You are here

Discussion Networks in the French Blogosphere

The Friday at ICA 2010 starts with the first of two panels on online network mapping (I'll be presenting in the second one, later today). My brilliant PhD student Tim Highfield is the first presenter. His interest is in topical discussion networks in the French political blogosphere: such topical networks comprise sites commenting on specific events or issues, and the links between them. This observation comes out of a larger dataset collected over a longer period of time.

Such networks can be found on Twitter as well - around the #spill keyword, tracking the change of leadership in Australia over the last few days, as well as around otehr keywords; in blogs, keywords and links can help to identify communities. (On Twitter, for example, Tim has identified a network of reweets of a satirical video mocking Perth by users based in or related to that city.)

Tim's project examines French and Australian political blogging, using data from January to August 2009; we scrape the content of some 217 active sites in France, 148 of which are bona fide political blogs. The set includes over 20,000 blog posts, then. Tim examined this, first, by exploring the network of sites as identified by their blogrolls - more static links which identify explicit affiliation rather than signal day-to-day interaction, and he now shows a visualisation of this (with sites clustering more or less according to political affiliation - the far right is an almost separate cluster, as would be expected; alternative media sites also feature highly).

Additionally, Tim conducted a pilot study examining blog posts from ten days around the inauguration of Barack Obama in early 2009, and visualised the network of links in the 170 posts relating to this event - imptorantly also including mainstream media sites as well as video sharing sites such as YouTube and the French DailyMotion. Alternative media sites didn't feature strongly here. (Neither did the far right.)

So, this study advances our network mapping methodologies by distinguishing between different link types - what emerges from this is also a question of how much the in-text links specifically can be seen as markers of more permanent affiliation in the way that blogroll links can be (in other words, what's the lifespan of such discursive links?); additionally, how do we delimit the temporal boundaries of specific events? Further, what will also be possible here is to explore the dynamics of day-by-day linking - and of course, the link network study can also be augmented by further content analysis, and be applied to non-blog networks.

Technorati : , , , , , : , , , , ,