ANZCA09

description

Inauguration from afar
Mapping Obama discussion in Australian political blogs,
13 - 27 January 2009
Tim Highfield
t.highfield@qut.edu.au
andthentheworld.wordpress.com
A preliminary study
First work and findings from data collected since the
start of January 2009

Forms part of a larger, ongoing project collecting blog
posts from Australia and France

Work in collaboration with Lars Kirchhoff and Thomas
Nicolai of Sociomantic Labs
Methods
First output from the larger project, acting as a test for
various approaches to studying blog activity
An initial list of 245 blogs, news websites, commentators, and feeds
crawled for several months (ongoing) from January 2009
From the resulting data collection, each post and story posted between 13-27 January 2009 extracted, concepts identified using Leximancer.
This particular study looks at post
content and activity only.

Future work will examine links
and other aspects of blogging in more detail
Inauguration of Barack Obama
20 January 2009
A major international political event, acting  as the central point in the period covered in this study:

Collection of blog posts from one week before
to one week after the inauguration
Inauguration provides a clear event around
which study can be positioned

Other, recent examples that may be used for further study
include the initial coverage of swine flu or the 'Utegate' incident.
What happened?
Above average posting on 20/01, but not extreme spike - nor the most active day
Limitations
Inauguration not as likely to have attracted
same volume of posts as election victory

Live coverage of inauguration not as likely as
election to have large-scale coverage
because of time difference

Conclusions and additional notes
Importance of international events to Australian political blogs -
Gaza, Obama - also smaller coverage of whaling, Afghanistan -
international events with some Australian involvement

Easier to identify key international events than Australian in this sample,
but may also be due to period analysed


Future work and development
Refinement of analysis and processes

Cleaning up data, revisiting population to evaluate activity, check feeds

Identification of further events/themes for similar analysis

Evaluation of posting patterns, weekly rhythms in blog activity

Further analysis of inauguration data/other case studies around links, blog-blog, blog-mainstream media relationships
What was discussed?
Leximancer used to identify concepts, key words and names in the collected posts

Goes through several stages of analysis, allowing filtering of results by excluding concepts and merge variations on terms (e.g. Obama and Mr Obama/Barack Obama/President-elect Barack Obama)


MSM concept map
Blogs concept map
Discussion themes


Obama --> African-american, inauguration, security...

Coverage focussing on the event itself
(e.g. security surrounding the inauguration)
as well as the work of the president(-elect)


In general used the inauguration as means of
looking at economic response to GFC, policy and
strategy (environment, technology, Guantanamo Bay),
or to analyse Obama's speech itself...
Examples:
Larvatus Prodeo
Online Opinion
Andrew Bartlett
Crikey
Despite the population containing
more than 200 blogs and websites,

only around 50 blogs published at least
one post during the two week period
MSM posts per day
However, the inauguration is by no means the major focus of the period studied here - related terms and names only form small clusters of the concept maps featured here
Moving away from identifying blog networks over lengthy periods to studying interactions between sites around events or themes

Seeing how the structure of networks may change with topics and events as well as over time
In this analysis, kill concepts included writes,
post, update, publish.

Both Australia and Australian also removed
from the concept list.
Mainstream media sites/news feeds:
Blog posts (and commentary sites):
Does not take into account twitter coverage,
nor liveblogging using tools such as CoverItLive

Refinement needed of method,
cleaning up data and source lists