You are here

Thinking through Approaches to Mapping Blog Networks

Gothenburg.
The final speaker in our social media mapping session at AoIR 2010 is my excellent PhD student Tim Highfield, whose focus is on comparing the French and Australian political blogospheres. Here, he’s examining blog network mapping, which enables an investigation of links, affiliations, friendships, clusters, references, and oppositions between blogs; this can also easily lead to simply pretty visualisations which ultimately don’t tell us much, however.

Strengths are that larger and longer-term datasets can be created, and dominant groups can be identified over time – however, many studies still focus on all links on a page, rather than only on the discursive links in blog posts and/or the static affiliations in blogrolls. for example. Additionally, it would also be used to distinguish supportive and oppositional links, and to weight repeated links more strongly than less frequent interlinkage.

Tim’s study covers the period of January to August 2009, tracking a set of French and Australian political blogs; the time-based nature of his data enables him to look at specific periods in this overall timeframe, and compare activities there. In the first place, he’s also able to compare linkage patterns in blogrolls and in-post citations. Blogrolls, for example, show that in the French blogosphere, there is only a specific group of blogs which frequently link to mainstream media in their blogrolls; those MSM sites are more central for link networks based on links in the posts themselves.

During specific times during the overall period, there are also substantial variations in linking activity. During the inauguration of Barack Obama, for example, a number of mainstream media are linked to, as well as video sites Dailymotion and YouTube, and a specific subset of blogs are especially active. When combining this with content analysis of the posts, it becomes possible to see what the key themes of these sites were – and while some blogs did indeed cover the Obama inauguration during this period, others continued to focus mainly on French domestic politics; Obama was not able to replace French president Nicolas Sarkozy as the most cited politician in those sites.

It would also be interesting to represent these data more geographically, for example to identify where those bloggers are based, and whether rural and regional topics are therefore addressed at all. Further, the hyperlink analysis performed here would also be suitably complemented by user traffic analysis.