import nltk
text = "the little yellow dog barked at the cat."
sens = nltk.sent_tokenize(text)
words = [nltk.word_tokenize(sentence) for sentence in sens]
tags = [nltk.pos_tag(tokens) for tokens in words]
grammar = r"""
  NP: {<DT|PP\$>?<JJ>*<NN>}
      {<NNP>+}
"""
cp = nltk.RegexpParser(grammar)
result = cp.parse(tags[0])
print(result)

得到的结果是：
(S
  (NP the/DT little/JJ yellow/JJ dog/NN)
  barked/VBD
  at/IN
  (NP the/DT cat/NN)
  ./.)

这个结果对我来说没有用。我想得到的是['the little yellow dog','barked','at',' the cat'].因为我相做分块的词频分析。得到的Tree怎么可以转换？或者有没有别的办法实现？