Mapping Online Publics: Researching the Uses of Twitter


Published on

Public lecture presented at the University of Amsterdam, 1 September 2011.

Published in: Education, Technology
  • Be the first to comment

  • Be the first to like this

No Downloads
Total views
On SlideShare
From Embeds
Number of Embeds
Embeds 0
No embeds

No notes for slide

Mapping Online Publics: Researching the Uses of Twitter

  1. 1. Mapping Online Publics:Researching the Uses of Twitter<br />Assoc. Prof. Axel Bruns<br />ARC Centre of Excellence for Creative Industries and Innovation<br />Queensland University of Technology, Brisbane, Australia<br /> – – @snurb_dot_info<br /><br />
  2. 2. Why Twitter?<br />Researching Twitter:<br />Significant world-wide social network<br />~200 million users (but how many active?)<br />Varied range of uses: from phatic communication to emergency coordination<br />Healthy third-party ecosystem (for now)<br />Strong history of user innovation: @replies, #hashtags<br />Flat and open network structure: non-reciprocal following, public profiles by default<br />Good API for gathering data for research<br />
  3. 3. (Google Maps)<br />
  4. 4. The 2011 Queensland Floods<br />Chronology:<br />December 2010 to January 2011: unprecedented rainfall<br />Emergency declared for more than 50% of Queensland<br />Wivenhoe dam reaches 180% capacity<br />December 2010: Flooding in northern Queensland<br />January 2011: Floods in southeast Queensland<br />10 January 2011: flash flooding in Toowoomba<br />10 January 2011: ‘inland tsunami’ in the Lockyer Valley<br />11 January 2011: flooding begins in Ipswich<br />12-16 January 2011: major flooding in Brisbane<br />
  5. 5. (Google Maps)<br />
  6. 6. (ABC News)<br />
  7. 7. (ABC News)<br />
  8. 8. (<br />
  9. 9. (ABC News)<br />
  10. 10. (ABC News)<br />
  11. 11. (ABC News)<br />
  12. 12. Social Media during the Floods<br />Various platforms:<br />Facebook, Twitter – updates and information<br />YouTube, Flickr, Twitpic – first-hand video and photos<br />Google Maps, Ushahidi – map-based information mashups<br /><ul><li>Different tools for different purposes</li></ul>Various levels of maturity:<br />Uses and use practices still developing<br />Different demographic reach<br />Technological differences:<br />e.g. Facebook: built around personal networks; semi-private; discussion threads<br />e.g. Twitter: open, flat network; public #hashtag conversations; update stream<br />
  13. 13. #qldfloods Tweets<br />10 Jan 2011 11 Jan 2011 12 Jan 2011 13 Jan 2011 14 Jan 2011 15 Jan 2011<br />
  14. 14. 10 Jan 2011 11 Jan 2011 12 Jan 2011 13 Jan 2011 14 Jan 2011 15 Jan 2011<br />#qldfloods from Toowoomba to Brisbane<br />
  15. 15. #qldfloods @replies<br />authorities<br />mainstream media<br />
  16. 16. @QPSmedia as Central #qldfloods Information Source<br />
  17. 17. #qldfloods Network Map – Most Active Accounts Only(Degree >= 15 / Node size: indegree / node colour: outdegree)<br />
  18. 18. The Queensland Floods Community<br />Self-organisation:<br />Rapid establishment of #qldfloodshashtag<br />Ad hoc development of community structures<br />Highlighting of leading accounts, vigilant against disruption<br />Suspension of petty squabbles (e.g. state politics)<br />Innovation and rapid prototyping:<br />Adjunct hashtags (#Mythbuster, #bakedrelief)<br />Sharing and gathering of online resources<br />Additional tools (Google Maps, Ushahidi Maps)<br />Emergency services rapidly adopting social media tools (despite lack of established strategies)<br /> ‘Go where they are’ rather than ‘build it and they will come’<br />
  19. 19. Image by Maproom Systems<br />
  20. 20. New Media and Public Communication: Mapping Australian User-Created Contentin Online Social Networks<br />Australian Research Council (ARC) Discovery Project (2010-13) – $410,000<br />QUT (Brisbane), Sociomantic Labs (Berlin)<br />First comprehensive study of Australian social media use<br />Computer-assisted cultural analysis: tracking, mapping, analysing blogs, Twitter, Flickr, YouTubeas ‘networked publics’<br />Builds on previous work of the research team (UCC, YouTube, blogosphere mapping)<br />Advances beyond established approaches - beyond political blogospheres, beyond snapshots<br />Addressing the problem of scale (‘Big Data’) and disciplinary change in media, cultural and communication studies– natively digital methods<br />
  21. 21. Data Gathering<br />yourTwapperkeeper+ in-house crawler<br />Data Processing<br />Gawk – open source, multiplatform, programmable command-line tool for processing CSV documents<br />Textual Analysis<br />Leximancer – commercial (University of Queensland), multiplatform: extracts key concepts from large corpora of text, examines and visualises concept co-occurrence<br />WordStat – commercial, PC-only text analysis tool; generates concept co-occurrence data that can be exported for visualisation<br />Visualisation<br />Gephi – open source, multiplatform network visualisation tool<br />Tools<br />
  22. 22. Analysis – Twapperkeeper (#hashtags)<br />
  23. 23. But Why?<br />Possible research questions:<br />Ad hoc events and publics:<br />How do online publics form and dissolve? How do they interact, what structures do they form?<br />Where do they draw information from? What do they share?<br />Do they simply consist of the usual suspects? How insular and disconnected are online publics?<br />Hashtags in context:<br />How do different hashtag events compare? Are there common types of hashtags/publics?<br />How ‘big’ are they? What topics attract attention on Twitter?<br />What community (?) structures emerge?<br />
  24. 24. #royalwedding<br />
  25. 25. Hashtag Publics, Hashtag Metrics<br />How big is the central core of users?<br />Long tail distribution: most active users responsible for the majority of content<br />90/9/1 rule: how much does the top 1% of users contribute?<br />#royalwedding: ~10% of all tweets<br />#qldfloods: ~17% of all tweets<br />#libya: ~49% of all tweets<br />What do they do: inform, share, chat?<br />How many links to they share?<br />How much retweeting do they do (edited/unedited)?<br />How many @replies do they send / receive?<br />… etc.<br />
  26. 26. Tweets by Top 1% of Contributors<br />
  27. 27. Distinguishing Apples and Oranges<br />(top 1% most active users in each hashtag; size = percentage of total tweet volume)<br />Top 1% most active users<br />
  28. 28. Distinguishing Apples and Oranges<br />All contributors<br />
  29. 29. Towards a Typology of Twitter Uses<br />How are hashtags used (during acute events)?<br />Gatewatching: <br />Finding and sharing information about breaking news (before the mainstream media do?)<br />Ad hoc publics: many URLs, many retweets(even unedited)<br />Audiencing:<br />Shared experience of major (foreseen) events<br />Imagined community of fellow participants: few URLs, limited retweeting<br />What other uses are there?<br />Continuing discussions (#auspol, #bundesliga, …)<br />Memes (#ghettohurricanenames, …)<br />Emotive hashtags (#fail, #win, #headdesk, …)<br />What about keywords?<br />
  30. 30. Beyond Hashtags<br />Publics on Twitter:<br />Micro: @reply and retweet conversations<br />Meso: hashtag ‘communities’<br />Macro: follower/followee networks<br /> Multiple overlapping publics / networks<br />What drives their formation and dissipation?<br />How do they interact and interweave?<br />How are they interleaved with the wider media ecology?<br />Twitter doesn’t contain publics: publics transcend Twitter<br />
  31. 31. Understanding Australian Twitter Use<br />What is the Australian Twitteruserbase?<br />Large-scale snowballing project<br />Starting from selected hashtag communities (e.g. #ausvotes, #qldfloods, #masterchef)<br />Identifying participating users, testing for ‘Australianness’:<br />Timezone setting, location information, profile information<br />Retrieving follower/followee information for each account (very slow)<br />Progress update:<br />~550,000 Australian users identified so far<br />
  32. 32. The Australian Twittersphere<br />Follower/followee network:~150,000 Australian Twitter users(of ~550,000 known accounts so far) colour = outdegree, size = indegree<br />
  33. 33. The Australian Twittersphere<br />
  34. 34.<br />@snurb_dot_info<br />@jeanburgess<br />