Bryan Routledge on "Mapping the Economic Crisis: Some Preliminary Investigations" by Pierre Bourreau and Thierry Poibeau:

This paper builds a social network from people/companies mentioned in the crisis documents. The relations between two mentioned entities (appear in same sentence) is extracted from the text. The text here is the summary documents of the financial crisis. entity resolution was done using a series of rules -- seems sensible and efficient. The visualizations (Fig 1 and 2) are OK but are not really conclusive of much.

The time evolution exercise is much more interesting.

This is a “node” analysis across time. Can you add the network over time and does that make anything more interesting?