基于MLlib的商品推荐

实验名称

基于MLlib的商品推荐

实验目的

掌握基于Spark MLLib的协同过滤：

• 基于用户（User-Based）的协同过滤
• 基于物品（Item-Based）的协同过滤
• 基于ALS协同过滤

实验环境

• VMware Workstation
• Ubuntu 16.04
• Pycharm
• Pyspark

实验步骤

#-*-coding:utf-8
from  pyspark import  SparkContext
from pyspark.mllib.recommendation import  ALS,Rating
 
 
#数据格式 用户id，商品id，评分
sc=SparkContext("local[2]","second spark app")
#读取数据
rawData=sc.textFile("/home/lxs/Downloads/ratingdata.txt")
print(rawData.first())
#分离数据
rawRating=rawData.map(lambda line:line.split("\t")[:3])
print(rawRating.first())
 
 
ratings=rawRating.map(lambda x:Rating(int(x[0]),int(x[1]),float(x[2])))
print(ratings.first())
#模型训练
model=ALS.train(ratings,50,10,0.01)
 
print(model.userFeatures().count())
print(model.productFeatures().count())
#商品预测
predictedRating=model.predict(789,123)
print(predictedRating)
#推荐10个商品
userId=789
K=10
topKRecs=model.recommendProducts(userId,K)
print(topKRecs)