Social network analysis (SNA) & applications

Social media tools such as email, discussion forums, blogs, micro-blogs, and wikis are used by billions of people worldwide. As they communicate through these media via desktop and web-based applications on fixed and mobile devices the result is the creation of multiple complex social network structures. The lively interaction and networks of relationships created through these technologies is of growing importance to individuals, organizations, and communities. Understanding how these social media networks grow, change, fail, or succeed is a growing concern to researchers and professionals. The field of social network analysis provides a set of concepts and metrics to systematically study these dynamic processes. The methods of information visualization have also become valuable in helping users to discover patterns, trends, clusters, and outliers, even in complex social networks.

These days, in our class, we learned social network analysis (SNA), the useful network analyzing method. It is the study of the pattern of interaction between actors. In SNA, every activator is present as a node (vertex) and the action between each other is shown by lines (edges) connected to nodes. With these vertices and edges, we can generate some social graph from which we can clearly find the relationship of everyone there.

To analyze a network, several software could be applied, such as UCINET, Pajek, NetMiner, Multinet, Stocnet, Strucuture, NodeXL, ect.

This is one of social grams I made using software NodeXL. The data is collected by Oct 29, and it just shows people in our class.

Social graph can be classified to undirected graph and directed graph. Undirected graph just displays the connecting condition and directed graph can show the direction of each information flow.

Some important indexes could be obtained from SNA.

Degree( degree centrality)-----analyzing in undirected graph. It’s a count of the number of edges that are connected to it. If the edges represented strong friendship ties of individuals in a class, we might say that XX is the most popular person. In directed graph, we analyze in-degree and out-degree, which respectively show the number of edges that point toward the vertex of interest or the vertex of interest points towards.

The following graph is a sub graph of Su Jing, who is the one with the biggest degree(the most popular one) by Oct 29. Generating subgragh image is a useful way to understand complex networks is to view individual sections of the larger graph. It represents the information exchanging condition on a certain individual in the organization.

If we say degree is about popularity, then betweenness centrality is about if the vertex is important in connecting the whole system. Vertices that are included in many of the shortest path between other vertices have a higher Betweenness Centrality than those that are not included.

Closeness centrality presents how close each person is to others in the network. It’s a measure of the average shortest distance from each vertex to each other. A lower closeness centrality score indicates a more central position in the network.

The parameters above are all important when dealing with individuals in the network. If we put the focus on the whole system, we should discuss the density and clustering coefficient…

Nowadays, SNA is widely used in fields such as business link, emergency service, academic collaboration, ect. The methods of information visualization have also become valuable in helping users to discover patterns, trends, clusters, and outliers, even in complex social networks.