Question

我正在尝试使用NLTK的SklearnClassifier和BernoulliNB对数据进行分类。这是一些代码：

train = posFeatures[174:]+negFeatures[174:]+neuFeatures[174:]
devtest = posFeatures[124:174]+negFeatures[124:174]+neuFeatures[124:174]
test = posFeatures[:124]+negFeatures[:124]+neuFeatures[:124]
dev, tag_dev = zip(*devtest)
def score(classifier):
    classifier = SklearnClassifier(classifier)
    classifier.train(train)
    pred = classifier.batch_classify(dev)
    return accuracy_score(tag_dev,pred)
print('BernoulliNB  accuracy is %f'%score(BernoulliNB()) )

然后我收到了这个错误：

Traceback (most recent call last):
  File "G:/pycharm/quanbu.py", line 53, in <module>
    print('BernoulliNB  accuracy is %f'%score(BernoulliNB()) )
  File "G:/pycharm/quanbu.py", line 51, in score
    pred = classifier.batch_classify(dev)
AttributeError: 'SklearnClassifier' object has no attribute 'batch_classify'

我正在使用python3.5.2，nltk3.2.1，scikit-learn 0.18。

Answer 1

您一直在查看过时的文档。使用nltk 3，名称已更改为classify_many()。（同样在其他模块中，batch_pos_tag()和batch_parse()分别更改为pos_tag_sents()和parse_sents()。）

'SklearnClassifier'对象在NLTK中没有属性'batch_classify'

1 个答案: