Describes an investigation into “social data understanding” (they use the term sensemaking which I find too pretentioussounding). The actual project they did was pretty cool. They built a website with visualizations of a bunch of US Census data which provided the ability to annotate the data and comment on it (in a blog-comment-like format). The annotations were linked to the comments, so you could draw some arrows pointing to interesting knees in the curve of, say, bartender employment in the US over the last hundred years, and ask questions or make observations. Other people could either take a comment-centric view, and navigate to the annotated visualizations from the comments, or a data-centric view and see comments pop up as they encountered visualizations that others had also seen and commented on.
I’m not sure what the “result” is here other than that more heads are better when it comes to understanding data, but I had fun reading the paper. I also love a good visualization, and the researchers involved in this project have made the visualization toolkit they used for this and other projects available, so they get high marks in my book for that.