Once a sentence containing “I feel” or “I am feeling” is found, the system looks backward to the beginning of the sentence, and forward to the end of the sentence, and then saves the full sentence in a database.
Once saved, the sentence is scanned to see if it includes one of about 5,000 pre-identified “feelings”. This list of valid feelings was constructed by hand, but basically consists of adjectives and some adverbs. The full list of valid feelings, along with the total count of each feeling, and the color assigned to each feeling, is here.
If a valid feeling is found, the sentence is said to represent one person who feels that way.
More information on the methods can be found here, or you can just watch the visualizations themselves. The project is a collaboration between artist Jonathan Harris and Sep Kamvar, who works for Google and is a computational mathematician.
Increasingly, I find students want to do studies that depend on collecting online data. The problem with many of these proposals (in my experience, anyway) is that the students often think that simply doing some internet-related sociology automatically makes you cutting-edge, and that any old trawl of myspace pages or what have you amounts to a radically interesting fusion of network theory and Goffmanian approaches to the presentation of self. But work in this area has been going on for well over a decade now, and the kind of stuff that We Feel Fine is doing—which does not even present itself as sociology—suggests where the cutting edge really is in terms of purely web-driven data collection and visualization of the structure and content of online networks.