Sunday, February 7, 2010

Nice package of software an tutorials for corpus analysis

I was about to try to provide some help in using the \(G^2\) implementations that I use and give away for a graduate student trying to do some corpus analysis and I found this lovely site describing how to do corpus analysis using R.   Marco Baroni and Stefan Evert have done a really lovely job there of describing how to do a large number of simple tasks.  Kudos to them!

For not necessarily the most noble reasons, I was happy to see from their slide show on computing association measures that \(G^2\) is still essentially on top as a measure of association after all of these years.

2 comments:

Sander Stepanov said...

web links to blog are not correct
may you pls find correct ones

Ted Dunning ... apparently Bayesian said...


Fixed the link. (simple google search found it)