SNA is methodology for describing and analysis a social
networking, on basis of graph theory.
In SNA, we usually represent a ‘person’, or ‘entity’
as a Node. (i.e. the Social Actor) Communications or Interactions between (like send a
message, subscribe/follow the blog) between is regards a link between them. (i.e.
the Relationship Actor)
Thus studying the relationship between the nodes
and links, formed the base on Social Network Analysis.
By measuring the Centrality (Outgoing direction)
and Prestige (Incoming direction), in terms of:
- Degree, the number of direct connections
- Closeness, the inverse of geodesic distance,
- Betweenness, the number of times a node connects pair of other nodes.
However, in real case, it is almost impossible to
count and calculate each nodes and relationship, on the other hand, we will usually calculate
the ‘Group’ Value, that normalized by dividing by the maximum possible value,
to obtain the generic Group value for ease of calculations and analysis.
Consider the following social network by 5 actors:
2a) The above network is a undirected network, which means the direction of each node is not important.
A cut-off point occurs at David, that means if David is removed from the network, it will make the network disconnected.
Node {A,B,C,D} also form a 2-clique, that means there exist an sub-network , that maximum distance of any nodes is 2.
Also, Node {A,B,C,D} also form a 2-plex, that means every node is connected to 2 others within the network.
For the above network,
Density = 2L / g (g-1) = 2 x 6 / 5 x 4 = 0.6
2b)
To access how influential and also the quality of a social network, we could use SNA techniques Degree / Closeness / Betweeness Centrality to analysis it.
Because centrality identify which nodes are in the center of network, by number of direct connections (Degree), average distance (Closeness) and number of shortest path that it was relied on (Betweeness).
Degree Centrality
Consider the following social network by 5 actors:
A cut-off point occurs at David, that means if David is removed from the network, it will make the network disconnected.
Node {A,B,C,D} also form a 2-clique, that means there exist an sub-network , that maximum distance of any nodes is 2.
Also, Node {A,B,C,D} also form a 2-plex, that means every node is connected to 2 others within the network.
For the above network,
Density = 2L / g (g-1) = 2 x 6 / 5 x 4 = 0.6
2b)
To access how influential and also the quality of a social network, we could use SNA techniques Degree / Closeness / Betweeness Centrality to analysis it.
Because centrality identify which nodes are in the center of network, by number of direct connections (Degree), average distance (Closeness) and number of shortest path that it was relied on (Betweeness).
Degree Centrality
| Node | Density | Degree Centrality |
| A | 3 | 3/5 = 0.6 |
| B | 2 | 2/5 = 0.4 |
| C | 2 | 2/5 = 0.4 |
| D | 4 | 4/5 = 0.8 |
| E | 1 | 1/5 = 0.2 |
Closeness Centrality:
(g-1) = 4
| Node | Sum of Geodesic distance | Closeness Centrality | Normalized Closeness Centrality |
| A | (1+1+1+2) = 5 | 1/5 = 0.25 | 4/5 = 0.8 |
| B | (1+2+1+2) = 6 | 1/6 = 0.16 | 4/6 = 0.67 |
| C | (1+2+1+2) = 6 | 1/6 = 0.16 | 4/6 = 0.67 |
| D | (1+1+1+1) = 4 | 1/4 = 0.2 | 4/4 = 1 |
| E | (2+2+2+1) = 7 | 1/7 = 0.142 | 4/7 = 0.57 |
Betweeness Centrality:
(g-1) = 4
| Node | Betweeness Centrality |
| A | (1+1+0+0)/2 = 1 |
| B | 0 |
| C | 0 |
| D | (1+2+2+3)/4 = 2 |
| E | 0 |
From the above founding, David is the most influential, since it got highest score in both Degree / Closeness / Betweeness Centrality.
2c)
From the result of Degree / Closeness / Betweeness Centrality, David is most influential in terms of number of direct connections (Degree), average distance (Closeness) and number of shortest path that it was relied on (Betweeness).
This show that David is most likely the 'Key' player in the social network.
沒有留言:
張貼留言