Digg
Digg.com is one of the most-visited sites on the internet. The site's ever-changing content, all submitted by members, provides a vivid window into what's interesting online, right now.
Since early 2006, Stamen has been collaborating with Digg on numerous ongoing visualization projects. Our work together focuses on research into the astonishingly rich and diverse ecosystem of the Digg user community.
Some of this work is released to the public in the form of live data-driven interactive projects at labs.digg.com. We are also in continuous exploration mode with Digg's data, generating internal research artifacts which provide new and useful views of the Digg community's behavior (see "Research," below). Our goal is to continually extend the possibilities of live data visualization, while addressing Digg's real business needs.
Digg Labs:
Digg moves very quickly, and has a great many stories submitted every day, so good material can sometimes fly by before you even know it. Stamen's live interactive visualizations for Digg look beneath the surface of this active community's activities, and allow for a broader (and deeper) view of the site.
Digg Swarm:
Digg Swarm is a lyrical view of Digg. Stories come in as circles with the title inside of them, and diggers "swarm" around these stories when they digg them. Every time a story gets dugg, it increases in size — so the bigger the story, the more active it is. As people digg more stories, they move from circle to circle, and increase in size. You might see enormous diggers moving quickly from story to story; those seem to be people digging without taking the time to read stories...
Stories that are closer together are being dugg by the same users, and you can roll over stories to see these connections. The thicker the line, the more diggers in common that story has—which starts to suggest connections between stories over time.
Digg Stack:
Digg Stack shows diggs occuring in real time on up to 100 stories at once. As stories are dugg, they appear along the bottom of the screen, colored according to how popular they are. Users digging these stories appear as falling blocks that stack up as activity increases. The zoom level is adjustable, so you can focus in on the most recent stories, or pull back for a broader overview. You can also pause the visualization if it's flying by too quickly, or if you want to focus on a specific story more readily.
Clicking on an individual item pulls up more information on that story—who dugg it, how many comments it has, etc.—as well as providing a sparkline-style graph that shows a more detailed hour-by-hour display of activity on that story.
Research:
One of the best thing about working with Digg is the access it affords us to a wildy rich ecosystem in a constant state of flux. Our public-facing work on Digg Labs is buttressed and supported by an extended set of investigations into the parameters of this ceaselessly shifting landscape.
One of our first stabs at the digg data was a simple time mapping; we wanted to know what an hour of activity on Digg looked like. The left-right axis in the image above represents a digger's newness to the site, and the top-bottom axis shows the time that a story was submitted. Every dot, then, is a given site member digging a given story. New site members digging current stories will show up in the bottom right, long-time members digging older stories will be represented in the upper left, etc.
Since some stories get more attention than others, we expected to see highly popular stories appear as more or less continuous horizontal lines across the whole spectrum of users; and in the image above this is just what you see. The vertical lines, on the other hand, are individual users digging story after story after story—or maybe bots.
Stories on Digg have their own arc in their path from submission to promotion; a unique fingerprint that can be used to better understand the particular characteristics and circumstances that lead to promotion. The image above shows a variety of sparkline/historgrams that each story leaves as it's dugg; this reasearch would ultimately find it's way into the public release of Digg Stack.