基于hadoop的健身帖子推荐系统
舟率率 4/22/2026 pythonscalaspringbootkafkaredisjavavue
# 项目概况
# 数据类型
健身帖子数据
# 开发环境
centos7
# 软件版本
python3.8.18、hadoop3.2.0、spark3.1.2、mysql5.7.38、jdk8、kafka2.8.2、redis6.2.9
# 开发语言
python、Scala、Java
# 开发流程
数据上传(hdfs)->数据清洗(spark)->数据分析(spark)->内容相似度(spark)->机器学习(spark)->数据存储(mysql)->后端(springboot)->前端(vue)
# 可视化图表
# 操作步骤
# python安装包
pip3 install pandas==2.0.3 -i https://pypi.tuna.tsinghua.edu.cn/simple/
pip3 install pymysql==1.1.0 -i https://pypi.tuna.tsinghua.edu.cn/simple/
pip3 install jieba==0.42.1 -i https://pypi.tuna.tsinghua.edu.cn/simple
pip3 install hanlp==2.1.3 -i https://pypi.tuna.tsinghua.edu.cn/simple
1
2
3
4
5
6
2
3
4
5
6
# 启动MySQL
# 查看mysql是否启动 启动命令: systemctl start mysqld.service
systemctl status mysqld.service
# 进入mysql终端
# MySQL的用户名:root 密码:123456
# MySQL的用户名:root 密码:123456
# MySQL的用户名:root 密码:123456
mysql -uroot -p123456
1
2
3
4
5
6
7
8
9
2
3
4
5
6
7
8
9
# 启动Hadoop
# 离开安全模式: hdfs dfsadmin -safemode leave
# 启动hadoop
bash /export/software/hadoop-3.2.0/sbin/start-hadoop.sh
1
2
3
4
5
2
3
4
5
# 启动kafka
# 启动zookeeper
sh /export/software/kafka_2.12-2.8.2/bin/zookeeper-server-start.sh -daemon /export/software/kafka_2.12-2.8.2/config/zookeeper.properties
# 启动kafka
sh /export/software/kafka_2.12-2.8.2/bin/kafka-server-start.sh -daemon /export/software/kafka_2.12-2.8.2/config/server.properties
# 创建topic
/export/software/kafka_2.12-2.8.2/bin/kafka-topics.sh --create --topic user_action_topic --replication-factor 1 --partitions 1 --zookeeper master:2181
# 启动消费者
/export/software/kafka_2.12-2.8.2/bin/kafka-console-consumer.sh --bootstrap-server master:9092 --topic user_action_topic
# 关闭kafka
# sh /export/software/kafka_2.12-2.8.2/bin/kafka-server-stop.sh
# 关闭zookeeper
# sh /export/software/kafka_2.12-2.8.2/bin/zookeeper-server-stop.sh
1
2
3
4
5
6
7
8
9
10
11
12
13
14
2
3
4
5
6
7
8
9
10
11
12
13
14
# 启动redis
redis-server /export/software/redis-6.2.9/redis.conf
1
2
3
2
3
# 准备目录
mkdir -p /data/jobs/project/
cd /data/jobs/project/
# 解压 "data" 目录下的 "data.7z" 到当前目录下
# 上传 "project-hadoop-fitness-post-recommendation-system" 整个文件夹 到 "/data/jobs/project/" 目录
1
2
3
4
5
6
7
2
3
4
5
6
7
# 上传文件到hdfs
cd /data/jobs/project/project-hadoop-fitness-post-recommendation-system/data/
hdfs dfs -mkdir -p /data/input/
hdfs dfs -rm -r /data/input/*
hdfs dfs -put data.txt /data/input/
hdfs dfs -ls /data/input/
1
2
3
4
5
6
7
8
2
3
4
5
6
7
8
# 程序打包
cd /data/jobs/project/project-hadoop-fitness-post-recommendation-system/spark-job/
# 对 "spark-job" 目录下的项目 "spark-job" 进行打包
# 打包命令
mvn clean package -DskipTests
1
2
3
4
5
6
7
2
3
4
5
6
7
# 创建MySQL表
cd /data/jobs/project/project-hadoop-fitness-post-recommendation-system/mysql/
# 请确认mysql服务已经启动了
# 快速执行.sql文件内的sql语句
mysql -uroot -p123456 < mysql.sql
1
2
3
4
5
6
7
2
3
4
5
6
7
# spark数据清洗
cd /data/jobs/project/project-hadoop-fitness-post-recommendation-system/spark-job/
spark-submit \
--master local[*] \
--class org.example.DataCleaner \
target/spark-job.jar /data/input/ /data/output/clean_posts/
1
2
3
4
5
6
7
8
2
3
4
5
6
7
8
# 文本深度分析
cd /data/jobs/project/project-hadoop-fitness-post-recommendation-system/text_reduce/
python3 fitness_post_analysis.py
1
2
3
4
5
2
3
4
5
# spark内容相似度计算
cd /data/jobs/project/project-hadoop-fitness-post-recommendation-system/spark-job/
spark-submit \
--master local[*] \
--class org.example.PostSimilarityRecommender \
target/spark-job.jar /data/output/clean_posts/
1
2
3
4
5
6
7
8
2
3
4
5
6
7
8
# spark实时行为采集与增量更新
cd /data/jobs/project/project-hadoop-fitness-post-recommendation-system/spark-job/
spark-submit \
--master local[*] \
--class org.example.RealtimeSGDRecommender \
target/spark-job.jar
1
2
3
4
5
6
7
8
2
3
4
5
6
7
8
# 启动后端
cd /data/jobs/project/project-hadoop-fitness-post-recommendation-system/后端/fitness-forum-backend/
# 对 "fitness-forum-backend" 目录下的项目 "fitness-forum-backend" 进行打包
# 打包命令
mvn clean package -DskipTests
java -jar target/fitness-forum-backend-1.0.0.jar
1
2
3
4
5
6
7
8
9
2
3
4
5
6
7
8
9
# 启动前端
cd /data/jobs/project/project-hadoop-fitness-post-recommendation-system/前端/front_ui/
# 安装node
npm install --registry=https://registry.npmmirror.com
chmod -R 755 node_modules/.bin
npm run dev
# http://localhost:8081
# 登录用户: admin/user
# 登录密码: 12345678
1
2
3
4
5
6
7
8
9
10
11
2
3
4
5
6
7
8
9
10
11
# spark协同过滤模型训练
# 需要存在用户点赞/收藏行为以后,才能执行
cd /data/jobs/project/project-hadoop-fitness-post-recommendation-system/spark-job/
spark-submit \
--master local[*] \
--class org.example.OfflineALSTrainer \
target/spark-job.jar
1
2
3
4
5
6
7
8
9
2
3
4
5
6
7
8
9