关于词性标注的最动态的语料库是树库语料库。然而,Brown Corpus Just 拒绝使用 HMM 和 TnT 标记器产生结果。对此有何解释?
size = int(len(brown.tagged_sents())*0.9)
train = brown.tagged_sents()[:size]
test = brown.tagged_sents()[size:]
trainer = hmm.HiddenMarkovModelTrainer()
tagger = trainer.train_supervised(train)
print(tagger.evaluate(test))
tnt_tagger = tnt.TnT()
tnt_tagger.train(train)
print(tnt_tagger.evaluate(test))