Its amazing to me how much interest there is in analyzing the huge body of work that is now available in blogs and other social media. I just heard a very interesting talk by Hugo Liu about Gender Modeling.
For example, consider this graph that shows the sorts of “time” words used by men and women. It seems that women write much more about the here-and-now, near future, and on a day-of-the-week scale. And men write more about the future, and on a month scale.
And this graph shows use of pronouns by gender. Top is female, bottom is male.
And perhaps coolest of all, they applied their algorithms to create male and female versions of Google news. You can see the demo here.
The full paper is available online and has many more interesting tidbits. All graphs in this post were taken from it.