In today's SNA Wednesday, we begin the discussion of clustering methods and measures for social networks. Identifying clusters of nodes, in a variety of ways, tells us about the cohesion, connectedness, density, compactedness of networks; are all the nodes connected and how tightly connected. This is related to node centrality measures but is a much more global set of network measures.
Lets start with some definitions and common concepts:
Clustering is the overall term for studying how "connected" the nodes of a network are to each other. It is the overarching term used in SNA for these type of measures.
A component is a subset of nodes of the network that are connected; ie. each node-pair has a path between them. Not all networks are one big component, but social networks very often have one giant component which comprises a large percentage of the network's nodes.
Cohesion is generally thought of as how "connected" a node is to another set of nodes or how "dense" a network or giant component is.
A clique is a subset of nodes which are all connected to each other. In slightly more technical language, a clique is a maximally set of connected nodes. Nodes may be part of multiple cliques and the structure of cliques are fragile.
These measure are very useful when understanding how information or disease might spread.
Next week's SNA Wednesday we will begin with simple density and cliquishness measures.
Here is a video to get you started: