def analyze(sentence):
s = SnowNLP(sentence)
if (round(s.sentiments) >= 0.5):
return (round(round(s.sentiments * 10)))
else:
return (round(round(s.sentiments * 10) * (-2)))
query_id = []
passage = []
query = []
answer = []
alternatives = []
passage_len = []
query_in_char = []
query_in_word = []
query_in_char_set = []
query_in_word_set = []
ques_mark = []
for i in range(0, len(lines)):
l = lines[i]
ge = json.loads(l)
que = SnowNLP(ge.get('query', '')).han
query.append(que)
query_id.append(ge.get('query_id', ''))
pas = SnowNLP(ge.get('passage', '')).han
try:
pas = pas.split('？')[-1].strip()
except:
pass
pas = pas.replace(que, '')
if len(pas) < 2:
pas = SnowNLP(ge.get('passage', '')).han
print(ge.get('passage', ''))
print(ge.get('query', ''))
print(ge.get('answer', ''))
def split_all():
input_list_1 = 'list/data_location/mycode/china_loc_list_full.txt'
input_list = 'list/data_location/mycode/china_prov_list_simple.txt'
f_in = open(input_list, 'r')
provs = f_in.readlines()
f_in.close()
provs = [prov.strip() for prov in provs]
save_dir = './data/select_prov/'
f = open('./list/stop_words.txt', 'r')
stop_words = f.readlines()
f.close()
stop_words = [i.strip() for i in stop_words]
count_all = {}
for prov in provs:
for year in xrange(2002, 2018):
year = str(year)
label = save_dir + year + '/' + ''.join(
[str(i) for i in SnowNLP(unicode(prov)).pinyin])
#print label
year_dict = read_dict(label + '_all_count.txt')
for key in year_dict:
if key not in count_all:
count_all[key] = np.zeros(17, 'float')
count_all[key][0] += year_dict[key]
count_all[key][int(year) - 2001] += year_dict[key]
for prov in provs:
loc_counts = make_dict(input_list_1)
word_counts = {}
for year in xrange(2002, 2018):
year = str(year)
label = save_dir + year + '/' + ''.join(
[str(i) for i in SnowNLP(unicode(prov)).pinyin])
#print label
year_dict = read_dict(label + '_all_count.txt')
for key in loc_counts.keys():
if key in year_dict:
loc_counts[key][0] += year_dict[key]
loc_counts[key][int(year) - 2001] = year_dict[key]
for key in year_dict:
if key in word_counts:
word_counts[key] += year_dict[key]
else:
word_counts[key] = year_dict[key]
for key in loc_counts.keys():
for i in xrange(0, 13):
if loc_counts[key][i] != 0:
loc_counts[key][i] /= count_all[key][i]
# for key in word_counts.keys():
# word_counts[key] /= count_all[key][0]
print(prov)
word_counts = sorted(word_counts.items(),
key=lambda x: x[1],
reverse=True)
loc_counts = sorted(loc_counts.items(),
key=lambda x: x[1][0],
reverse=True)
# _all_stop: all_count - stop_words
f = open(
save_dir + 'all/' +
''.join([str(i) for i in SnowNLP(unicode(prov)).pinyin]) +
'_all_stop.txt', 'w')
for k, v in word_counts:
if k not in stop_words and len(k.decode()) > 1:
f.write(k + ' ' + str(v) + '\n')
f.close()
## _all_count: all_count(word, count)
#f = open(save_dir+'all/'+''.join([str(i) for i in SnowNLP(unicode(prov)).pinyin])+'_all_count.txt', 'w')
#for k,v in word_counts:
# if k not in stop_words:
# f.write(k+' '+str(v)+'\n')
#f.close()
## _all_count_list: [word_in_list, all_count, 16*year_count]
#f_1 = open(save_dir+'all/'+''.join([str(i) for i in SnowNLP(unicode(prov)).pinyin])+'_all_count_list.txt', 'w')
#for k,v in loc_counts:
# f_1.write(k+' '+array_2_str(v)+'\n')
#f_1.close()
count_all = sorted(count_all.items(), key=lambda x: x[1][0], reverse=True)
f_1 = open(save_dir + 'all_count.txt', 'w')
for k, v in count_all:
if k not in stop_words:
f_1.write(k + ' ' + array_2_str(v.astype('int')) + '\n')
f_1.close()
def _sentiment(text):
s = SnowNLP(text)
return s.sentiments
# -*- coding: utf-8 -*-
from snownlp import SnowNLP
import codecs
def load_sentences(path):
with codecs.open(path, encoding='utf-8-sig') as f1:
text = f1.read()
return text.replace("\r", "").split("\n")
for sent in s.sentences:
print(sent, SnowNLP(sent).sentiments)
def test_new_word():
new_word = {"positive_word": [], "negative_word": []}
with codecs.open(Path1 + "new_word.txt", encoding="utf-8-sig") as f1:
for line in f1.readlines():
direction, word = line.replace("\r\n", "").split(" ")[:2]
if direction == u"positive":
new_word["positive_word"].append(word)
if direction == u"negative":
new_word["negative_word"].append(word)
def sentiment_analysis(text):
snownlp = SnowNLP(text)
return Comment.rank_sentiments(snownlp.sentiments)
from snownlp import SnowNLP
text="皱皱的颜色。也不鲜艳。很喜欢。但总归。还是嫌弃。"
s = SnowNLP(text)
for i in s.sentences:
#print(i)
s1 = SnowNLP(i)
#print(s1.sentiments)
text="皱皱的颜色也不鲜艳很喜欢但总归还是嫌弃"
a = SnowNLP(text)
print(text)
print(a.sentiments)
text="皱皱的颜色也不鲜艳很嫌弃但总归还是喜欢"
a = SnowNLP(text)
print(text)
print(a.sen timents)
def NewsProcess(data):
s = SnowNLP(data)
print(s.words)
print("##" + str(s.sentiments))
# sents = normal.get_sentences(t)
# doc = []
# for sent in sents:
# words = seg.seg(sent)
# words = normal.filter_stop(words)
# doc.append(words)
# rank = textrank.TextRank(doc)
# rank.solve()
# for index in rank.top_index(5):
# print(sents[index])
# keyword_rank = textrank.KeywordTextRank(doc)
# keyword_rank.solve()
# for w in keyword_rank.top_index(5):
# print(w)
sentimentslist = []
data = xlrd.open_workbook('MOBICK.xlsx')
table = data.sheets()[0]
nrows = table.nrows
for i in range(nrows):
str = "".join('%s' % id for id in table.row_values(i))
s = SnowNLP(str)
print(table.row_values(i), s.sentiments)
sentimentslist.append(s.sentiments)
fileObject = open('MOBICK.txt', 'w+')
for ip in sentimentslist:
fileObject.write(__builtins__.str(ip))
fileObject.write('\n')
fileObject.close()
fig1 = plt.figure("sentiment")
plt.hist(sentimentslist, bins=np.arange(0, 1, 0.02))
plt.show()
def comment_analysis(pagedata):
jsondata = str(pagedata)
df = pd.DataFrame(eval(jsondata))
df = df.drop_duplicates()
#Constuct Features
ud = []
ff = []
fo = []
for i in range(0, len(df)):
try:
#trim Comments
s = df.loc[i, "Comments"]
df.loc[i, "Comments"] = s[1:len(s) - 1]
#drop rows with pure repost
if (df.loc[i, "Comments"] == "转发微博"):
df = df.drop(i)
else:
if ("回复@" in df.loc[i, "Comments"]):
s = df.loc[i, "Comments"]
df.loc[i, "Comments"] = s[s.index(':') + 1:len(s)]
#label user description attribute
if (str(df.loc[i, "User_description"]) == "nan"):
ud.append(0)
else:
ud.append(1)
#label verified user attribute
if (df.loc[i, "Verified"] == 0):
df.loc[i, "Verified"] = 1
elif (df.loc[i, "Verified"] == -1):
df.loc[i, "Verified"] = 0
#calculate follower/following and follower/originalPost ratio
if df.loc[i, "Follower"] == 0:
df.loc[i, "Follower"] = 1
ff.append(df.loc[i, "Following"] / df.loc[i, "Follower"])
fo.append(df.loc[i, "Original_post"] / df.loc[i, "Follower"])
except KeyError:
ud.append(-1)
ff.append(-1)
fo.append(-1)
df = df.reset_index(drop=True)
#add features from post/user information
df['description'] = ud
df['ffRatio'] = ff
df['foRatio'] = fo
for i in range(0, len(df)):
if df.loc[i, "description"] == -1:
df = df.drop(i)
ud.remove(-1)
ff.remove(-1)
fo.remove(-1)
df = df.reset_index(drop=True)
#Append Sentiment Score
pynlpir.open()
segPosts = []
sentiScore = []
translator = Translator()
r = '[’！？：；【】，《》!"#$%&\'()（）“”…*+,-./:;<=>?@[\\]^_`{|}~]+'
#Append Sentiment Score for all Comments under original post
for i in range(0, len(df)):
df.loc[i, "Comments"] = re.sub(r, '', df.loc[i, "Comments"])
try:
if "en" in str(detect(df.loc[i, "Comments"])):
transText = str(
translator.translate(df.loc[i, "Comments"],
src='en',
dest='zh-cn').text)
line = transText.strip()
s = SnowNLP(line)
senti = (s.sentiments +
SnowNLP(re.sub("[0-9]", "", line)).sentiments) / 2
sentiScore.append(senti)
seg = pynlpir.segment(line, pos_tagging=False)
segPosts.append(seg)
elif "zh-cn" or "zh-tw" or "ko" in str(
detect(df.loc[i, "Comments"])):
line = df.loc[i, "Comments"].strip()
s = SnowNLP(line)
senti = (s.sentiments +
SnowNLP(re.sub("[0-9]", "", line)).sentiments) / 2
sentiScore.append(senti)
seg = pynlpir.segment(line, pos_tagging=False)
segPosts.append(seg)
else:
#drop rows without valid sentiment scores
print("error1")
df = df.drop(i)
except:
print("error2")
df = df.drop(i)
df = df.reset_index(drop=True)
df['Sentiment'] = sentiScore
df = df.drop('Follower', axis=1).drop('Following', axis=1).drop(
'User_description',
axis=1).drop('Original_post',
axis=1).drop("Comments", axis=1).drop("Username",
axis=1).drop("UID",
axis=1)
pd.set_option('display.max_rows', 400)
df = df.round({'Sentiment': 5})
return df
def senti(x): return (x[0],[SnowNLP(x[1]).sentiments,1])
plt.xlabel('Epochs')
plt.ylabel('Loss and Acc')
plt.legend()
#plt.show()
pre_y = model.predict(pre_x, batch_size=32)
pre_y_class = model.predict_classes(pre_x, batch_size=32)
pre_y = [round(x[0], 3) for x in pre_y]
pre_y_class = [x[0] for x in pre_y_class]
print(pre_y[:10])
print(pre_y_class[:10])
# 关键词情感分类
star = []
for it in x_pre:
s = SnowNLP(it)
t = s.sentiments
star.append(round(t, 3))
print(star[:10])
##汇总
star_class = np.array(star)
star_class = np.where(star_class >= 0.5, 1, 0).tolist()
x_pre = x_pre.tolist()
y_label = y_label.tolist()
# pre_y = pre_y.tolist()
# pre_y_class = pre_y_class.tolist()
# 验证模型准确率
acc1 = accuracy_score(y_label, pre_y_class)
acc2 = accuracy_score(y_label,star_class)
def fenci(string):
s = SnowNLP(string)
#result = s.words
#print(len(result), '/'.join(result)) # 。words 分词
print('Sentiments:', s.sentiments) # 。sentiments 情感分析
from snownlp import SnowNLP text = '我来到河北保定河北大学上学' s = SnowNLP(text) print(s.words)
def snownlp_segment(self, sentence):
# snownlp分词
# unicode_sentence = sentence.decode('gbk')
sentence = SnowNLP(sentence).words
return ' '.join(sentence)
（7）文本关键词和文本摘要提取（TextRank算法）
（8）计算文档词频（TF，Term Frequency）和逆向文档频率（IDF，Inverse Document Frequency）
（9）Tokenization（分割成句子）
（10）文本相似度计算（BM25）
SnowNLP的最大特点是特别容易上手，用其处理中文文本时能够得到不少有意思的结果，但不少功能比较简单，还有待进一步完善。
# In[10]:
from snownlp import SnowNLP
s=SnowNLP(u'杭州西湖风景很好，是旅游胜地,每年吸引大量前来游玩的游客！')
#分词
print(s.words)
# In[11]:
#情感词性计算
print("该文本的情感词性为正的概率:" + str(s.sentiments))
# In[12]:
_s=SnowNLP(u'今天又是下雨又是刮风,真是糟糕透了!')
print("该文本的情感词性为正的概率:" + str(_s.sentiments))
# -*- coding: utf-8 -*-
import pandas as pd
from snownlp import SnowNLP
if __name__ == '__main__':
test = pd.read_csv(r"TestModel.csv")
review_list = [review for review in test['review']]
label_list = [label for label in test['label']]
list_test = [(label, review)
for label, review in list(zip(label_list, review_list))
if type(review) != float]
for j in list_test:
print(j[1], j[0], SnowNLP(j[1]).sentiments)
senti = [SnowNLP(review).sentiments for label, review in list_test]
newSenti = []
for i in senti: # 预测结果为pos的概率,大于0.6我们认定为积极评价
if (i >= 0.6):
newSenti.append(1)
else:
newSenti.append(0)
counts = 0
for i in range(len(list_test)):
if (newSenti[i] == list_test[i][0]):
counts += 1
accuracy = float(counts) / float(len(list_test))
print("准确率为:%.2f" % accuracy)
def qingxu_number(text):
s = SnowNLP(text)
print('本句情绪指数:%f' % s.sentiments)
""" 中文文本的情感分析 @Date 2020.04.28 pip3 install snownlp """ from snownlp import SnowNLP text = '人们日常所犯最大的错误，是对陌生人太客气，而对亲密的人太苛刻了，而努力改掉这个习惯。' s = SnowNLP(text) # 分词 print(s.words) # 词性标注 tags = [x for x in s.tags] print(tags) # 断句 print(s.sentences) # 拼音 print(s.pinyin) # 情绪判断，返回值为正面情绪的概率，越接近1表示正面情绪，越接近0表示负面情绪 text1 = '这部电影太棒了' text2 = '这部电影简直是烂到爆' s1 = SnowNLP(text1) s2 = SnowNLP(text2) # 这部电影太棒了 0.9829468270441747 print(text1, s1.sentiments) # 这部电影简直是烂到爆 0.2296519477554504 print(text2, s2.sentiments) # 关键字抽取
# coding=utf8 from snownlp import sentiment from snownlp import SnowNLP s = SnowNLP(u"这个东西真赞") print(s.sentiments)
def Txtmine(indata):
data = indata[6]
title = indata[2]
dresult["title"] = title
dresult["article"] = data
mylen = len(data)
line_num = data.count("\n")
"""字数行数"""
dresult["wordnums"] = mylen
dresult["linenums"] = line_num
"""词频"""
words = jieba.lcut(data)
counts = {}
for word in words:
if len(word) == 1:
continue
elif word.isdigit():
continue
else:
rword = word
counts[rword] = counts.get(rword, 0) + 1
items = list(counts.items())
items.sort(key=lambda x: x[1], reverse=True)
dresult["words"] = items[0:3]
"""情感"""
s = SnowNLP(data)
dresult["sentiments"] = s.sentiments
"""摘要"""
s = SnowNLP(data.replace("不", ""))
# print(s.keywords())
swords = list()
for word in s.keywords():
if len(word) < 2:
continue
else:
swords.append(word)
dresult["keywords"] = swords
summary = list(set(s.summary()))[:3:1]
# print(summary)
lines = ""
for line in summary:
# print(line)
lines = lines + line + " //// "
dresult["summary"] = lines
"""特征值"""
values = [i for i in indata[7:]]
dresult["val"] = values
"""超链接数目"""
results = re.findall("(?isu)(https\://[a-zA-Z0-9\.\?/&\=\:]+)", data)
dresult["urlnums"] = len(results)
"""错字检测"""
r = requests.post("http://www.cuobiezi.net/api/v1/zh_spellcheck/json",
data={
'content': '我最喜欢的就是元啊节吃汤圆。 ',
'check_mode': 'advanced',
'action': 'show'
})
""""""
loc = 0
org = 0
peo = 0
words, ners = fool.analysis(data)
for ner in ners[0]:
if ner[2] == "location":
loc += 1
elif ner[2] == "org":
org += 1
elif ner[2] == "person":
peo += 1
else:
continue
dresult["loc"] = loc
dresult["org"] = org
dresult["peo"] = peo
# print(dresult)
return dresult
# coding:utf-8
# SnowNLP是python中用来处理文本内容的，
# 可以用来分词、标注、文本情感分析等，情感分析是简单的将文本分为两类，
# 积极和消极，返回值为情绪的概率，越接近1为积极，接近0为消极。代码如下：
import numpy as np
from snownlp import SnowNLP
import matplotlib.pyplot as plt
f = open('avengers.txt', 'r')
list = f.readlines()
sentimentslist = []
for i in list:
s = SnowNLP(i)
# print s.sentiments
sentimentslist.append(s.sentiments)
plt.hist(sentimentslist, bins=np.arange(0, 1, 0.01), facecolor='g')
plt.xlabel('Sentiments Probability')
plt.ylabel('Quantity')
plt.title('Analysis of Sentiments')
plt.show()
def process_data_8(text_name, gt_name, save_name, simple):
data = {}
with open(text_name, 'r', encoding='utf-8') as f:
for line in f.readlines():
line = line.strip()
id = line.split()[0][5:-1]
text = ''.join(line.split()[1:])
if simple:
if len(text) != len(SnowNLP(text).han):
print(text)
han = list(SnowNLP(text).han)
new_han = []
k = 0
for i in range(len(han)):
if i < k:
continue
if i+3 < len(han) and han[i:i+4] == ['公', '共', '汽', '车']:
new_han += ['公', '车']
k = i + 4
else:
new_han += han[i]
new_text = ''.join(new_han)
if len(text) != len(new_text):
han = list(new_text)
new_han = []
k = 0
for i in range(len(han)):
if i < k:
continue
if i + 2 < len(han) and han[i:i + 3] == ['出', '租', '车']:
new_han += ['的', '士']
k = i + 3
else:
new_han += han[i]
new_text = ''.join(new_han)
print(new_text)
if len(text) != len(new_text):
han = list(new_text)
new_han = []
k = 0
for i in range(len(han)):
if i < k:
continue
if i + 2 < len(han) and han[i:i + 3] == ['因', '特', '网']:
new_han += ['网', '际', '网', '络']
k = i + 3
else:
new_han += han[i]
new_text = ''.join(new_han)
print(new_text)
assert len(text) == len(new_text)
text = new_text
else:
text = SnowNLP(text).han
data[id] = {'text': text}
with open(gt_name, 'r', encoding='utf-8') as f:
for line in f.readlines():
line = line.strip().split(', ')
answer = []
# extract wrong word id and replaced word
if line[1] == '0':
data[line[0]]['answer'] = []
else:
for i in range(1, len(line), 2):
if simple:
line[i + 1] = SnowNLP(line[i+1]).han
answer.append([int(line[i]), line[i+1]])
data[line[0]]['answer'] = answer
with open(save_name, 'w', encoding='utf-8') as f:
json.dump(data, f)
from snownlp import SnowNLP
text1 = '垃圾东西，很难吃'
def rank_sentiments(sentiments):
if sentiments >= 0.8:
return 5
elif sentiments >= 0.6:
return 4
elif sentiments >= 0.4:
return 3
elif sentiments >= 0.2:
return 2
else:
return 1
s1 = SnowNLP(text1)
print(rank_sentiments(s1.sentiments))
def split_all():
input_list_1 = 'list/bag.txt'
input_list = 'list/data_location/mycode/china_prov_list_simple.txt'
f_in = open(input_list, 'r')
provs = f_in.readlines()
f_in.close()
provs = [prov.strip() for prov in provs]
save_dir = './data/select_prov/'
f = open('./list/stop_words.txt', 'r')
stop_words = f.readlines()
f.close()
stop_words = [i.strip() for i in stop_words]
for prov in provs:
loc_counts = make_dict(input_list_1)
word_counts = {}
for year in xrange(2002, 2018):
year = str(year)
label = save_dir + year + '/' + ''.join(
[str(i) for i in SnowNLP(unicode(prov)).pinyin])
#print label
year_dict = read_dict(label + '_count_search.txt')
for key in loc_counts.keys():
for key_split in key.split():
if key_split in year_dict:
loc_counts[key][0] += year_dict[key_split]
loc_counts[key][int(year) -
2001] = year_dict[key_split]
for key in year_dict:
for key_split in key.split():
if key in word_counts:
word_counts[key] += year_dict[key_split]
else:
word_counts[key] = year_dict[key_split]
print(prov)
word_counts = sorted(word_counts.items(),
key=lambda x: x[1],
reverse=True)
loc_counts = sorted(loc_counts.items(),
key=lambda x: x[1][0],
reverse=True)
# _all_stop: all_count - stop_words
#f = open(save_dir+'all/'+''.join([str(i) for i in SnowNLP(unicode(prov)).pinyin])+'_all_stop.txt', 'w')
#for k,v in word_counts:
# if k not in stop_words:
# f.write(k+' '+str(v)+'\n')
#f.close()
## _all_count: all_count(word, count)
#f = open(save_dir+'all/'+''.join([str(i) for i in SnowNLP(unicode(prov)).pinyin])+'_all_count.txt', 'w')
#for k,v in word_counts:
# if k not in stop_words:
# f.write(k+' '+str(v)+'\n')
#f.close()
# _all_count_list: [word_in_list, all_count, 16*year_count]
f_1 = open(
save_dir + 'all/' +
''.join([str(i) for i in SnowNLP(unicode(prov)).pinyin]) +
'_count_search_bag.txt', 'w')
for k, v in loc_counts:
f_1.write(k + ' ' + array_2_str(v) + '\n')
f_1.close()
def get_sentiment(word):
text = u'{}'.format(word)
s = SnowNLP(text)
print(s.sentiments)
from snownlp import SnowNLP
import matplotlib.pyplot as plt
import numpy as np
conn = pymysql.connect(host='127.0.0.1',
user='******',
password='',
charset='utf8')
with conn:
cur = conn.cursor()
cur.execute("SELECT * FROM sinaweibo.macomment")
rows = cur.fetchall()
commentlist = []
for row in rows:
row = list(row)
del row[0]
if row not in commentlist:
commentlist.append(row[2])
conn.close()
print("Finish fetching the comment data...")
# print(commentlist)
# snowanalysis
print("Start SnowNLP data")
sentimentslist = []
for com in commentlist:
s = SnowNLP(com)
sentimentslist.append(s.sentiments)
# print(sentimentslist)
plt.hist(sentimentslist, bins=np.arange(0, 1, 0.02))
plt.show()
from snownlp import SnowNLP
text = '手机质量非常好，非常漂亮！'
s = SnowNLP(text)
# print(s.words)
# tags = [x for x in s.tags]
# print(tags)
# print(s.sentiments)
# print(s.keywords(limit=20))
print(s.sentences)
for sen in s.sentences:
t = SnowNLP(sen)
tags = [x for x in t.tags]
for k, v in tags:
if (v == 'n'):
print(k)
print(tags)
#encoding=utf-8 from snownlp import SnowNLP """汉字转拼音""" s = SnowNLP(u'这个东西真心很赞') print s.pinyin # [u'zhe', u'ge', u'dong', u'xi', # u'zhen', u'xin', u'hen', u'zan']
def saveComments():
index = 0
mongoClient = MongoClient('172.24.177.30', 27017)
db = mongoClient['futures_data']
posts = db['comments']
sinanews = db['sinanews']
for data in loadUrls():
try:
index += 1
tempList = []
url = data['url']
id = data['id']
print(url)
url = getCUrl(url)
request.urlopen(url)
response = request.urlopen(url + '&page=1&page_size=200')
jsonStr = response.read()
print(jsonStr[9:])
jsonObj = json.loads(jsonStr[9:])
tempList.extend(jsonObj['result']['cmntlist'])
count = int(jsonObj['result']['count']['show'])
pageNum = math.ceil(count / 200)
if pageNum > 1:
for page in range(2, pageNum + 1):
response = request.urlopen(url +
('&page=%s&page_size=200' %
page))
jsonStr = response.read()
jsonObj = json.loads(jsonStr[9:])
tempList.extend(jsonObj['result']['cmntlist'])
print('page: %s' % page)
print(len(tempList))
# break
sentiments = []
positive = 0
negative = 0
for temp in tempList:
s = SnowNLP(temp['content'])
sentiments.append(s.sentiments)
if s.sentiments > 0.5:
positive += 1
else:
negative += 1
if len(sentiments) != 0:
sentiment = np.mean(sentiments)
else:
sentiment = -1
# saveObj = {'id': id, 'url': url, 'comment': tempList, 'sentiment': sentiment, 'positive': positive, 'negative': negative}
print('第' + str(index) + '篇', sentiment, data['id'])
# posts.insert_one(saveObj)
sinanews.update(
{'_id': ObjectId(data['id'])},
{'$set': {
'positive': positive,
'negative': negative
}})
except Exception as e:
print(e)
