Apache Mahout之协同过滤原理与实践

1,101,5
1,102,3
1,103,4
1,104,4
2,101,3
2,102,1
2,103,2
2,104,3
2,105,3
3,101,4
3,102,3
3,103,4
3,104,3
3,105,5
4,101,3
4,102,3
4,103,1
4,104,5
4,105,4
5,101,1
5,102,5
5,103,5
5,104,2
5,105,1

<?xml version="1.0" encoding="UTF-8"?>
<project xmlns="http://maven.apache.org/POM/4.0.0"
         xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance"
         xsi:schemaLocation="http://maven.apache.org/POM/4.0.0 http://maven.apache.org/xsd/maven-4.0.0.xsd">
    <modelVersion>4.0.0</modelVersion>
    <groupId>com.leboop</groupId>
    <artifactId>mahout</artifactId>
    <version>1.0-SNAPSHOT</version>
    <properties>
        <!-- mahout版本号 -->
        <mahout.version>0.13.0</mahout.version>
        <!-- scala版本号 -->
        <scala.version>2.10</scala.version>
    </properties>
    <dependencies>
        <!-- mahout -->
        <dependency>
            <groupId>org.apache.mahout</groupId>
            <artifactId>mahout-math</artifactId>
            <version>${mahout.version}</version>
        </dependency>
        <dependency>
            <groupId>org.apache.mahout</groupId>
            <artifactId>mahout-integration</artifactId>
            <version>${mahout.version}</version>
        </dependency>
        <!-- mahout-scala -->
        <dependency>
            <groupId>org.apache.mahout</groupId>
            <artifactId>mahout-math-scala_${scala.version}</artifactId>
            <version>${mahout.version}</version>
        </dependency>
        <!-- mahout-spark -->
        <dependency>
            <groupId>org.apache.mahout</groupId>
            <artifactId>mahout-spark_${scala.version}</artifactId>
            <version>${mahout.version}</version>
        </dependency>
    </dependencies>
</project>

package com.leboop.recommendation;

import org.apache.mahout.cf.taste.impl.common.LongPrimitiveIterator;
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.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.util.Arrays;
import java.util.List;

/**
 * 推荐思路
 * 1、读取用户-物品-评分数据，转换成推荐数据模型
 * 2、基于用户相似度计算用户N-最近邻
 * 3、使用推荐引擎推荐物品
 */
public class BasedUserRecommendationTest {
    public static void main(String[] args) {
        //用户-物品-评分数据文件
        String filePath = "data\\alice.txt";
        //数据模型
        DataModel dataModel = null;
        try {
            //文件数据转换成数据模型
            dataModel = new FileDataModel(new File(filePath));
            /**
             * 用户相似度定义
             */
            //余弦相似度
//          UserSimilarity userSimilarity= new UncenteredCosineSimilarity(dataModel);
            //欧几里得相似度
//          UserSimilarity userSimilarity= new EuclideanDistanceSimilarity(dataModel);
            //皮尔森相似度
            UserSimilarity userSimilarity = new PearsonCorrelationSimilarity(dataModel);

            //定义用户的2-最近邻
            UserNeighborhood userNeighborhood = new NearestNUserNeighborhood(2, userSimilarity, dataModel);
            //定义推荐引擎
            Recommender recommender = new GenericUserBasedRecommender(dataModel,userNeighborhood, userSimilarity);
            //从数据模型中获取所有用户ID迭代器
            LongPrimitiveIterator usersIterator = dataModel.getUserIDs();
            //通过迭代器遍历所有用户ID
            while (usersIterator.hasNext()) {
              System.out.println("================================================");
                //用户ID
                long userID = usersIterator.nextLong();
                //用户ID
                LongPrimitiveIterator otherusersIterator = dataModel.getUserIDs();
                //遍历用户ID，计算任何两个用户的相似度
                while (otherusersIterator.hasNext()) {
                    Long otherUserID = otherusersIterator.nextLong();
                    System.out.println("用户 " + userID + " 与用户 " + otherUserID + " 的相似度为："
                            + userSimilarity.userSimilarity(userID, otherUserID));
                }
                //userID的N-最近邻
                long[] userN = userNeighborhood.getUserNeighborhood(userID);
                //用户userID的推荐物品，最多推荐两个
                List<RecommendedItem> recommendedItems = recommender.recommend(userID, 2);
                System.out.println("用户 "+userID + " 的2-最近邻是 "+ Arrays.toString(userN));
                if (recommendedItems.size() > 0) {
                    for (RecommendedItem item : recommendedItems) {
                        System.out.println("推荐的物品"+ item.getItemID()+"预测评分是 "+ item.getValue());
                    }
                } else {
                    System.out.println("无任何物品推荐");
                }
            }
        } catch (Exception e) {
            e.printStackTrace();
        }
    }
}

package com.leboop.recommendation;

import org.apache.mahout.cf.taste.impl.common.LongPrimitiveIterator;
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.GenericItemBasedRecommender;
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.impl.similarity.UncenteredCosineSimilarity;
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.ItemSimilarity;
import org.apache.mahout.cf.taste.similarity.UserSimilarity;

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

/**
 * 推荐思路
 * 1、读取用户-物品-评分数据，转换成推荐数据模型
 * 2、定义物品相似度（余弦相似度、皮尔森相似度等）
 * 3、预测评分
 * 4、使用推荐引擎推荐物品
 */
public class BasedItemRecommendationTest {
    public static void main(String[] args) {
        //用户-物品-评分数据文件
        String filePath = "data\\alice.txt";
        //数据模型
        DataModel dataModel = null;
        try {
            //文件数据转换成数据模型
            dataModel = new FileDataModel(new File(filePath));
            /**
             * 物品相似度定义
             */
            //余弦相似度
            ItemSimilarity itemSimilarity = new UncenteredCosineSimilarity(dataModel);
            //欧几里得相似度
//            ItemSimilarity itemSimilarity= new EuclideanDistanceSimilarity(dataModel);
//            //皮尔森相似度
//            ItemSimilarity itemSimilarity = new PearsonCorrelationSimilarity(dataModel);
            //定义推荐引擎
            Recommender recommender =new GenericItemBasedRecommender(dataModel, itemSimilarity);
            //获取物品迭代器
            LongPrimitiveIterator itemIDIterator = dataModel.getItemIDs();
            //遍历所有物品
            while(itemIDIterator.hasNext()){
                System.out.println("==================================================");
                Long itermID=itemIDIterator.next();
                LongPrimitiveIterator otherItemIDIterator=dataModel.getItemIDs();
                //打印物品相似度
                while (otherItemIDIterator.hasNext()){
                    Long otherItermID=otherItemIDIterator.next();
                    System.out.println("物品 "+itermID+" 与物品 "+otherItermID+" 的相似度为： "+itemSimilarity.itemSimilarity(itermID,otherItermID));
                }
            }
            //获取用户迭代器
            LongPrimitiveIterator userIDIterator =dataModel.getUserIDs();
            //遍历用户
            while(userIDIterator.hasNext()){
                //获取用户
                Long userID=userIDIterator.next();
                //获取用户userID的推荐列表
                List<RecommendedItem> itemList= recommender.recommend(userID,2);
                if(itemList.size()>0){
                    for(RecommendedItem item:itemList){
                        System.out.println("用户 "+userID+" 推荐物品 "+item.getItemID()+",物品评分 "+item.getValue());
                    }
                }else {
                    System.out.println("用户 "+userID+" 无任何物品推荐");
                }
            }
        } catch (Exception e) {
            e.printStackTrace();
        }
    }
}

hadoop jar mahout-examples-0.13.0-job.jar

org.apache.mahout.cf.taste.hadoop.item.RecommenderJob

-i /input/mahout-demo/itemdata.data

-o /output/cf/item

-s SIMILARITY_LOGLIKELIHOOD

--tempDir /tmp/cf/item