The Mathematics of Social Networking

In essence a social network is made up by an number of points all connected to each other by a specific relationship.  They could be members of a certain group, be ’friends’ or have sent someone a message or ’like’.  The units don’t have to be just people though, they can be places, groups, companies or even photos or articles.

The main issues with social networks can be quite easy to define but when you try and assess them things get a little more confusing.  For a company like Facebook understanding these relationships is core to their ideal of bringing people together (and making money out of them!).  Trying to work out which individuals can be connected is a key aim and to understand it you  have to ask some basic questions.

  • Who is connected to who else?
  • How Can you Connect Sub Groups of People
  • Are Some Groups of People better connected than others.

It can be summaried into defining communities and understanding their relationships.  So how can we use maths to understand and help answer these questions?  Well the easiest way is to graph the various networks – this can be achieved by partitioning a social network into specific groups and sub networks,.

The subject can become quite complicated but if you’re interested in a good starting point try Moses Boudourides, search online for “An Introduction to Community Detection in Graphs”

He helps explain some excellent techniques and approaches to help breakdown the networks into more manageable subsets. This helps the overall analysis of the different nodes and their relationships between each other.  For a primer into the SEO science of social networking then this article comes recommended – How to Use Social Networking in SEO.  Although the analysis of social networks may seem quite confusing at first, in reality they offer a fascinating way to study relationships and social groups interacting with each other.

Henry Haversham

Author of British VPN

facebooktwittergoogle_plusredditpinterestlinkedinmailby feather

Tags: