【机器学习】Mahout 基于用户的协同过滤推荐

J2EE实现:

maven SpringMVC项目可部署到服务器，由于比较复杂就不再这里写了感兴趣的可访问

https://github.com/XVIIB/UCF-XVII

JavaSE实现

1：输入数据（26M大小的csv文件，包含1000000+条淘宝用户浏览购买信息）：

用户id   商品id   评分

2：maven依赖：

http://mvnrepository.com/artifact/org.apache.mahout/mahout-core

http://mvnrepository.com/artifact/org.apache.mahout/mahout-integration

3：主程序代码（推荐系统主体）

import org.apache.mahout.cf.taste.common.TasteException;
import org.apache.mahout.cf.taste.impl.model.file.FileDataModel;
import org.apache.mahout.cf.taste.impl.neighborhood.NearestNUserNeighborhood;
import org.apache.mahout.cf.taste.impl.recommender.GenericUserBasedRecommender;
import org.apache.mahout.cf.taste.impl.similarity.EuclideanDistanceSimilarity;
import org.apache.mahout.cf.taste.impl.similarity.PearsonCorrelationSimilarity;
import org.apache.mahout.cf.taste.model.DataModel;
import org.apache.mahout.cf.taste.neighborhood.UserNeighborhood;
import org.apache.mahout.cf.taste.recommender.RecommendedItem;
import org.apache.mahout.cf.taste.recommender.Recommender;
import org.apache.mahout.cf.taste.similarity.UserSimilarity;

import java.io.File;
import java.io.IOException;
import java.util.List;

public class C {
    public static void main(String[] args) {
        //准备源数据
        File file=new File("D:\\data\\user.csv");
        DataModel dataModel= null;
        System.out.println(file.toString());
        //从文件读取并整理所有用户id
        ArrayList<String> user=FileOperation.getUserId();
        try {
            //加载数据
            dataModel = new FileDataModel(file);
            //计算相似度这里采用欧式距离
            UserSimilarity similarity=new EuclideanDistanceSimilarity(dataModel);
            //基于5个邻居计算相似度
            UserNeighborhood neighborhood=new NearestNUserNeighborhood(1000000,similarity,dataModel);
            //基于用户协同推荐
            Recommender recommender=new GenericUserBasedRecommender(dataModel,neighborhood,similarity);
            
            long a=0;
            File outfile=new File("F:\\out.csv");
            FileWriter fileWriter=new FileWriter(outfile);
            for (String item:user){
                a= Long.parseLong(item);
                System.out.println(a);
                List<RecommendedItem> list=recommender.recommend(Long.parseLong(item),1);
                if (list.size()==0) continue;
                try {
                //输出到本地文件
                    fileWriter.write(a+","+list.get(0).getItemID()+"\r\n");
                } catch (FileNotFoundException e) {
                    e.printStackTrace();
                } catch (IOException e) {
                    e.printStackTrace();
                }
            }
            fileWriter.close();
        } catch (IOException e) {
            System.out.println("filed");
            e.printStackTrace();
        } catch (TasteException e) {
            System.out.println("filed");
            e.printStackTrace();
        }
    }
}

 用户id处理类：

public class FileOperation {
        public static ArrayList getUserId(){
            Set set=new TreeSet();
            ArrayList list=null;
            File file = new File("D:\\data\\user.csv");
            try {
                FileInputStream fos = new FileInputStream(file);
                DataInputStream dos = new DataInputStream(fos);
                BufferedReader br = new BufferedReader(new InputStreamReader(dos));
                int a = 0;
                String s = null;
                String[] ss;
                try {
                    do{
                        a++;
                        ss = br.readLine().split(",");
                        set.add(ss[0]);
                    }
                    while ((s=br.readLine()) != null) ;
 
                } catch (IOException e) {
                    e.printStackTrace();
                }
                System.out.println("共发现  " + a);
                System.out.println("去重后 " + set.size());
                list=new ArrayList(set);
            } catch (FileNotFoundException e) {
                e.printStackTrace();
            }
        return list;
        }
}

4：实现效果：

       输出为所有用户的推荐细则（ 单机运行很慢大概用了一个小时）

       数据输出格式为（用户id ，推荐商品）