We're sharing an interview we conducted this month with Severino Ribecca. He is the InfoVis (Information Visualization) expert behind The Data Visualization Catalogue. There is a lot of interesting insight into his work. In particular, check out his tips for working with big data and making charts for mobile.

Meet Severino Ribecca

Severino is a graduate of the University of the Arts in London and an award-winning designer of dataviz and infovis projects.
severino ribecca, infovis expert
He is a graphic designer with a particular expertise when it comes to visualizing data. He originally started The Data Visualization Catalogue to develop his own knowledge and create a personal reference tool. Over time, he thought the site would be useful to both designers and anyone in a field that regularly uses data visualisation.

We've used it several times at ZingChart to ensure we're selecting the right chart type for the job. And we were lucky enough to ask him all our biggest infovis questions, which we're sharing with you now.

Can you tell us about your first data visualization project? Did that first project pique your interest in dataviz, or did it take a few design projects before you realized this was an area in which you wanted to specialize?

I wouldn’t say there was a single project that excited me about the field. It was more of a gradual process. Originally, I was actually more interested in getting into motion graphic design and knew very little about information design or data visualisation.

That changed during my first year of work towards my bachelor’s degree. After a persuasive presentation and the endorsement from a previous tutor, Ian Noble, I decided to try out the “information design” pathway.

information is beautiful website
Over the course of a few small, related projects and after seeing more examples of design work visualising data, I became more interested in the field of data visualisation and information design. I was particularly inspired by David McCandless’ blog Information is Beautiful.

With information design, you have to think about it more and there is more problem solving involved. Other design fields were more artistic, self-expressive and subjective. Of course there’s strategy to apply in any field of design. But with information design, I enjoy the challenge of balancing both styling aesthetic form and engineering something that works. Information design combines both logical problem solving and artistic expression, which keeps it interesting.

I also have a lot of interest in science. At one point, I considered a degree in engineering instead of graphic design. Information design bridged my interests: I liked how there was actually some science behind the design principles used in information design and data visualisation. There’s an engineering and architectural-like quality to them. You could say data visualisation is like a marriage of art and mathematics.

You say “I believe design to be an important tool to educate and inform people on complex subjects like science and world finances. By making information more visual, people can understand something a lot quicker then reading through large paragraphs of text or massive databases.”

Can you give an example of your work or another that you think best illustrates this ease of understanding through visual rather than text communication?

debt crisis infovis
Well in terms of world finances, I think my graphic on the European debt crisis is a good example of this. Very few people actually look at their nation’s budget or the amount of debt they owe. But even if they do, it can be hard to grasp numbers at such a large scale - especially when they’re just displayed in a spreadsheet.

The visualisation I produced brought the scale of national debt into perspective by comparing the proportional debt and GDP of each EU country to each other.

Follow up question: do you think adding interactivity to data visualization helps aid understanding?

Oh definitely. Interactivity not only allows for more layers of information to be added onto the visualisation, but it can also make it much more engaging.

Is there a particular type of data that inspires you to work on visualizations (non-client work)?

Historical and cultural information is interesting to me.

greek myth infovis
Someday I would like to have the time to produce something like another Greek Myth Family Tree Diagram, but for another religion.

Your site is mostly about choosing the right chart type. Which type do you see most misused/which is the most misunderstood?

I don’t think there is necessarily a “right” chart to for any given situation. Infovis is still an art to a certain degree. So you have to take other factors into consideration. However, I hope my site helps steer people in the right direction.

I’ve actually heard from a number of people, that the site is being used more as a source of inspiration when creating charts, which is interesting.

But anyway, in terms of misused charts, it’s hard to say exactly because I’m not exposed to a large enough quantity of other people’s charts. Jon Schwabish would be a good person to ask about this since he runs the helpmeviz.com website.

Which type do you come back to time and again?

I think that would be the trusty bar graph and all its incarnations, for a number of reasons:
- It’s versatile: bar graphs come in multiple forms, such as stacked, grouped, vertical, horizontal.
- It can be adapted or tweaked visually to look more interesting for infographics.
- Its values are displayed in a way that makes them easy to accurately compare.
- Almost everyone is familiar with how it works.

Do you hate 3D pie charts like most infovis experts? :-D

“Hate” is such a harsh word… 3D pie charts have never caused me enough harm to feel that way towards them! But I would say to avoid them like they’re some kind of freak.

Making the chart 3D doesn’t do anything to help visually communicate the data - it’s just a gimmick. In fact, displaying a pie chart 3D with a tilt distorts the area of each slice (some appear bigger or smaller then they should be) and this makes the chart harder to interpret and misleading.

How do you stay on top of user research/information design trends and news?

I usually check out certain websites every once in awhile, like Visual Loop. I also follow data visualization feeds on Twitter.

data stories
There are also a couple of good infovis podcasts out there, like Data Stories. However, I don’t keep as up-to-date as much as I’d like. I wish just I had more time to do so because the field is developing so fast!

You are based in Lithuania now. Do you work with clients remotely for the most part? What are some of the keys to understanding style and functionality with people who are not in the same room?

Well I’m still back-and-forth between the UK. But yes: most of the freelance projects I do are done remotely. It’s been like that for awhile, so I didn’t see the need to stay in one place when I’m getting clients from around the world. The internet has now enabled me, like so many others, to live a location-independent lifestyle.

The key to working with clients you’ve never met is good communication and to make sure you’re regularly in contact with them. This can be done through clear, detailed e-mails and Skype. Project management tools such as Basecamp or even just a simple Google Drive help as well.

As a designer, how do you decide which charting library to use (assuming the libraries have the chart type you are seeking to create)?

A lot the charts I create are custom. Therefore, I tend to draw my charts from scratch in Adobe Illustrator. Personally, I rarely use any chart libraries. But I have suggested D3 a couple times.

What kind of chart types or charting features would you like to see in a library?

I would like to see a product that makes developing custom chart designs easier. So you could create your own chart type by assigning variables to different forms of visual encoding (like length, area, angle etc), which might come in the form of “smart shapes.” Somebody is already working on this, but it’s in the development stages I think.

What recommendations do you have for people designing charts that visualize big data/lots of data points?

My first piece of advice would be to maintain a good knowledge base of which chart types have a high data capacity. For example:
- Scatter plots
- Spiral plots
- Line graphs
- Area graphs

Some other chart types really struggle and become illegible when you use a huge dataset.

But if you’re unfamiliar with the chart type you’re using or it’s a new design, then I would recommend testing out the chart to see how it handles the data. Sometimes you will need to tweak the chart design in order for it to work well with large datasets. There is no single way to adjust design for this, it really depends on each chart and the visual system.

We love to know your thoughts on infovis for mobile devices.

What do you think are some of the most important things to keep in mind when a chart might be viewed on a large display and a mobile phone? Or would you recommend two totally separate projects?

I would avoid any charts with a lot of detail on mobile, especially if that detail is meant to be read closely. A possible exception would be a complex “data art” visualisation that’s communicating a more generalised view of the data.

But when you’re producing charts that will be displayed on a mobile device, it is safer to opt for more simple charts such as a bar graph, pie chart or area graph. You need to make sure all the important visuals and text will be just as legible on a mobile device, as it is on a desktop screen.

Responsive design can help… sometimes. For example, on a line graph, you could reduce the number of lines on an axis scale and enlarge the text to fit well on a mobile screen.

Also, it is best to avoid having any charts that require zooming on mobile. You should just be able to scroll down and read it clearly.

What are your questions?

Share your infovis questions with us in the comments section below. We're happy to give you our two cents, and maybe Severino will see them too and provide an answer!