS): Our goal is to give people a good overview of the principles, methods, and tools for creating great visualisations.

In your opinion, what are some of the biggest mistakes people are making with visualisations and their treatment of big data?

TM: Using color correctly is very tricky, because the way that people tend to think about it from a computer graphics point of view doesn't serve you well when you're trying to encode information visually. People are often over-eager to use 3D for nonspatial data; it's best suited for spatial data when the task involves understanding its intrinsic shape.

Programming languages like R and software like Tableau are becoming common methods for preparing visualisations. What programming languages and/or software would you recommend?

TM: Those two are a great start. The next obvious place to go D3: browser-based visualization with Javascript has become wildly popular.

Will people attending the Data Visualisation Masterclass be learning any programming languages or software? What level are you pitching the class?

CS: We won't have enough time to get into any programming language in-depth, but we will go over some examples in R and JavaScript. That said, programming is not the main focus of the class, but we want to make people aware that you can achieve much more effective results by teaming up with a programmer.

Can you give us an example of where data visualisations have really helped in telling an important story or presenting important results?

CS: Scientific journals are full of examples, but to cite one that actually helped save lives, there was John Snow's map of cholera cases in the London epidemic of 1854: it allowed him to trace the origin of the outbreak to a contaminated water pump. See below.

If an analyst is presented with and overwhelmed by a huge data set are there any suggestions you can offer in terms of tackling the task of visualisations?

TM: A single view doesn't have to be all things to all people. A powerful approach to visualization is to have multiple linked views, where each them is showing different aspects of the data and they're linked together with shared highlighting, so that you can see where the same items fall in the different visual encodings.

Is there anything else you'd like to tell the readers of The AU Review about the Data Visualisation Masterclass, big data and visualisations, etc.?

TM: People who want to understand more about visualization might like to check out my new book, Visualization Analysis and Design. It provides a systematic, comprehensive framework for thinking about visualization in terms of principles and design choices. The book features a unified approach encompassing information visualization techniques for abstract data, scientific visualization techniques for spatial data, and visual analytics techniques for interweaving data transformation and analysis with interactive visual exploration. It emphasizes the careful validation of effectiveness and the consideration of function before form.

Here are a few recent systems:

Vismon is a visual tool for fisheries data analysis. The input is a simulation with two input parameters (the management options) and several output indicators. The purpose of Vismon is to help decision makers to quickly narrow down all possible management options to only few that are agreeable to all stakeholders. Then, a detailed trade-off among the few chosen management options can be performed.

MizBee is a multiscale browser for exploring relationships between chromosomes between different species using comparative genomics data. Using side-by-side linked views, MizBee enables efficient data browsing across a range of scales, from the genome to the gene.

-----

Tamara Munzner and Christian Stolte will appear at the Data Visualisation Masterclass at the University of Sydney for Vivid Festival on June 8. For more information and tickets please visit: http://www.vividsydney.com/event/ideas/data-visualisation-masterclass or datavismasterclass.org
For more information about Tamara Munzner’s book, “Visualization Analysis and Design” please refer to: http://www.cs.ubc.ca/~tmm/vadbook/

More information on Vismon can be found at: http://www.vismon.org/

More information on MizBee can be found at: http://www.cs.utah.edu/~miriah/mizbee