Viégas on Visual Analysis of Social Networks

What a treat! I had had the hon­our of meet­ing and spend­ing the last two days chat­ting with Fernanda Vié­gas from the Visual Com­mu­nic­a­tions Lab. Her work has been and con­tin­ues to be inspir­a­tional for me per­son­ally and to the inform­a­tion visu­al­isa­tion com­munity more sub­stan­tially. She presen­ted a tan­tal­iz­ing talk at the Social Network/ing con­fer­ence at OISE/UofT. ‘Visu­al­iz­ing and Ana­lyz­ing Social Net­works’ quickly demon­strated a small facet of Many Eyes to a new audi­ence and gave us a sneak pre­view of a new tool soon to be avail­able through ManyEyes called Pivot­Graph. The logic of the Pivot­Graph is one of those ah-ha moments — it makes all the sense in the world, but leave it to Fernanda and Mar­tin Wat­ten­berg to visu­al­ize the prob­lem, and come up with a bril­liant way to solve it. Con­sider that social net­works have tra­di­tion­ally been visu­al­ized in two ways: the node-link map and the mat­rix. The com­mon to node-link method is very intu­it­ive, but also becomes quickly cluttered and loses visu­al­iz­a­tion value as the scale of the net­work being mapped grows. The second is the rep­res­ent­at­ive mat­rix, which scales very well, but sac­ri­fices intu­ition for clar­ity. Real­iz­ing that there had to be a way of com­bin­ing the strengths and min­im­iz­ing the weak­nesses, the Pivot­Graph hybrid­ize these two forms using a col­lapsible node-link meta­phor that, inter­act­ively aggreg­ates like nodes and allows for focus on indi­vidual vec­tors. It’s noth­ing short of amaz­ing to see in action! As always, the care­fully visu­al­ized trans­ition anim­a­tions of the Visual Com­mu­nic­a­tions Lab make the products par­tic­u­larly visu­ally and exper­i­en­tially impress­ive. I feel com­pelled to note that this is not to detract from the intrinsic ana­lyt­ical power of these applic­a­tions, or to sug­gest visual super­fi­ci­al­it­ies, but instead to observe and emphas­ize the import­ance of exper­i­ence within a visual ana­lysis product. This is some­thing that is rein­forced through­out the infoviz com­munity in par­tic­u­lar the power­ful effect of tween­ing and time lapse in Hans Rosling’s gap­Minder to appre­ci­ate change over time.

Fernanda also gave a quick user story that demon­strates a Visual Com­mu­nic­a­tion Lab suc­cess. One of their early users (two days after they went live) pos­ted a data­set and visu­al­iz­a­tion of name col­loc­a­tion from the New Testament.

He was quite taken with the res­ults and repor­ted these on his blog with accom­pa­ny­ing screen shots. He was shortly over­whelmed with hits and story picked up on the tra­di­tional media. In fact, there was even a video uploaded to You­Tube demon­strat­ing a another user’s inter­ac­tion with the visu­al­isa­tion. Fernanda emphas­ized that this was not because of the nov­elty of the tool or its product, but because of the con­tent and res­ult of the visu­al­iz­a­tion itself. She stresses, and I can cer­tainly attest to this myself, that we are truly start­ing to see the tre­mend­ous value of inform­a­tion visu­al­iz­a­tion when we make new real­iz­a­tions because we are able to see the data in new ways. When we are able not just to present find­ings in the most clear and effect­ive man­ner, but when we are able to use the visual tools for ana­lysis. ManyEyes is push­ing the envel­ope of PopViz!

More inform­a­tion on the Pivot­Graph is avail­able at the Visual Com­mu­nic­a­tion Lab site — and keep your eyes on the ManyEyes tool­set for avail­ab­il­ity — I sense that this is a tool that is cross­ing an inter­est­ing visual bound­ary and will shift the way in which we con­sider the object being rep­res­en­ted as much as how we choose to rep­res­ent it.

I apo­lo­gize to Fernanda for my rather gush­ing post about her talk, as she is very earn­estly seek­ing con­struct­ive cri­ti­cism of ManyEyes — but I do prom­ise to share thoughts on the effect­ive­ness of Pivot­Graphs when I get my chance to play with it online. As for ManyEyes cur­rent offer­ings — they simply beg us to come up with data to visu­al­ize and if you haven’t already, I hope that you will take an oppor­tun­ity to enjoy some cool data visu­al­iz­a­tion with any sets that you hap­pen to have kick­ing around.

Update: BTW, Fernanda, as I men­tioned in a pre­vi­ous entry, blogged a num­ber of the talks from InfoVis this year. Her recap of her own panel is well worth a look.