Link: https://kieranhealy.org/blog/archives/2013/06/09/using-metadata-to-find-paul-revere/
In this project, the author creates an alternate history scenario in which the British Crown is using metadata analysis to find potential “terrorists”, ultimately finding Paul Revere as particularly suspicious. It’s actually a little funny to get drawn into this alternate world, but the IRL implications are quite a bit darker: when organizations harvest your metadata and claim it’s not identifying, that’s nonsense.
The project itself is rather simple, with the people of interest or organizations of interest representing the nodes and their connections to each other being edges. At first, the relationships of people to organizations are just given binary indicators, but this simple data is then turned into a matrix of the number of mutual members of each organization which is used for most of the analysis. There are then some further tests, using betweenness centrality and eigenvector centrality, to prove that Paul Revere is indeed a key figure in this network of enemies of the state.
The project is rather unfortunately not interactive (unless I missed something), but we do get a rather thorough walkthrough of how our agent of the crown found Paul Revere to be a major suspect: he’s the biggest connection between most of these groups, and the little bits of historical humor (eigenvector centrality will never be useful was my favorite) keep it entertaining.
Hi Evelyn, this blog post was really interesting. I really like how they set up an alternative scenario and added humor–that feels very interactive even though the project itself can’t be adjusted. That’s something I felt that was missing in the project I looked at (the MoMA Abstraction).
This site also caught my eye, but I love your explanation and perspective of the data. Do you think the site is better off not being interactive? Would that construe the given data or would being interactive be more helpful?