CVPR关键词分析生成热词汇云图

任务目标：    根据CVPR论文生成热点词汇云图

 

任务分解： 

1.          爬取CVPR2019论文到数据库中；
2.          分析、查找出关键词，并排序；
3.          生成热词汇云图；
4.          点击热词云中的热词可以找到与之对应的文章题目；

 

一：Python爬虫

    分析网站源代码，利用Python爬取数据，然后存储到Mysql数据库中。爬取时将题目拆解成关键词，也可以直接将摘要拆解。

     

import requests
import pymysql
from bs4 import BeautifulSoup

db = pymysql.connect('127.0.0.1',
                     port=3306,
                     user='root',
                     password='123456',
                     db='mytest',
                     charset='utf8')

cursor = db.cursor()

headers={
        "User-Agent":"Mozilla/5.0 (Windows NT 10.0; WOW64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/74.0.3729.169 Safari/537.36"
        }
url="http://openaccess.thecvf.com/CVPR2019.py"
html=requests.get(url)

soup=BeautifulSoup(html.content,'html.parser')

soup.a.contents=='pdf'

pdfs=soup.findAll(name="a",text="pdf")

lis = []
jianjie=""
for i,pdf in enumerate(pdfs):
    pdf_name=pdf["href"].split('/')[-1]
    name=pdf_name.split('.')[0].replace("_CVPR_2019_paper","")
    link="http://openaccess.thecvf.com/content_CVPR_2019/html/"+name+"_CVPR_2019_paper.html"
    url1=link
    html1 = requests.get(url1)
    soup1 = BeautifulSoup(html1.content, 'html.parser')
    weizhi = soup1.find('div', attrs={'id':'abstract'})
    if weizhi:
        jianjie =weizhi.get_text();
    print("这是第"+str(i)+"条数据")
    keyword = str(name).split('_')
    keywords = ''
    for k in range(len(keyword)):
        if (k == 0):
            keywords += keyword[k]
        else:
            keywords += ',' + keyword[k]
    info = {}
    info['title'] = name
    info['link'] =link
    info['abstract']=jianjie
    info['keywords']=keywords
    lis.append(info)

cursor = db.cursor()
for i in range(len(lis)):
    cols = ", ".join('`{}`'.format(k) for k in lis[i].keys())
    print(cols)  # '`name`, `age`'

    val_cols = ', '.join('%({})s'.format(k) for k in lis[i].keys())
    print(val_cols)  # '%(name)s, %(age)s'

    sql = "insert into lunwen(%s) values(%s)"
    res_sql = sql % (cols, val_cols)
    print(res_sql)

    cursor.execute(res_sql, lis[i])  # 将字典a传入
    db.commit()
    num=1
    print(num)
    print("ok")

Python