
Martin Wattenberg is a leading expert in the field of information visualization. Even if you know nothing about the field, you are likely to recognize some of his projects: for his wife's book, he created the
baby name voyager, which went viral on the web;
history flow tracks the edit history of wikipedia pages; one of his earliest projects was the
map of the market; and many other projects can be viewed at
his website.
He is the founding manager of IBM’s Visual Communication Lab, which researches new forms of visualization and how they can enable better collaboration. The group is making strides to democratize information visualization with
Many Eyes and the
New York Times Visualization Lab, which provide users with the tools to visualize and analyze text and data in the style of Martin Wattenberg.
I am sincerely grateful to Martin for his thoughtful responses.
Q. What can I say to help people better understand the value of information visualization?
A. A jeweler selling a ring always asks the customer to try it on. The same trick works with data visualization: if you can show people their own data in the form of a chart, they find it hard to go back to a table.
Q. How might visualization be used in ways that it is not currently being used?
A. My collaborator Fernanda Viegas and I believe that visualization is an expressive medium. One of the themes of our work is extending this technology to spark memories, to spur creativity, to convey subtle thoughts and emotions.
Q. Many Eyes and the New York Times Visualization Lab are tremendous efforts to democratize visualization. How do you think it’s going so far? How would you like to see it evolve?
A. I'm very happy with how they're working. At the same time, one of the exciting aspects of the field is just how far we still have to go. One important area for improvement is handling data other than numbers: text, images, and video. We've done some work here, but there's huge progress still to be made. I'd also like to work on ways to deepen the analytic discussions that take place around visualizations.
Q. Briefly, please describe the process of creating visualizations start to finish. I’m sure no two are the same, but generally how do you come up with the idea, how much planning is involved, how long does it take, how many people work on it and in what roles, what tools are used, etc.?
A. The process can vary tremendously, and it's different for different people. The one constant is that every single good idea in the visualization world has come from intense engagement with real-life data sets. That's why my first piece of advice for people starting to work on visualizations is to begin by finding data they care about deeply.
Q. What makes a visualization really good?
A. Different projects succeed in different ways. Florence Nightingale saved millions of lives with her charts. The Hertzsprung-Russell diagram and the periodic table of elements led to huge scientific advances. But I also value gentler aims. When I look at the work of Jason Salavon, for instance, I feel like I'm making contact with a world of ethereal platonic essences.
Q. What sort of education would you recommend for aspiring information visualizers? Which subjects should they study, and in what order (computer science, design, mathematics/statistics, human-computer interaction, other)? Are there particular programs such as MIT’s Media Lab that you would recommend?
A. Anyone doing visualization will have to think about colors, mouse clicks, numbers, algorithms, and so on. That doesn't mean you have to take academic courses in design, HCI, statistics, and programming, however. You can get a working knowledge of these areas in many ways.
Looking at people whose work I admire, their educations had three things in common. First, they all learned at least one subject deeply--though there's no pattern in the subjects they studied! Second, they didn't simply learn theory, but also became good at the craft aspects their field, whether design, programming, or architecture. And finally, everyone I know who does good work is intensely curious about the world and always learning pieces of new fields--as a result they find it easy to see the big picture in a project, and to communicate with collaborators.
For people choosing a graduate school, a big issue is finding the right advisor. Your best bet may be to look at who is writing the academic papers that interest you the most. You can also learn a lot outside of academia. When it comes to visualization there's real intellectual action at corporate research labs and at media companies.
Q. What should aspiring visualizers know that would not be obvious to ask?
A. The time is right to make an impact in visualization! This field is new, and so is much more accessible than doing algebraic geometry or playing the violin. Today, if you're interested in data and graphics and have a creative approach--especially if it's based on a deep idea from a seemingly unrelated field--it's entirely possible for your visualization to sweep the web. The key to success is finding data you think is tremendously important, and that you want to share with other people.
---
Related:
Q & A with Nathan Yau of FlowingDataHe creates ways of seeing information (Boston Globe 12/12/08)
Rise of the Data Scientist (FlowingData 6/4/09)