Bagofwords没有属性精确和召回

Question

import os


ENGLISH_OPINION_LEXICON_LOCATION = os.path.join('opinion-lexicon-English')

POS_WORDS_FILE = os.path.join（ENGLISH_OPINION_LEXICON_LOCATION，'positive-words.txt'）     NEG_WORDS_FILE = os.path.join（ENGLISH_OPINION_LEXICON_LOCATION，'negative-words.txt'）

pos_words = [（{'amazing'：True}，'positive'），（{'great'：True}，'positive'）]

neg_words = [（{'pathetic'：True}，'negative'）]

pos_words = []
neg_words = []

打开pos_word（POS_WORDS_FILE，'r'）。readlines（）[35：]：     pos_words.append（（{pos_word.rstrip（）：True}，'positive'））

for neg_word in open(NEG_WORDS_FILE, 'r').readlines()[35:]:
neg_words.append(({neg_word.rstrip(): True}, 'negative'))

print ("First 5 positive words %s "  % pos_words[:5])
print ("First 5 negative words %s"  % neg_words[:5])

print ("Number of positive words %d" % len(pos_words))

print ("Number of negative words %d" % len(neg_words))

all_words_with_sentiment = pos_words + neg_words

print ("Total number of words %d" % len(all_words_with_sentiment))

from nltk.classify import NaiveBayesClassifier

classifier = NaiveBayesClassifier.train(all_words_with_sentiment)


def to_dictionary(words):
return dict([(word, True) for word in words])


test_data = []


def predict_sentiment(text, expected_sentiment=None):
text_to_classify = to_dictionary(text.split())
result = classifier.classify(text_to_classify)
test_data.append([text_to_classify, expected_sentiment])
return result


POLARITY_DATA_DIR = os.path.join('polarity-data', 'rt-polaritydata')
POSITIVE_REVIEWS_FILE = os.path.join(POLARITY_DATA_DIR, 'rt-polarity-pos.txt')
NEGATIVE_REVIEWS_FILE = os.path.join(POLARITY_DATA_DIR, 'rt-polarity-neg.txt')

import collections

import nltk.classify
import nltk.metrics
#import nltk.metrics.scores
from nltk.util import LazyConcatenation, LazyMap

from nltk.util import LazyConcatenation, LazyMap
from decimal import *
from nltk.metrics.scores import (precision, recall)
from nltk.metrics import precision, recall
#from sklearn.metrics import precision_score


def run_sentiment_analysis_on_rt():
rt_positive_reviewers = open(POSITIVE_REVIEWS_FILE, 'r')

expected_pos_set = collections.defaultdict(set)
actual_pos_set = collections.defaultdict(set)

for index, review in enumerate(rt_positive_reviewers.readlines()):
    expected_pos_set['positive'].add(index)
    actual_sentiment = predict_sentiment(review, 'positive')
    actual_pos_set[actual_sentiment].add(index)

print ("Total Negative found in positive reviews %s" % 
len(actual_pos_set['negative']))

rt_negative_reviews = open(NEGATIVE_REVIEWS_FILE, 'r')

expected_neg_set = collections.defaultdict(set)
actual_neg_set = collections.defaultdict(set)

for index, review in enumerate(rt_negative_reviews.readlines()):
    expected_neg_set['negative'].add(index)
    actual_sentiment = predict_sentiment(review, 'negative')
    actual_neg_set[actual_sentiment].add(index)

print ("Total Positive found in negative reviews %s" % 
len(actual_neg_set['positive']))

print ('accuracy: %.2f' % nltk.classify.util.accuracy(classifier, test_data))
#print ('pos precision:', nltk.metrics.precision(refsets['pos'], testsets['pos']))
print ('pos precision: %.2f' % 
nltk.metrics.scores.precision(expected_pos_set['positive'], actual_pos_set['positive']))
print ('pos recall: %.2f' % 
nltk.metrics.scores.recall(expected_pos_set['positive'], actual_pos_set['positive']))
print ('neg precision: %.2f' % 
nltk.metrics.scores.precision(expected_neg_set['negative'], actual_neg_set['negative']))
print ('neg recall: %.2f' % 
nltk.metrics.scores.recall(expected_neg_set['negative'], actual_neg_set['negative']))


run_sentiment_analysis_on_rt()

我得到的错误如下： AttributeError：module'nltk.translate.metrics'没有属性'scores'

Answer 1

你可以尝试

from nltk import precision

并直接调用precision方法。它会工作！

Answer 2

我认为你错误地调用了精确功能。 从docs来看，你似乎应该使用它 nltk.metrics.scores.precision代替nltk.metrics.precision。回忆也是如此。

Answer 3

只需导入以下内容。

from nltk import precision

并按如下所示直接调用precision方法。

nltk.precision(expected_neg_set['negative'], actual_neg_set['negative']))

肯定会起作用！