It’s amazing how much great, free research is available on the web these days – if you can find it. One place a lot of people might not know about is the HP Social Computing Lab, which “focuses on methods for harvesting the collective intelligence of groups of people in order to realize greater value from the interaction between users and information.” It appears they have a team of about 14 researchers, led by Senior Fellow Bernardo A. Huberman, and they publish a couple of papers a month on the topic.

One of the papers I found most interesting lately was Revealing the Long Tail of Office Conversations, by Michael J. Brzozowski & Sarita Yardi. What the authors were interested in exploring tied was how social media tools (blogs, wikis, etc.) could breakdown geographic distances, work group boundaries, and organizational hierarchy in the organization. More importantly, they wanted to look at what motivated individuals to “invest their own time in creating content for public consumption.”

In order to do so, HP decided to conduct a study of themselves – the design and use of Watercooler, their social media platform. Included in their study was a survey of 144 users, 12 months of log files from the internal blog server, and interviews with approximately 96 blog users. The paper is broken up into sections on the design of Watercooler, results of the experiments, perceived benefits for users. The paper is also quite short- only about four pages – so it is well worth the read.

What I really liked were the readership and commenting network “social graphs”. I know they’re kind of hard to see on the screen, but in the first each node is an individual user color coded by business group, and each arrow represents someone having read someone else’s blog at least 3 times. It is notable that a couple of authors clearly emerge as “hubs” in the network, and also that 55% of links went outside of one’s own business group. It is also clear that the two green dots in the upper left corner are having an affair of some sort (just kidding).

The second figure focuses on comments instead of views – and the results are even more striking. One individual (J) emerges as the center of the giant component, due to a bit of a virtuous circle – she comments a lot on other people’s posts, and they in turn comment on hers. It’s also notable that there are several “disconnected clusters” where people are chatting away. And those two green dots on the side are still clearly having an affair (just kidding again).