Hello, the ngram package can be used to compute the bigram counts, but there is no language model support.
Jörn On 10/02/2012 02:41 AM, Adam Goodkind wrote:
Hi, I would like to write a program that tests the likelihood of a given program, against a "generic" language model. I realize that "generic" is a dangerous term: We'll be using data from American college students, if that helps to better define the sort of model we'd like to test against. Does OpenNLP offer anything like this? Thanks, Adam
