Putting the science into speeches

Posted by Martin Stabe on 26 July 2007 at 16:04
Tags: visualisation

Political blogger Guido Fawkes this week boasted of having obtained a pirated PDF version of the Alastair Campbell Diaries.

The nice thing about having PDFs of large documents is that unlike reams of paper, they are searchable. This allowed the blogger’s “co-conspirators” to count 135 references to the word “fuck”.

A more systematic content analysis would make far more interesting journalism, of course. Using Many Eyes, a free online data visualisation tool developed by IBM, it is possible to create a heatmap that graphically reveals the frequency distribution of words or phrases in a chunk of text.

The tool was demonstrated by New York Times database editor Aron Pilhofer at the Investigative Journalism
Summer School held by the Centre for Investigative Journalism at City University last weekend.

It’s particularly useful for analysing political documents because it quickly reveals the key messages being repeated throughout a speech. Rhetorical spin is rendered instantly visible.

City’s head of journalism, Adrian Monck, tried out the tool using the text of Gordon Brown’s speeches in Parliament since becoming Prime Minister. They reveal that Brown’s favourite two-word phrases include “British people”, “rising aspirations”, “young people”, “national security”, “affordable housing” and, thanks to the no-doubt riveting Planning Gain Supplement Bill, “planning gain”.

Now, if only that Alastair Campbell PDF were legally obtainable…