计算机毕业设计hadoop+spark+hive漫画推荐系统 动漫视频推荐系统 漫画分析可视化大屏 漫画爬虫 漫画推荐系统 漫画爬虫 知识图谱 大数据

流程：
1.DrissionPage+Selenium自动爬虫工具采集漫画视频、详情、标签等约200万条漫画数据存入mysql数据库；
2.Mapreduce对采集的动漫数据进行数据清洗、拆分数据项等，转为.csv文件上传hadoop的hdfs集群；
3.hive建库建表导入.csv动漫数据；
4.一半指标使用hive_sql分析得出，一半指标使用Spark之Scala完成；
5.sqoop对分析结果导入mysql数据库；
6.Flask+echarts搭建可视化大屏；
创新点：Python全新DrissionPage+Selenium双爬虫使用、海量数据、爬虫、可视化大屏、离线hive+实时Spark双实现
可选装：推荐系统、预测系统、知识图谱、后台管理等。

核心算法代码分享如下：

import csv
 
import pymysql
import requests
connect = pymysql.connect(host="bigdata",port=3306, user="root",
                                           password="123456", database="hive_zymk")
 
cur = connect.cursor()
cur.execute("select * from tb_zymk " )
rv = cur.fetchall()
lines=0
for result in rv:
    id=result[0]
    title=result[1]
    update_times=result[2]
    tags=result[3]
    content=result[4]
    readings=result[5]
    subscribes=result[6]
    rewards=result[7]
    monthtickets=result[8]
    recommends=result[9]
    comments=result[10]
    scores=result[11]
    author=result[12]
    zps=result[13]
    ctime=result[14]
    img=result[15]
    url=result[16]
 
    #title字段去掉特殊字符
    title = title.strip().replace(',', '，').replace('"', '').replace("'", '').replace("\n", '').replace('\r', '').replace('\t', '')
    # 内容字段去掉特殊字符
    content = content.strip().replace(',', '，').replace('"', '').replace("'", '').replace("\n", '').replace('\r', '').replace('\t', '')
    #zps
    zps = zps.strip().replace(',', '，').replace('"', '').replace("'", '').replace("\n", '').replace('\r', '').replace('\t', '')
    # 阅读量替换汉字数值入亿 万
    temp_readings=readings
    if '亿' in readings:
        temp_readings=float(readings.replace('亿',''))*100000000
    elif '万' in readings:
        temp_readings = float(readings.replace('万', '')) * 10000
    readings = str(int(temp_readings))
 
    #订阅量subscribes
    temp_subscribes=subscribes
    if '亿' in subscribes:
        temp_subscribes=float(subscribes.replace('亿',''))*100000000
    elif '万' in subscribes:
        temp_subscribes = float(subscribes.replace('万', '')) * 10000
    subscribes = str(int(temp_subscribes))
 
    #打赏rewards
    temp_rewards=rewards
    if '亿' in rewards:
        temp_rewards=float(rewards.replace('亿',''))*100000000
    elif '万' in rewards:
        temp_rewards = float(rewards.replace('万', '')) * 10000
    rewards = str(int(temp_rewards))
 
    #monthtickets月票
    temp_monthtickets=monthtickets
    if '亿' in monthtickets:
        temp_monthtickets=float(monthtickets.replace('亿',''))*100000000
    elif '万' in monthtickets:
        temp_monthtickets = float(monthtickets.replace('万', '')) * 10000
    monthtickets = str(int(temp_monthtickets))
 
    #recommends 推荐数
    temp_recommends=recommends
    if '亿' in recommends:
        temp_recommends=float(recommends.replace('亿',''))*100000000
    elif '万' in recommends:
        temp_recommends = float(recommends.replace('万', '')) * 10000
    recommends = str(int(temp_recommends))
 
 
    #ctime需要去掉空格
    ctime=ctime.strip()
    if update_times == None or len(update_times) == 0 or update_times == '' or update_times == 'None':
        update_times = '无'
    #tags标签字段拆分
    # 标签
    tags_arr=tags.split('|')
    for tag in tags_arr:
        print(tag)
        zps_arr = zps.split('|')
        for zp in zps_arr:
            print(zp)
            zymk_file = open("zymk.csv", mode="a+", newline='', encoding="utf-8")
            zymk_writer = csv.writer(zymk_file)
            zymk_writer.writerow(
                [id,title,update_times,tag,content,readings,subscribes,rewards,monthtickets,recommends,comments,scores,author,zp,ctime,img,url])
            zymk_file.close()
            print('国产动漫转CSV',id,title,update_times,tag,content,readings,subscribes,rewards,monthtickets,recommends,comments,scores,author,zp,ctime,img,url)
            lines=lines+1
            print('正在处理,当前总计处理数据', lines,'条')
print('全部处理完毕','处理数据量',lines,'条')