I have a project I'm using to map out the propaganda matrix using some subject extraction and contextual mapping techniques I developed. It's based off of words or phrases and how many times they are used with each other in the same articles. I'm not going to go into detail about all of it, but I'll just say this. Looking at my graph I knew that obama was going to win because he was more contextually "close" to the word "president" in the major media sources than Mccain.
I don't know how many of you have read that book "A Brave New World" by Aldous Huxley? Basically, they did brainwashing in that book through tape-recorders in pillows which was a crude metaphor for the sleep-like state you go into when you're watching TV... and they keep just throwing these phrases at you together in certain sequence and repetitiveness. If you know anything about how the mind works, one of the most simple forms of long term memory commitment is to repeat the input you want to go in. And your memory is like a matrix of interconnected thoughts. They are connected by the details in those thoughts that are similar to the other thoughts. This is what I am mapping out... the contextual map of subjects, but it's not your brains thoughts, it's the media that a lot of us as citizens of our countries read in their news media and thus get brainwashed with the rest of their fellow citizens.
The link to the graph is at http://www.newsbubbles.org/graph.php?q=president&p=mccain:obama:bush:debate:election&start=2007-08-01&end=2008-12-20&unit=month
To explain what you're seeing... the different plots represent each of the subjects in the legend. The higher a subject is on the graph, the "closer" it is in context to the context word... in this graph the context word is "president" ... and as you can see, obama was just more related to the word president for the entire time they were even within proximity of the word. I included the plots for the "debate" bubble (subject), (which got a little bit of earlier coverage before it's spike a month before the election), and "election" which spikes in the weeek of the election, for some sort of visual reference of time. Also notice how "bush", also added for contextual orientation, was always really high up on the list each week, but the month of the election, obama soars, and overtakes bush... in the context of president.
There are many more interesting graphs I've done... one for the context of the word "crisis" shows that it has only recently... like in August, gained any real media coverage, because it is one of the new hot words just like "bailout"... you can use the selectors on the right to plot other words in the context of president to see what other things are more popularly related or closer to the word over whatever time period you want.