In 2010 we began the DataArt project with the BBC exploring different ways of organising, visualising and mining their data and programme archives. Some of the most interesting work coming from this collaboration involved developing visualisations of historic trends in news coverage employing both news archives from the BBC and the Guardian.
Visualisation promises to organise disparate data into rationale, smooth surfaces promising to make complexity fleetingly graspable, if not entirely explicable. For example, launching NewsTraces enables us to search the online news coverage published by the BBC and Guardian newspaper over a 10-year period. By searching this news archive, fascinating topical strata are generated and animate across the screen pointing to the frequency of certain news foci and by implication their importance. Similarly, with NewsCloud we can track, trace and intuit key emerging themes, and scientific discoveries around environmental change through the prevalence of news coverage of it.
What is interesting about these machines however is the tension between the functional rhetoric of visualisation they inherit and the hidden life of data that they grid up, slice and dice, and neatly organise. Or to put it another way, somewhere between the promise to help us navigate the suffocating overabundance of news information and the messy realities of human failure, hubris and overweening ambition that they point to, is revealed a different set of possibilities one of which forces us to consider what these visualisations tell us of the priorities, desires and inclinations of both the producers of news and its consumers.
So as NewsTraces and other works featured on this page show, information visualisation won't necessarily help us tame the beast of information overload, but as affect mining engines, are remarkably effective at baring witnesses to and mirroring a desiring public back to itself.