基于内容的推荐算法实现

Job1.java

package com.oracle.moviecf;

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

import org.apache.hadoop.conf.Configuration;
import org.apache.hadoop.fs.Path;
import org.apache.hadoop.io.LongWritable;
import org.apache.hadoop.io.Text;
import org.apache.hadoop.mapreduce.Job;
import org.apache.hadoop.mapreduce.Mapper;
import org.apache.hadoop.mapreduce.Reducer;
import org.apache.hadoop.mapreduce.lib.input.FileInputFormat;
import org.apache.hadoop.mapreduce.lib.output.FileOutputFormat;

/**
 * 与Job2共同完成：
 * 	用户对电影的评价矩阵 × 电影特征建模矩阵 = 用户对某种特征的偏好权重
 * （矩阵的相乘）
 * */
public class Job1 {
	public static class Map extends Mapper<LongWritable, Text, Text, Text> {
		@Override
		protected void map(LongWritable key, Text value, Context context)
				throws IOException, InterruptedException {
			String line = value.toString();
			String split[];
			String val;
			String mvID;
			String userID;
			
			if(line.startsWith("u")) {
				split = line.split("\t");
				userID = split[0];
				for(int i = 1; i < split.length; i++) {
					val = split[i].substring(split[i].indexOf("_")+1);
					mvID = split[i].substring(0, split[i].indexOf("_"));
					context.write(new Text(mvID), new Text(userID+"_"+val));
				}
			}
			if(line.startsWith("m")) {
				split = line.split("\t");
				for(int i = 1; i < split.length; i++) {
					context.write(new Text(split[0]), new Text(split[i]));
				}
			}
			
		}
	}
	
	public static class Reduce extends Reducer<Text, Text, Text, Text> {
		@Override
		protected void reduce(Text key, Iterable<Text> values, Context context)
				throws IOException, InterruptedException {
			List<String> list_user_val = new ArrayList<>();
			List<String> list_tag_val = new ArrayList<>();
			for(Text i: values) {
				if(i.toString().startsWith("u")) {
					list_user_val.add(i.toString());
				}else {
					list_tag_val.add(i.toString());
				}
			}
			
			String userID;
			double userVal;
			String tagID;
			double tagVal;
			for(String i: list_user_val) {
				userID = i.substring(0, i.indexOf("_"));
				userVal = Double.parseDouble(i.substring(i.indexOf("_")+1));
				for(String j: list_tag_val) {
					tagID = j.substring(0, j.indexOf("_"));
					tagVal = Double.parseDouble(j.substring(j.indexOf("_")+1));
					context.write(new Text(userID+"_"+tagID), new Text(String.valueOf(userVal*tagVal)));
				}
			}
		}
	}
	
	public static void main(String[] args) throws Exception {
		Configuration conf = new Configuration();
		
		Job job = Job.getInstance(conf);
		job.setJarByClass(Job1.class);
		job.setMapperClass(Map.class);
		job.setReducerClass(Reduce.class);
		
		job.setMapOutputKeyClass(Text.class);
		job.setMapOutputValueClass(Text.class);
		job.setOutputKeyClass(Text.class);
		job.setOutputValueClass(Text.class);
		
		FileInputFormat.setInputPaths(job, new Path("/moviecf/input"));
		FileOutputFormat.setOutputPath(job, new Path("/moviecf/output1"));
		
		job.waitForCompletion(true);
	}
}

Job2.java

package com.oracle.moviecf;

import java.io.IOException;
import java.util.ArrayList;
import java.util.Collections;
import java.util.HashMap;
import java.util.List;

import org.apache.hadoop.conf.Configuration;
import org.apache.hadoop.fs.Path;
import org.apache.hadoop.io.LongWritable;
import org.apache.hadoop.io.Text;
import org.apache.hadoop.mapreduce.Job;
import org.apache.hadoop.mapreduce.Mapper;
import org.apache.hadoop.mapreduce.Reducer;
import org.apache.hadoop.mapreduce.lib.input.FileInputFormat;
import org.apache.hadoop.mapreduce.lib.output.FileOutputFormat;

/**
 * 与Job1共同完成
 * 目的：
 * 	用户对电影的评价矩阵 × 电影特征建模矩阵 = 用户对某种特征的偏好权重
 * */
public class Job2 {
	public static class Map extends Mapper<LongWritable, Text, Text, Text> {
		@Override
		protected void map(LongWritable key, Text value, Context context)
				throws IOException, InterruptedException {
			
			String line = value.toString();
			String split[] = line.split("\t");
			String userID;
			String tagID;
			
			userID = split[0].substring(0, split[0].indexOf("_"));
			tagID = split[0].substring(split[0].indexOf("_")+1);
			context.write(new Text(userID), new Text(tagID+"_"+split[1]));
		}
	}
	
	public static class Reduce extends Reducer<Text, Text, Text, Text> {
		@Override
		protected void reduce(Text key, Iterable<Text> values, Context context)
				throws IOException, InterruptedException {
			
			HashMap<String, Double> map = new HashMap<>();
			List<String> list = new ArrayList<>();
			String tagID;
			double val;
			
			for(Text i: values) {
				tagID = i.toString().substring(0, i.toString().indexOf("_"));
				val = Double.parseDouble(i.toString().substring(i.toString().indexOf("_")+1));
				if(!map.containsKey(tagID)) {
					map.put(tagID, val);
					list.add(tagID);
				}else {
					map.put(tagID, val+map.get(tagID));
				}
			}
			
			StringBuffer line = new StringBuffer();
			Collections.sort(list);
			for(String i: list) {
				line.append(i+"_"+map.get(i)+"\t");
			}
			
			context.write(key, new Text(line.toString()));
		}
	}
	
	public static void main(String[] args) throws Exception {
		Configuration conf = new Configuration();
		
		Job job = Job.getInstance(conf);
		job.setJarByClass(Job2.class);
		job.setMapperClass(Map.class);
		job.setReducerClass(Reduce.class);
		
		job.setMapOutputKeyClass(Text.class);
		job.setMapOutputValueClass(Text.class);
		job.setOutputKeyClass(Text.class);
		job.setOutputValueClass(Text.class);
		
		FileInputFormat.setInputPaths(job, new Path("/moviecf/output1"));
		FileOutputFormat.setOutputPath(job, new Path("/moviecf/output2"));
		
		job.waitForCompletion(true);
	}
}

Job3.java

package com.oracle.moviecf;

import java.io.IOException;
import java.util.ArrayList;
import java.util.HashMap;
import java.util.List;

import org.apache.hadoop.conf.Configuration;
import org.apache.hadoop.fs.Path;
import org.apache.hadoop.io.LongWritable;
import org.apache.hadoop.io.Text;
import org.apache.hadoop.mapreduce.Job;
import org.apache.hadoop.mapreduce.Mapper;
import org.apache.hadoop.mapreduce.Reducer;
import org.apache.hadoop.mapreduce.lib.input.FileInputFormat;
import org.apache.hadoop.mapreduce.lib.output.FileOutputFormat;/**
 * 相似度
 * 与Job4同用
 * 根据 用户对特征的偏好权重矩阵 与 电影特征建模矩阵 ，求用户对电影的余弦相似度，即得到每个用户对每个电影的喜欢程度
 * 
 * */
public class Job3 {
	public static class Map extends Mapper<LongWritable, Text, Text, Text> {
		@Override
		protected void map(LongWritable key, Text value, Context context)
				throws IOException, InterruptedException {
			
			String line = value.toString();
			String split[] = line.split("\t");
			String userID = split[0];
			String tagID;
			double val;
			double a = 0;	//分母
			
			for(int i = 1; i < split.length; i++) {
				val = Double.parseDouble(split[i].substring(split[i].indexOf("_")+1));
				a += val*val;
			}
			
			for(int i = 1; i < split.length; i++) {
				tagID = split[i].substring(0, split[i].indexOf("_"));
				val = Double.parseDouble(split[i].substring(split[i].indexOf("_")+1));
				context.write(new Text(tagID), new Text(userID+"_"+val+"_"+a));
			}
		}
	}
	
	public static class Reduce extends Reducer<Text, Text, Text, Text> {
		@Override
		protected void reduce(Text key, Iterable<Text> values, Context context)
				throws IOException, InterruptedException {
			
			
			List<String> list_user = new ArrayList<>();
			List<String> list_movie = new ArrayList<>();
			String user_a;
			String mv_a;
			double user_val;
			double mv_val;
			String userID;
			String mvID;
			
			for(Text i: values) {
				if(i.toString().startsWith("u")) {
					list_user.add(i.toString());
				}else {
					list_movie.add(i.toString());
				}
			}
			
			String split[];
			for(String i: list_user) {
				split = i.split("_");
				userID = split[0];
				user_val = Double.parseDouble(split[1]);
				user_a = split[2];
				for(String j: list_movie) {
					split = j.split("_");
					mvID = split[0];
					mv_val = Double.parseDouble(split[1]);
					mv_a = split[2];
					context.write(new Text(userID+"_"+mvID), new Text((mv_val*user_val)+"_"+user_a+"_"+mv_a));
				}
			}
			
		
			
		/*for(Text i: values) {
			context.write(key, i);
		}*/
			
		}
	}
	
	public static void main(String[] args) throws Exception {
		Configuration conf = new Configuration();
		
		Job job = Job.getInstance(conf);
		job.setJarByClass(Job3.class);
		job.setMapperClass(Map.class);
		job.setReducerClass(Reduce.class);
		
		job.setMapOutputKeyClass(Text.class);
		job.setMapOutputValueClass(Text.class);
		job.setOutputKeyClass(Text.class);
		job.setOutputValueClass(Text.class);
		
		
		FileInputFormat.setInputPaths(job, new Path("/moviecf/output2"));
		FileInputFormat.addInputPath(job, new Path("/moviecf/input/data1.txt"));
		FileOutputFormat.setOutputPath(job, new Path("/moviecf/output3"));
		
		job.waitForCompletion(true);
	}
}

Job4.java

package com.oracle.moviecf;

import java.io.IOException;
import java.util.ArrayList;
import java.util.Collections;
import java.util.HashMap;
import java.util.List;

import org.apache.hadoop.conf.Configuration;
import org.apache.hadoop.fs.Path;
import org.apache.hadoop.io.LongWritable;
import org.apache.hadoop.io.Text;
import org.apache.hadoop.mapreduce.Job;
import org.apache.hadoop.mapreduce.Mapper;
import org.apache.hadoop.mapreduce.Reducer;
import org.apache.hadoop.mapreduce.lib.input.FileInputFormat;
import org.apache.hadoop.mapreduce.lib.output.FileOutputFormat;

/**
 * 与Job3同用
 * 根据 用户对特征的偏好权重矩阵 与 电影特征建模矩阵 ，求用户对电影的余弦相似度，即得到每个用户对每个电影的喜欢程度
 * */

public class Job4 {
	public static class Map extends Mapper<LongWritable, Text, Text, Text> {
		@Override
		protected void map(LongWritable key, Text value, Context context)
				throws IOException, InterruptedException {
			String line = value.toString();
			String split[] = line.split("\t");
			String userID = line.substring(0, line.indexOf("_"));
			String mvID = split[0].substring(split[0].indexOf("_")+1);
			
			context.write(new Text(userID), new Text(mvID+"\t"+split[1]));
		}
	}
	
	public static class Reduce extends Reducer<Text, Text, Text, Text> {
		@Override
		protected void reduce(Text key, Iterable<Text> values, Context context)
				throws IOException, InterruptedException {
			HashMap<String, Double> map_a = new HashMap<>();
			HashMap<String, Double> map_b = new HashMap<>();
			List<String> keyList = new ArrayList<>();
			
			String[] split;
			String[] stemp;
			double a;
			double b;
			for(Text i: values) {
				split = i.toString().split("\t");
				stemp = split[1].split("_");
				a = Math.sqrt(Double.parseDouble(stemp[1])*Double.parseDouble(stemp[2]));
				b = Double.parseDouble(stemp[0]);
				if(!map_a.containsKey(split[0])) {
					map_a.put(split[0], a);
					map_b.put(split[0], b);
					keyList.add(split[0]);
				}else {
					map_b.put(split[0], b+map_b.get(split[0]));
				}
			}
			
			Collections.sort(keyList);
			StringBuffer result = new StringBuffer();
			String userID = key.toString();
			double val;
			for(String i: keyList) {
				val = map_b.get(i) / map_a.get(i);
				result.append(i+"_"+String.format("%.1f", val)+"\t");
			}
			context.write(key, new Text(result.toString()));
		}
	}
	
	public static void main(String[] args) throws Exception {
	Configuration conf = new Configuration();
		
		Job job = Job.getInstance(conf);
		job.setJarByClass(Job4.class);
		job.setMapperClass(Map.class);
		job.setReducerClass(Reduce.class);
		
		job.setMapOutputKeyClass(Text.class);
		job.setMapOutputValueClass(Text.class);
		job.setOutputKeyClass(Text.class);
		job.setOutputValueClass(Text.class);
		
		
		FileInputFormat.setInputPaths(job, new Path("/moviecf/output3"));
		FileOutputFormat.setOutputPath(job, new Path("/moviecf/output4"));
		
		job.waitForCompletion(true);
	}
}