Found out about a interesting program called "We Feel Fine" from SEOBOOK - I wrote up a little bit about this in ArtNYC tonight. You might want to check it out.
I was able to use various metrics to slice the feelings of people, in mass, based on blog posts that are harvested by this program, which runs in an applet.
We Feel Fine has 6 styles of presenting the data called Movements. All are intersting, but Metrics is the one movement that really relates to Webmetricsguru. Here’s what Metrics looks like for We Feel Fine:
Metrics, the fifth movement, consists of five smaller movements. Whereas Mobs expresses the notion of “Most Common”, Metrics expresses the notion of “Most Salient”. A full discussion of the difference between “Most Common” and “Most Salient” is here. Essentially, the traits listed in Metrics are those that best distinguish the sample population from the global average (i.e. how is this population different?). Metrics displays the most representative traits of the sample population, along five axes: feeling, gender, age, weather, and location.
Metrics (Feeling) displays the most representative feelings of the sample population, along with some statistical data indicating the significance of the findings. The feelings themselves are listed along the left edge of the screen, ranked by the number of times their frequency in the sample population exceeds the global average. A large red circle holds this number. On the right side of the screen are a series of bar charts showing the number of times each feeling occurred in the sample population, and the number of times each feeling typically occurs.
Metrics (Gender) displays a comparison between men and women for the sample population, indicating whether either gender is particularly salient (i.e. do women feel happy more than men?). The labeling and graphing follow the same conventions of Metrics (Feeling).
Metrics (Age) displays the most representative ages (in ten year increments) of the sample population, indicating whether a given age range is particularly salient (i.e. do 40 year olds feel old more than most people?). The labeling and graphing follow the same conventions of Metrics (Feeling).
Metrics (Weather) displays a comparison between the four weather types (sunny, cloudy, rainy, snowy) for the sample population, indicating whether a given weather type is particularly salient (i.e. do people feel depressed more often when it’s rainy?). The labeling and graphing follow the same conventions of Metrics (Feeling).
Metrics (location) displays the most representative locations (countries, states, and cities) of the sample population, indicating whether a given location is particularly salient (i.e. do Canadians feel cold more than most people?). The labeling and graphing follow the same conventions of Metrics (Feeling).
I also found Mounds facinating, visually:
"Mounds
Mounds, the sixth and final movement, is independent of the sample population, always displaying every feeling in our database, scaled and sorted in order of frequency. Each feeling is portrayed as a large bulbous mound, colored to correspond to the feeling it represents. The mounds jiggle slightly when undisturbed, and bend away as the mouse cursor approaches their perimeter. Clicking the screen caused the mounds to jiggle wildly.
A small scrollbar below the mounds represents the entire database of feelings, and allows the viewer to jump to a specific point in the list. The viewer can also position the mouse near the left or right edge of the screen to cause the mounds list to self-scroll. Above each mound is listed its feeling, along with the rank of that feeling, and the total number of occurrences of that feeling contained in our database. Clicking the feeling above a mound retrieves feelings from people who feel that way.
The Panel
The Panel allows the viewer to control the sample population on screen at any one time. At all times, the red bar atop the screen presents a concise summary of the current sample population. Clicking that red bar causes the panel to open, and within the panel, viewers can constrain the population along any combination of the following axes:
- Feeling (happy, sad, depressed, etc.)
- Age (in ten year increments - 20s, 30s, etc.)
- Gender (male or female)
- Weather (sunny, cloudy, rainy, or snowy)
- Location (country, state, and/or city)
- Date (year, month, and/or day)
When satisfied, the viewer can press “Find Feelings” and We Feel Fine will retrieve any matching feelings from our database. Because of web browser limitations (the applet is highly CPU intensive), we limit the maximum number of feelings returned at any one time to 1,500 (on PCs) and 1,000 (on Macs). This means that searching for a very specific population (e.g. males from Afghanistan in their 20s when it is cloudy), will yield few or no feelings, and the feelings that are returned will likely stretch back several weeks or months in time. On the other hand, searching for a more general population (e.g. people who feel sad), will likely return the maximum number of 1,000 or 1,500 feelings, and those feelings will likely stretch back only hours or days.
I’m going to play a little more with this program in the coming days and if there’s anything that I find relevant, from a metrics perspective, I’ll write it up here.
