Acceso usuarios
Mike Thelwall, Iina Hellsten
Introduction. We assess the extent to which published media timelines reflect contemporary electronic discussions of major media events, using the London Attacks of July 2005 as a case study. The main objective is to judge whether timelines could, in principle, be automatically generated from contemporary debates. A secondary objective is to analyse the reasons for differences between contemporary debates and retrospective media timelines.
Method. Our method exploits the new opportunities for large-scale analysis afforded by electronic news feeds and blogs. We compared published media timelines with blog postings and news stories related to the London attacks of July, 2005. Rich Site Summary (RSS) technology was used to gather data from 19,587 blog and news sites. For the period of July 7 to July 31 we identified an average of 275 sources a day that posted new information containing the word 'London'. These postings, a combination of blogging and news stories, formed the raw data for our analysis. In particular, we extracted and counted the individual words used in the postings.
Analysis. We compared contemporary word usage statistics from our RSS data with the text of three published media timelines, particularly seeking words and types of words that were present in one but not the other.
Results. We found that the timelines tended to ignore the role of communication during the event itself, to use less emotionally charged language, and to gloss over to some extent the wider context of the event. Hence some aspects (frames) of the unfolding discussion were ignored by the timelines. Moreover, all sources included a range of relatively trivial details, the timelines apparently using them as concept markers (metonymy) for broader issues.
Conclusions. It seems that it would be difficult to automatically generate media timelines from contemporary discussions because of frame changes, and because of changes in terminology and the difficulty in separating highly discussed relatively insignificant details from the key facts. Nevertheless, the comparative analysis techniques reported in this paper may provide an interesting new window on contemporary discussions and media representations of major events.
