Viégas focuses her research on the social side of visualization, exploring storytelling, collective sensemaking, and online identity. Her remarks help us visualize her ongoing social exploration in the Many Eyes project.

I think I have a somewhat different background than the other folks in the panel today. So I thought I'd give you a little glimpse of where I'm coming from. I come from graphic design and art history - very traditional training - sort of like all analogue and no digital, printing, making films, silkscreening and drawing is the idea here. The thing I was interested in with graphic design was "how can we use the visual form to communicate?".

How do you really use visuals to communicate a message and to convey information? After I was done with graphic design in the traditional sense I ended up in this place, the same place as Hiroshi - the MIT Media Lab, and this was really the first place where everything was sort of interdisciplinary and it was it was really exciting to get a sense to work with engineers and computational stars and people who used to think about things in a very different way than I did. It gave me a lot of room to explore things that I wouldn't otherwise. Because I'm a visual person, one of the things that I became interested was how could you use computation and technology to continue to convey information visually. That's how I ended up dealing with information visualization.

And I'm showing you these screenshots, they're sort of from the 1990s, because this is what traditional infovis used to look like. It's usually something that's done by scientists for scientists and it's about serious data, you know you have scientists sit in front of a computer working with lots of data and trying to get wonderful insights that they're going to share with the world. They're usually male as well. So, I was sort of you know interested in the techniques and the possibilities there - but I wasn't interested in using it in the traditional way. I was really interested in how can we bring visualization to a regular person. What would they use visualization for? And that's when I started seeing some lay-uses of visualization. So if I can give two brief examples - one from my own work and one of a colleague's work.

I started visualizing people's private data. These are things like email archives, chat conversations, mailing lists what have you - really social data that people were sort of leaving traces of on the web. These are three pieces of email visualizations that I did and because I wanted to have real users with real data using these systems, I had to be very careful about things like privacy so I had to explain to my users you know, if you give me your email to visualize - I'm never going to publish any images without your consent and I'm happy to anonymize all the names. And people would look at me and be like, "ok I guess I trust you, so here's my ten years worth - fifteen years worth of private email." An interesting and unexpected thing was that as soon as I put people in front of these visualizations, the first thing they wanted to do was to share these images with others.

And so they would take screenshots and send it to family and friends, they would bring friends and sit with them in front of the computer and play with visualizations and be like "oh do you remember this is when we started dating and this is when we broke up" and I'm sitting there thinking, privacy? Anyone, remember, we were really worried about privacy, this is your private data? And people were like "oh yeah, yeah - but you know what? I really want to share these. I really want to take these and share them with my friends and family". And all of sudden it struck me, that I was thinking about these visualizations in the very sort of orthodox way. Where it's a single thing, you do it by yourself, exploratory, and I was missing out on the social dimension of these things - people were using these as social artifacts, sort of like how we use photographs, around which you want to tell stories and you want to reminisce about the past.

So that was sort of a wake-up call for me. Meanwhile, on the other side of the galaxy, Martin Wattenberg, who is another visualization researcher, was creating this, what is now a very famous, visualization called Name Voyager. His wife wrote a book, Martin is right there by the way. His wife wrote a book about statistical trends on baby names and the United States since the 1800s and Martin decided to visualize this to help publish the book. So the visualization is on the web, it's available and anyone can use it. It's really cool because you can go and search for your own name and see if you're popular, if you're going down in popularity and so forth. And what Martin started realizing is that even though he had made this visualization and his wife had written the book specifically for expectant parents a whole lot of other people were coming and playing with his visualization. People who are not expecting babies, people who don't even like babies and they were starting to do data analysis for fun, so they would do things like, they would play little games with each other, they would do things like "oh can you find the name that was popular in the 1900s and then plummeted and is coming back in vogue right now", "what is the name that's most constantly popular over time" - and people were doing these pretty sophisticated sort of analysis tasks for fun which is something that we don't see very often.

So, finally Martin and I started working together and we thought you know, the fact that I'm seeing these trends with email visualizations, you're seeing these trends with the name voyager - there seems to be something out there that these visualizations are sort of catalysts for conversation and social interaction. So can we design for that, and that's how we came up with Many Eyes. So Many Eyes is a site that we launched last year where anyone can upload data, they can visualize data and share these visualizations with the world. And it looks this, this is a page from the visualization of in this case, it was United States government spending over time since the 1960s. Some user came to Many Eyes, uploaded this data set and created the visualization. It's an interactive visualization so you can drill down at a bunch of levels and see how your money, or America's money is being spent on say national defense or energy, or whatever other category you're interested in. But the other important thing about Many Eyes is that underneath the visualization you have a lot of comments because the idea is to support conversation around these data visualizations. People will come here, point to something, and be like "what is that big peak there, I don't understand, I don't remember what that was" and then someone else will come and answer "that seems to be the savings and load scandals of the 1980s" and so forth.

And every visualization on Many Eyes has a direct link to its data set which is really important because it frees the data. It allows people to go there and copy the data set if they want and take it themselves and do whatever they want with it on excel, or reuse it in some other way or re-visualize it on Many Eyes. We have a whole portfolio of visualization techniques that span very basic ones sort of like what you have on excel to more sophisticated visualization techniques to even visualizing text that turns out to be VERY popular. People love visualizing all sort of textual input.

So what are people doing with Many Eyes - very very diverse kinds of things. People love to visualize political data sets and speeches and right now we're in an election year so you can imagine how many speeches or debates have been visualized on the site. Citizen advocacy is starting to happen, so people have started to visualize data about their own communities. The interesting thing is that they will visualize, discuss to come up with an hypothesis about something like crime in the community and then take these visualization back to officials in the town to try to engage them in changing whatever situation they're going through.

Something unexpected, there are a lot of visualizations about religious data on the site. A lot of people are interested in analyzing the Bible statistically and it's a very active community. They've been doing all sorts of very interesting kinds of visualizations and very actively debating and talking about what these graphs and charts mean. There's a lot of 'naval gazing' even though Many Eyes is completely public people love to visualize themselves. They will visualize their own blogs, visualize things like oh this is my weight over time, so you know we're cheering for people who are losing weight over time. There are lot of running logs, swimming logs, you name it, it's all there.

Science, even though we designed Many Eyes for the 'regular person', scientists are coming and visualizing research data and actually finding new trends in the data so they're starting to publish papers with the Many Eyes visualization which is really exciting for us.

And finally people are having fun. Every once in a while someone has created a game on Many Eyes that's based on a visualization and how you play with the visualization and select little things. So again people are challenging each other to see patterns or to find things that other people haven't found. And doing art - actually also all sorts of interesting tag clouds and graphs that are more of an artistic manner.

Where are they doing this? All over the place - so they will come to Many Eyes and have discussions on Many Eyes but a lot of the interesting things that happen around these Many Eyes visualization are happening outside of our site. So people will blog about it, people will gives talks about it, Many Eyes made it into one of Lawrence Lessig's' less corruption lectures, there have been institutions who are interested in transparency in government data. Who have also used Many Eyes visualizations. People are creating youtube videos of themselves using the visualizations - to just sort of leave you with a little bit to think about.

I started in graphic design and ended in infovis and in doing infovis in a very weird way. So is there a method to what we're doing here? There is but it is not the 'usual method' of doing information visualization. We tend to go with real data, real users, in real world problems. In translation what does that mean? - instead of just doing a technique for the sake of a new visualization we base it on real data. and it's the data that drives a lot of the techniques that we come up with. We try to work with real users, which is something that a lot of times visualization people don't strive for. Think about it, it's a lot of headache. I can understand why people don't do it very often. The thing for us is that these real users are giving us a lot of insight. The whole reason why we made Many Eyes is because we realized how users had this social connection to these visualizations we were creating and finally - real world problems. You know its so inspirational the motivation that Hiroshi was talking about - the vision, it really inspires the vision with Many Eyes that you're going to give these tools, you're going to 'democratize' visualization. You're going to bring something that is usually done by experts and empower people to have debates around data.

So just to finish, when I first got to the media lab, you know, visualization for most people is this thing that scientists do and my research could have ended up looking very much like this, if this was the tag cloud of my research. The fact that I was trying to figure out what to do with these techniques in a very different context opened up a whole bunch of other opportunities and brought it back to the whole notion of how do you use use visuals to communicate. Is it journalism, is it art, is it literature, sports, you name it, visualization as a communication tool and as a medium is something that we're really interested in and brings together this notion of innovation and balancing art and science for us.