调用baidu-aip进行情感分析

 代码跟随b站up主“你好我是大河”的教程编写（有改动）。

流程：

 

import aip
import time
import pandas as pd
from tqdm import tqdm

content_list = []
positive_prob_list = []
negative_prob_list = []
sentiment_prob_list = []


def sentiment_classify(txt):
    client_appid = '**'  ##**个人账号内容，token需要先通过AK和SK获取
    client_ak = '**'
    client_sk = '**'
    my_nlp = aip.nlp.AipNlp(client_appid, client_ak, client_sk)
    token = '**'
    url = 'https://aip.baidubce.com/rpc/2.0/nlp/v1/sentiment_classify?charset=UTF-8&access_token={}'.format(token)

    result = my_nlp.sentimentClassify(txt)
    # print(my_nlp.sentimentClassify(txt))
    try:                #记录错误值并继续执行
        positive_prob = result['items'][0]['positive_prob']
        negative_prob = result['items'][0]['negative_prob']
        sentiment_prob = result['items'][0]['sentiment']                
    except:
        positive_prob ="Nan"
        negative_prob ="Nan"
        sentiment_prob="Nan"



    content_list.append(txt)
    positive_prob_list.append(positive_prob)
    negative_prob_list.append(negative_prob)
    sentiment_prob_list.append(sentiment_prob)
    return positive_prob


df = pd.read_excel('D:/desktop/试.xlsx')                #读取文件
txt_content = df["content"]                              #确定文本，表格第一行为content
print(txt_content)

positive_times = 0
negative_times = 0
for txt in tqdm(txt_content):
    time.sleep(1)                #延时访问
    if float(sentiment_classify(txt)) > 0.5:
        positive_times += 1
    else:
        negative_times += 1
#    print(sentiment_classify(txt))

df_res = pd.DataFrame({"content": content_list, "positive_prob": positive_prob_list, "negative_prob": negative_prob_list,"sentiment_prob": sentiment_prob_list})
df_res.to_excel("D:/desktop/结果.xlsx")                #写入文件
print("{}，{}".format(positive_times, negative_times))
# 正向{}条，负向{}条