This dataset is only the nodes with the most influence, measured by PageRank
Node Legend:
People: Blue
Organizations: Green
Federal Agencies: Pink
Clusters:
When rows and columns aren't the same type of node (i.e not clustered) they appear black.Blocks of color show influence by node type
Explanation of Centrality Measures
PageRank: How well-connected a node and its direct connections are. If you and your network have many connections, your pageRank is higher. Can be seen as power-players.
Degree: Who is the most connected to other nodes? This can relate to influence, or strategic importance for communication.
ClosenessWho can most quickly access other nodes? A high score means you have a lower distance to other nodes and can communicate easily.
BetweenessHigh scoring nodes are typically "bridges" between other nodes or clusters. They are the shortest pathways of communication within the network.
Based on Mike Bostock's
Les Mis Co-occurrence Matrix Example. Built with D3
and
Lynn Cherny from NetworkX analysis with accompanying talk slides and blog post.