If you haven’t seen this already, it’s worth a look. If you have, it’s worth another look. Thanks to Mashable for the YouTube version of the video.
A group of developers at Facebook have created this fantastic data visualization of Facebook network data overlaid on a globe. They’re calling it Project Palantir, (an LOTR reference), and it grew out of a hackathon project.
Palantir uses the open source jME framework for rendering, and the data visualized are real-time network interactions on Facebook, nicely geo-tagged so you can see visually the global usage of Facebook. The trailing mesh visualization of actions as they happen is most interesting to me because it shows the direction of actions, e.g. friend requests orginating in one location and “traveling” to another. Very cool stuff.
Rumor is that Facebook is considering productizing the visualization; aside from the cool factor, I’m not sure how much value it would add, unless you could watch your own activity flash around the globe. I think data visualizations work best when you’re trying to show benefits or scale, e.g. in a presentation format.
Someone on the video thread page on Facebook has pointed out that Google has a similar visualization at its HQ in of its real time searches plotted on a globe.
Again, this shows the scale (and I guess power) of the data processed, but if you had this visualization available to you on Google, would you use it?
I really do like data visualizations, but for social network data, I think new ways to look at the data are better because we’re still struggling to show value in the network.
What do you think? Do you like the eye-candy, or is it frivilous? Any suggestions for better visualizations?
Find the comments.
Such visualizations help to illustrate the increasingly organic nature of data flow, thanks to the emergence and evolution of social computing tools. The images remind me of infrared photos of the earth, or views from space of cities bathed in electric light, technological representations of various dimensions of a living planet.
On the flip side, how much of what is represented by the swirling colors and dancing icons is the equivalent of “OMG! The Jonas Brothers so totally ROCK!”
Let's hope the Galactic Overlords have patience and a sense of humor.
Agreed. Although I really like what data visualizations can do to explain and induce ah-ha moments, they're not day-to-day useful.
So, maybe the goal is to use them in the right situations, judiciously.
When flickrvision first came out, I was staring at it for a while, then when something I recognized appeared I hit Print Scrn. Pasted into Wordpad for reference, then was able to find the flickr page (IIRC the sequence of events), google for email, find all the usual stuff about the person (who was, it turned out, a few miles from me in meatspace). If I were so motivated, I could've scared the hell out of some random geek innocently uploading some pix in the middle of the night.
Also thought about posting something about it at the time, but I kinda scared myself, and could imagine the reaction of others who already have skewed ideas of my online persona. So I just posted an omg u c twittervision for someone who hates im-speak. 🙂
I imagine there might be some data mining possible with this kind of stream. Organic data flow… icky stuff grows under rocks and logs… who knows what evil lurks in the data visualizations of men?
Interesting comment, especially re. the potential impact of geo-tagging the stuff you do on the Intertubes.
When human interactions are added, e.g. all the connotations of friending someone, hurt feelings and jealousy could find an outlet IRL if it were possible to identify where, when and whom attributes of the interaction. Serious, unintended consequences could arise.
Jealous ex-boyfriend watches visualization for ex-girlfriend's friend requests, sees one for new man, complete with handy geo-tag for easy driving directions. Ta-da! Stalking (or worse) made easy.
Not that you're a stalker (AFAIK), your comment just points out an aspect I hadn't considered. Thanks.