What can we learn from South Africa’s most influential tweeters?

So far in our series on the South African Twittersphere, we explored what lessons we can learn from conversations happening around two key political figures, opposition leader Helen Zille and president Jacob Zuma. We then mapped a large portion of the South African Twitter network in order to identify who had the most followers and the sub-communities that exist within the network.

While it is useful to know who has the most followers, this is not necessarily the best way of understanding who truly wields influence. To get a better grasp on who is influential within the network, our last article in this series focuses on two statistics, both of which are widely considered as good proxies for understanding influence based on a person’s position within a network:

• Does the person connect different groups together i.e. “betweeness centrality”? A person might not have many followers but if they are the only person that connects two different groups together that might not otherwise interact, then that person acts as a kind of conduit for the flow of information and, thus, influence.
• Do other important people think that a person is important a.k.a. the Authority metric from the HITS algorithm (a precursor to Google’s PageRank)? Who do the people that already have many followers all follow? We recognise these people as uniquely authoritative.

Using our two measures mentioned above, we created a network that sizes people’s nodes based on their scores on each of our two metrics.

People like ad man Khaya Dlanga, actor and comedian Rob van Vuuren, political strategist Ryan Coetzee, journalist Mandy Wiener and publisher Matthew Buckland clearly bridge different communities and act as conduits for the information that flows through their communities.

Helen Zille, Jacob Zuma, weekly newspaper The Mail & Guardian, City Press editor Ferial Haffajee, parliamentary opposition leader Lindiwe Mazibuko and Mandy Wiener are viewed as authoritative people by people in the know:

We can further explore the issue of influence by looking at who the top ten individuals are in terms of betweeness centrality and Authority within each of our sub-communities.

These are the information conduits; the people that knit disparate communities together (betweeness centrality):

These are the people that others in the know think are important and worth listening to (HITS Authority):

It is important to remember that the metrics we have looked at above only rank people in terms of influence based on their position within the network. We still don’t know what they are actually saying to each other and whether people are acting on their statements and advice. To know this requires a deeper dive into the data that is beyond the scope of this current research. Regardless, the results are interesting and should make intuitive sense to most readers.

This final set of analyses bring this series of articles to an end. In it we have briefly explored and characterized the South African Twittersphere. Twitter’s influence on our national debates continue to grow so I hope you agree that it is been fascinating to take a brief look under the skin of South Africa’s collective online consciousness. It is important to note that these analyses are not the final word on the the country’s Twittersphere.

Twitter is an organic beast that continues to grow and evolve and the snapshot described in these articles was accurate as of April 2012.

Also, while every attempt was made to collect as much information as possible in order to construct these networks, data will always be missed due to the nature of colleting these type of data. Regardless, I hope these articles have been interesting and informative.

Caveats:

Our definition of influence in this case only extends to how people are connected to each other in our network. We do not touch on who is directly influencing whom through their actions. This requires even deeper analysis, beyond the scope of this piece of research.