World Trade network in 2000. Coolurs are associated to the different economic world regions considered (for example in clear colour the European Union). Source: Lothar Krempel's NetVis  Gallery. Click on the image to enlarge it.

In our issue number 113 we spoke about visualisation of social interaction and we also outlined four basic elements of representation:

• Presence, the people participating in a particular event.
  
  
• Identity or affiliation of the participants.
  
  
• Interacton between the participants.
  
  
• Communication that takes place among them.

In this issue we will focus on the third point: social interaction and the networks that said interaction creates. Social networks are important (for our purposes) for different reasons, among which we can count:

• We all belong to one or more of them.
  
  
• Social networks are vehicles of influence and power in the organisations.
  
  
• The inherent properties of networks determine their possibilities. (There are things you can do in a network and things you can’t)
  
  
• The advent of Internet, the network of networks, has favoured the appearance of thousands of social networks, with many members each.

For this reason social network visualisation has begun to receive a certain amount of attention.

One important point of networks –and social networks are just one more among the many possible networks-- is that there’s a mathematical theory, the graph theory, that allows you to study the behaviour of the networks and to know their properties. 

As its own name implies, graph theory has in its essence a component of a graphical nature, of easy representation. And, better yet, its implications are relatively easy to understand just by looking at a graph. All this makes it ideal as a visual metaphor for information visualisation of networked systems, a reason why most of the visualisations about social networks use this paradigm.

Capital relations in Germany in 1996 and 2000. Colours indicate the nature of the same. In beige you find fianancial entities (banks and insurance companies). In red industrial companies. Beige arrows indicate shares between financial institutions, the red ones link industrial entities. The remaining show crossed relationships (financial-industrial). Source: Lothar Krempel's NetVis  Gallery. Click on the image to enlarge it.

For example, Lothar Krempel from the Max-Planck-Institut für Gesellschaftsforschung (Max-Planck-Institute for the Study of Socities) in Germany, maintains a gallery of suggestive visualisations. Of particular interest is the visualisation of the capital ties existing among the 100 biggest German companies in the years 1996 and 2000. The plot shows a closely knit pattern of relations between banks, insurance companies and industrial companies where Deutsche Bank is a predominant node. 

Social Network Visualisation: e-mail between the members of a project. Every square depicts a participant. Its colour indicates which departments he/she belongs to. Grey lines indicate that there's at least a weekly frequency between the two linked participants. Source: Orgnet.com, as can be seen in its web page about social networks and e-mail. Click on the image to enlarge it.

Another interesting example of social network interaction can be seen in one of the pages of Orgnet.com, that used its “InFlow” software in order to visualise the e-mail links (disregarding its contents) among the participants of a project, a short while after failing to meet the deadline of the 4th month of the project. 

In the representation every participant is coded with a colour in accordance with the department he/she belongs to. The grey lines that link two people indicate exchanging e-mails on at least a weekly basis. 

From the analysis of the social network it can be easily seen that the relationship between departments was very limited and the links between their corresponding members were circumscribed to some people that acted as “bridges” between different departments. The solution that was adopted to overcome the situation was to physically co-locate people of different departments but related assignments.

Social network analysis reveals, in many cases, the structure of informal leadership (outside of the established organisation chart) that is usually led by experts in certain fields.

Building and visualising a graph of the social network can be an important decision factor. In all the networks there are nodes that accumulate links while other nodes are weakly linked to the others. Obviously the most outstanding nodes, the ones with most of the links, are also the most influential nodes. It isn’t always simple nor, above all, quick to detect them.

In order to do this it’s necessary to have available information about:

• which nodes do exist
  
  
• how are they interconnected.

With this information we can build a graph, represent it and study the indicators that can allow us to assess the network structure, discovering which nodes play a central role and which don’t.

Social network visualisation is in its beginnings, but promises an insight on the social structure of organisation and in particular on those of companies, most of them still organised in a hierarchical way that, nevertheless, appears to be starting to change, adopting some of the elements of the networks.

In the next issue we will speak about the power that graph representation has and the problems that appear when doing it.

In the INSNA (International Network for Social Network Analysis) web site we can find a long list list of software with programs for social network analysis. Those interested in the topic speaking Spanish can also consult the magazine Redes.

Links of this issue: