nltk中meteor_score的计算，报错

懒得在介绍来龙去脉了，反正就是找到的代码全是这种：

import nltk

hypothesis = ' '.join(['It', 'is', 'a', 'cat', 'at', 'room'])
reference = ' '.join(['It', 'is', 'a', 'cat', 'inside', 'the', 'room'])
#there may be several references
merteor_score = nltk.translate.meteor_score.single_meteor_score(reference, hypothesis)
print(merteor_score)

简单来说就是hypothesis和reference都是字符串，然后我就一直报这个错：TypeError: "hypothesis" expects pre-tokenized hypothesis (Iterable[str]): It is a cat at room。看起来是类型问题，然后找了一通，全都是用的字符串，最后没办法了看了眼源码，说要求是{hypothesis}格式，我觉得应该是set类型吧，然后就把

hypothesis和reference改成了set，然后就可以了，就emmm，我还费劲巴拉改一晚上，难道是nltk版本更新所以数据格式换了？
改后代码：

import nltk
# from nltk.corpus import wordnet
# nltk.download('wordnet')
hypothesis = ['It', 'is', 'a', 'cat', 'at', 'room']
reference = ['It', 'is', 'a', 'cat', 'inside', 'the', 'room']
hypothesis=set(hypothesis)
reference=set(reference)
#there may be several references
merteor_score = nltk.translate.meteor_score.single_meteor_score(reference, hypothesis)
print(merteor_score)

标签：room,reference,score,hypothesis,meteor,报错,nltk
来源： https://www.cnblogs.com/space130/p/15937987.html