Isn't Dekang Lin and Patrick Pantel's late-90's, early-00's stuff the best for this? Or their various recent innovations.

Isn't Dekang Lin and Patrick Pantel's late-90's, early-00's Lin's distributional similarity thesaurus-building stuff the best for this? Or their various recent innovations.From the late-90's.

If I had to bet, I'd bet they do this much better than the sorts of models currently popular in the machine learning community.

Isn't Dekang Lin's distributional similarity thesaurus-building stuff the best for this? From the late-90's.

If I had to bet, I'd bet they do this much better than the sorts of models currently popular in the machine learning community.community. Watching topic model people try to do this is kind of painful.

Isn't Dekang Lin's distributional similarity thesaurus-building stuff the best for this? From the late-90's.

If I had to bet, I'd bet they do this much better than the sorts of models currently popular in the machine learning community. Watching topic model people try to do this is kind of painful.painful. Or NN's, for that matter.