python - VADER:每句话的情感

Question

我是 python 新手，我有一个如下所示的数据集

我正在从数据集中提取评论，并尝试应用 VADER 工具来检查与每个评论相关的情绪权重。我能够成功检索评论，但无法将 VADER 应用于每条评论。这是代码

import nltk
    import requirements_elicitation
    from nltk.sentiment.vader import SentimentIntensityAnalyzer

c = requirements_elicitation.read_reviews("D:\\Python\\testml\\my-tracks-reviews.csv")
class SentiFind:
    def init__(self,review):
        self.review = review

for review in c:
    review = review.comment
    print(review)

sid = SentimentIntensityAnalyzer()
for i in review:
    print(i)
    ss = sid.polarity_scores(i)
    for k in sorted(ss):
        print('{0}: {1}, '.format(k, ss[k]), end='')
    print()

示例输出:

g
compound: 0.0, neg: 0.0, neu: 0.0, pos: 0.0, 
r
compound: 0.0, neg: 0.0, neu: 0.0, pos: 0.0, 
e
compound: 0.0, neg: 0.0, neu: 0.0, pos: 0.0, 
a
compound: 0.0, neg: 0.0, neu: 0.0, pos: 0.0, 
t
compound: 0.0, neg: 0.0, neu: 0.0, pos: 0.0, 

compound: 0.0, neg: 0.0, neu: 0.0, pos: 0.0, 
a
compound: 0.0, neg: 0.0, neu: 0.0, pos: 0.0, 
p
compound: 0.0, neg: 0.0, neu: 0.0, pos: 0.0, 
p

我需要为每个评论自定义标签，如下所示

"Total weight: {0}, Negative: {1}, Neutral: {2}, Positive: {3}".

Answer 1

您定义的review是一个字符串，因此当您迭代它时，您会得到每个字母:

for i in review:
   print(i)

g
r
e
a...

因此，您需要分析器进行每条评论:

sid = SentimentIntensityAnalyzer()

for review in c:
    review = review.comment
    ss = sid.polarity_scores(review)
    total_weight = ss.compound
    positive = ss.pos
    negative = ss.neg
    neutral = ss.neu
    print("Total weight: {0}, Negative: {1}, Neutral: {2}, Positive: {3}".format(total_weight, positive, negative, neutral))