Lately, I helped to organize a hackathon, together with my friends at Dataninja and Action Aid on the data published by Regione Emilia-Romagna about the network of firms that won public contracts to work in the post-emergency construction process.
During the hackathon, we cleaned the data on these firms, together with the activists, and tried some preliminary network analysis.
The stage of the research is far from being significant, and takes more work.
Our aim is to see any pattern emerging for participation in this process. At this stage, what you can see below is only a simple bipartite network projection weighted for the amount of money the public contract is. Firms are identified through their fiscal code, which needs more cleaning.
Although some form of clustering already appears, there are no clear-cut conclusions. What might be interesting is to project the public contract-firms relationship as a matrix multiplication so that the graph could become only a graph of firms, which relationships are not “mediated” by the ‘public contract’ type of node, but are accounted for in the edges.
This is an activism project not funded, so if you are interested in this give me a call!
originally published in italian