You are here

TrISMA (ARC LIEF)

ARC LIEF: TrISMA - Tracking Infrastructure for Social Media Analysis

Mapping a National Twittersphere: A ‘Big Data’ Analysis of Australian Twitter User Networks (ECREA 2014)

European Communication Conference (ECREA) 2014

Mapping a National Twittersphere: A ‘Big Data’ Analysis of Australian Twitter User Networks

Axel Bruns, Darryl Woodford, Troy Sadkowsky, and Tim Highfield

Twitter research to date has focussed mainly on the study of isolated events, as described for example by specific hashtags or keywords relating variously to elections (Larsson & Moe, 2012), natural disasters (Mendoza et al., 2010), entertainment (Highfield et al., 2013) and sporting events (Bruns et al., 2014), and other moments of heightened activity in the network. This limited focus is determined in part by the limitations placed on large-scale access to Twitter data by Twitter, Inc. itself. By contrast, only a handful of studies – usually by researchers associated with commercially funded research organisations or with Twitter, Inc. itself – have utilised the Twitter ‘firehose’ or similar more comprehensive sources of data to explore broader patterns of traffic flows or follower connections on the platform (e.g. Leetaru et al., 2013).

This project builds on a long-term, large-scale analysis of the global Twitter userbase which has managed to identify within the over 725 million global registered Twitter accounts some 2.5 million Australian accounts (by matching profile details such as location, description, and timezone against a set of relevant criteria). Further, we analysed the follower/followee connections of these 2.5 million accounts and from this developed a first comprehensive map of account relationships within the Australian Twittersphere. In-depth network analysis of this map reveals the existence of a range of clusters of especially tightly interconnected users, linked to each other by other accounts acting as bridges between the clusters. In turn, qualitative exploration of the leading account’s profiles in each cluster provides an indication of the various areas of thematic focus which have determined the formation of these clusters, and their association with other clusters in the same network vicinity. Further correlation with other relevant profile data (including the creation date for each account, its level of tweeting activity, and the date of the account’s last tweet) offers additional opportunities to trace the emergence and growing complexity of the Australian Twittersphere over time, from the earliest adopters of the platform to its most recent users, and to filter the overall network for the most active and most persistent users.

This study represents the first ever comprehensive investigation of the development of a national Twittersphere as an entity in its own right. While the global nature of Twitter as a social media platform means that Australian accounts will also be connected with their counterparts in other countries, it is still to be expected that shared interests and identity lead to the majority of connections between accounts to occur within the same national user population, and our analysis of these connection patterns provides an important indicator of the themes around which these connections crystallise, as well as of the longitudinal development of these clusters of interests.

Bruns, A., Weller, K., & Harrington, S. (2014). Twitter and Sports: Football Fandom in Emerging and Established Markets. In K. Weller, A. Bruns, J. Burgess, M. Mahrt, & C. Puschmann (Eds.), Twitter and Society (pp. 263–280). New York: Peter Lang.

Highfield, T., Harrington, S., & Bruns, A. (2013). Twitter as a Technology for Audiencing and Fandom: The #Eurovision Phenomenon. Information, Communication & Society, 16(3), 315–39. doi:10.1080/1369118X.2012.756053

Larsson, A.O., & Moe, H. (2011). Studying Political Microblogging: Twitter Users in the 2010 Swedish Election Campaign. New Media & Society, 14(5). doi:10.1177/1461444811422894

Leetaru, K., Wang, S., Cao, G., Padmanabhan, A., & Shook, E. (2013). Mapping the Global Twitter Heartbeat: The Geography of Twitter. First Monday, 18(5). doi:10.5210/fm.v18i5.4366

Mendoza, M., Poblete, B., & Castillo, C. (2010) Twitter under Crisis: Can We Trust What We RT? Paper presented at Social Media Analytics, KDD '10 Workshops, Washington, DC, 25 July 2010. Available from: http://research.yahoo.com/files/mendoza_poblete_castillo_2010_twitter_terremoto.pdf

The Global Demographics of Twitter

This final morning at AoIR 2015 opens with my paper with Darryl Woodford and Troy Sadkowsky which explores the global Twitter userbase. Our slides are below:

Coming Up Shortly

The annual end-of-year conference season is upon us again, and I’ll be heading off tomorrow to the annual Association of Internet Researchers conference – the most important conference in my field. In spite of the considerable troubles AoIR has faced this year – its first conference location, Bangkok, was no longer feasible following the military coup in Thailand, and there still seem to be some teething problems with the replacement location in Daegu, Korea – it will be great to catch up with leading colleagues in the field again.

This year, we’re presenting the first outcomes of our latest big data studies of the Australian and global Twitterspheres. One major paper will present what we’ve been able to glean so far of the overall patterns within the global Twitter userbase – we now have data on some 870 million Twitter profiles, which provides us with a unique perspective on how Twitter has grown and diversified as a platform. Further, we’ve also got a brand-new map of follower/followee  networks in the Australian Twittersphere, based on our dataset of some 2.8 million Australian users, and we’re using this to explore the footprint of recent television programming to shed new light on second-screen viewing practices, as part of our Telemetrics project (more on this at Darryl Woodford’s site). I’ll be live-blogging the conference again if I can get online, so expect to see much more over the next few days. As a preview, my slides for the two presentations are below.

Busy-ness as Usual

This blog has been somewhat slow again since the last round of conferences, and I'm hoping to do more in the future to change this. In the first place, I'm planning to post more regular updates again as I publish new articles and book chapters (watch out for a round-up of recent work soon, most of which already appear in my list of publications). There are also a number of new research projects which have started this year – and while more detailed updates about the day-to-day work of some of these will appear on Mapping Online Publics and the Website of the QUT Social Media Research Group, I'm planning to flag the most important outcomes from these projects here as well. And as always, updates are also available on Twitter through my own account @snurb_dot_info as well as @socialmediaQUT.

New Projects

Most importantly, I've just commenced my ARC Future Fellowship – a major four-year project which builds on my social media work and connects it with a number of other important data sources which shed light on the way Australian Internet users are engaging with news and current affairs. We'll continue to draw on large Twitter data (as well as, eventually, data from other social networks) which show the patterns of day-to-day activity around current events, and we'll correlate these patterns with data from Experian Hitwise, which track (anonymously and at very large scale) how Australian users search and browse the Web. Further, I'm also going to be able to incorporate some internal server data from Fairfax Digital (including its flagship mastheads Sydney Morning Herald and The Age) to investigate in more depth what articles users read and engage with on these sites.

Pages

Subscribe to RSS - TrISMA (ARC LIEF)