The Australian Twittersphere in 2016: Mapping the Follower/Followee Network

Twitter is now a key platform for public communication between a diverse range of participants, but the overall shape of the communication network it provides remains largely unknown. This article provides a detailed overview of the network structure of the Australian Twittersphere and identifies the thematic drivers of the key clusters within the network. We identify some 3.72 million Australian Twitter accounts and map the follower/followee connections between the 255,000 most connected accounts; we utilize community detection algorithms to identify the major clusters within this network and examine their account populations to identify their constitutive themes; we examine account creation dates and reconstruct a timeline for the Twitter adoption process among different communities; and we examine lifetime and recent tweeting patterns to determine the historically and currently most active clusters in the network. In combination, this offers the first rigorous and comprehensive study of the network structure of an entire national Twittersphere.

A summary of some of the study’s key findings was published in The Conversation in May 2017. Meanwhile, a paper by Axel Bruns at the Future of Journalism conference in Cardiff in September 2017 built on this new Twittersphere map to test for the existence of echo chambers and filter bubbles in Australian Twitter – and found little evidence to support the thesis:

Echo Chamber? What Echo Chamber? Reviewing the Evidence

The success of political movements that appear to be immune to any factual evidence that contradicts their claims – from the Brexiteers to the ‘alt-right’, neo-fascist groups supporting Donald Trump – has reinvigorated claims that social media spaces constitute so-called ‘filter bubbles’ or ‘echo chambers’. But while such claims may appear intuitively true to politicians and journalists – who have themselves been accused of living in filter bubbles –, the evidence that ordinary users experience their everyday social media environments as echo chambers is far more limited.

For instance, a 2016 Pew Center study has shown that only 23% of U.S. users on Facebook and 17% on Twitter now say with confidence that most of their contacts’ views are similar to their own. 20% have changed their minds about a political or social issue because of interactions on social media. Similarly, large-scale studies of follower and interaction networks on Twitter show that national Twitterspheres are often thoroughly interconnected and facilitate the flow of information across boundaries of personal ideology and interest, except for a few especially hardcore partisan communities.

Building on new, comprehensive data from a project that maps and tracks interactions between 4 million accounts in the Australian Twittersphere, this paper explores in detail the evidence for the existence of echo chambers in that country. It thereby moves the present debate beyond a merely anecdotal footing, and offers a more reliable assessment of the ‘echo chamber’ threat.

When did they join?

As you can see from the above graph, which shows Account creation date along the horizontal and ID along the vertical, a spattering of accounts were registered from late March to July 2006, with the first at approximately 20:50 on 21 March, ID#12. It is worth mentioning here that these are only accounts that are still active, and so it is impossible to access data for those which are no longer active. A slight increase in the registration rate of users occurred between November and December 2006, while late December ‘06 to early January ‘07 saw a sharp increase, corresponding to US publicity, which tapered off to a steadier rate until March ‘07 at which point we see a second publicity driven spike which took the IDs over the 1 million range. It is worth noting that of these 1 million IDs, only 48,546 accounts remain active.

Where in the world (is Twitter user x)?

This map contains some of the limitations discussed on Friday; namely that it is created by time zone, although interestingly does not show the same bias toward alphabetically prominent time zones such as Amsterdam as is present with new users. Also interesting here is the prominence of Italy, with about 4% of the active accounts coming from Italy, far higher than any non-English speaking country, and higher than Australia. The US is dominant, with well over 50%, however the neighbour to the north, Canada, accounts for only 0.3%, again a sharp contrast to the newly created data. Other hotbeds of early Twitter activity (those with over 1,000 of the 40,000 accounts) are limited to Australia and the United Kingdom.

What have they been up to since 2006?

The above chart shows total statuses posted, and shows some interesting patterns; whilst there are a few users with 400,000 plus tweets, the majority have managed to restrict themselves to 200,000 or less across the past 7 years, with the majority clustered below 50k. As we saw with the join times graph previously, there are a number of missing IDs (and thus 0 statuses) around the middle of the chart. The below chart, in which users are placed into ‘bins’ of 100,000 IDs again shows a fairly average status count among the users, suggesting that those early users whose account is still active (i.e. hasn’t been deleted) have tweeted more-or-less the same number of times as those joining during one of the publicity cycles.

Followers and Followees

Here, I have removed eight data points from the visualisation, which shows the numbers of followers by account ID; users with 29.3m (Barack Obama), 16.5m (Twitter themselves), 7.67million (New York Times), 7.5 million (CNN), 3.5 million (Starbucks), 3.2 million (BBC World), 3.17m (Mashable) and 2.6m (TechCrunch) followers. One random note from removing these is that Twitter themselves have an ID in the 783,000 range, while Starbucks are in the 31,000 range – clearly an early priority of the Twitter developers was not to create a corporate identity for themselves!

Again, we see a large number of IDs in the centre of the graph with little activity, replicating previous data, with two more populous clusters to either side. Here though, the later users show a marked increase in connectedness over those on the left side of the graph. The very early adopters (perhaps those who were in some way connected to a member of the development team), while tweeting regularly, may then be less connected than those tech aficionados who joined during the early phase of publicity.

The above diagram shows the number of accounts a user is following. Here, two accounts have been cut from this diagram for visualisation purposes; one following 665,279 users (Barack Obama) and the other 229,915. Otherwise, we see a fairly similar pattern as with statuses and followers, the missing IDs in the middle, with users to the left and right having fairly similar distributions to the followers graph above, re-enforcing the suggestion that later IDs seem to be more connected than the early users.

Overall then, an interesting distribution of early Twitter uses, which is in some ways similar and some ways different from the more recent users discussed on Friday. Now just to fill in the missing hundreds of million!