协同过滤算法(天池竞赛试题)

一:推荐算法分类：

　　1.按数据使用划分：

- 协同过滤算法：UserCF, ItemCF, ModelCF
- 基于内容的推荐: 用户内容属性和物品内容属性
- 社会化过滤：基于用户的社会网络关系

　　2.案例：天池大数据竞赛

　　　　我们会开放如下数据类型：

字段	字段说明	提取说明
user_id	用户标记	抽样&字段加密
Time	行为时间	精度到天级别&隐藏年份
action_type	用户对品牌的行为类型	包括点击、购买、加入购物车、收藏4种行为 (点击：0 购买：1 收藏：2 购物车：3）
brand_id	品牌数字ID	抽样&字段加密

提供的数据量，涉及千万级天猫用户，万级天猫品牌，时间跨度4个月的行为记录。
提供的训练数据在天池集群的表t_alibaba_bigdata_user_brand_total_1中，字段分别为：user_id，brand_id, type, visit_datetime。如图所示

3.用户4种行为类型(Type)对应代码分别为：
点击：0；购买：1；收藏：2；购物车：3

二:实现思路及代码

1、对原数据去重

 1 package com.oracle.www.TianChi_compition;
 2 
 3 import java.io.IOException;
 4 
 5 import org.apache.hadoop.conf.Configuration;
 6 import org.apache.hadoop.fs.FileSystem;
 7 import org.apache.hadoop.fs.Path;
 8 import org.apache.hadoop.io.LongWritable;
 9 import org.apache.hadoop.io.NullWritable;
10 import org.apache.hadoop.io.Text;
11 import org.apache.hadoop.mapreduce.Job;
12 import org.apache.hadoop.mapreduce.Mapper;
13 import org.apache.hadoop.mapreduce.Reducer;
14 import org.apache.hadoop.mapreduce.lib.input.FileInputFormat;
15 import org.apache.hadoop.mapreduce.lib.output.FileOutputFormat;
16 
17 /*
18  * 对原数据去重,去表头
19  */
20 public class Step01 {
21     static class MyMapper extends Mapper<LongWritable, Text, Text, NullWritable> {
22         @Override
23         protected void map(LongWritable key, Text value, Mapper<LongWritable, Text, Text, NullWritable>.Context context)
24                 throws IOException, InterruptedException {
25             if (key.get() > 0) {
26                 context.write(value, NullWritable.get());
27             }
28         }
29     }
30 
31     static class MyReducer extends Reducer<Text, NullWritable, Text, NullWritable> {
32         @Override
33         protected void reduce(Text key, Iterable<NullWritable> vlue,
34                 Reducer<Text, NullWritable, Text, NullWritable>.Context context)
35                 throws IOException, InterruptedException {
36             context.write(key, NullWritable.get());
37         }
38     }
39 
40     public static void main(String[] args)
41             throws IllegalArgumentException, IOException, ClassNotFoundException, InterruptedException {
42         Configuration conf = new Configuration();
43         Job job = Job.getInstance(conf);
44         job.setJarByClass(Step01.class);
45 
46         job.setMapperClass(MyMapper.class);
47         job.setReducerClass(MyReducer.class);
48 
49         job.setMapOutputKeyClass(Text.class);
50         job.setMapOutputValueClass(NullWritable.class);
51 
52         job.setOutputKeyClass(Text.class);
53 
54         Path outPath = new Path("hdfs://192.168.9.13:8020/deweight");
55         FileSystem fs = outPath.getFileSystem(conf);
56         if (fs.exists(outPath)) {
57             fs.delete(outPath, true);
58         }
59         FileInputFormat.addInputPath(job, new Path("hdfs://192.168.9.13:8020/TianmaoData"));
60         FileOutputFormat.setOutputPath(job, outPath);
61         job.waitForCompletion(true);
62 
63     }
64 
65 }

2、获得所有物品之间的同现矩阵

  1 package com.oracle.www.TianChi_compition;
  2 
  3 import java.io.IOException;
  4 import java.util.ArrayList;
  5 
  6 import org.apache.hadoop.conf.Configuration;
  7 import org.apache.hadoop.fs.FileSystem;
  8 import org.apache.hadoop.fs.Path;
  9 import org.apache.hadoop.io.IntWritable;
 10 import org.apache.hadoop.io.Text;
 11 import org.apache.hadoop.mapreduce.Job;
 12 import org.apache.hadoop.mapreduce.Mapper;
 13 import org.apache.hadoop.mapreduce.Reducer;
 14 import org.apache.hadoop.mapreduce.lib.input.FileInputFormat;
 15 import org.apache.hadoop.mapreduce.lib.input.KeyValueTextInputFormat;
 16 import org.apache.hadoop.mapreduce.lib.output.FileOutputFormat;
 17 
 18 /*
 19  * 生成同现(显)矩阵
 20  * map端<商品1-商品2,1>拆分,发送
 21  * reduce端<商品1-商品2,1,1,1...>统计
 22  */
 23 public class Step03 {
 24 
 25     static class MyMapper extends Mapper<Text, Text, Text, IntWritable> {
 26         Text k = new Text();
 27         IntWritable v = new IntWritable();
 28 
 29         @Override
 30         protected void map(Text key, Text value, Mapper<Text, Text, Text, IntWritable>.Context context)
 31                 throws IOException, InterruptedException {
 32             ArrayList<String> itemList = new ArrayList<>();
 33             String line = value.toString();
 34             String[] datas = line.split("\t");
 35             for (String data : datas) {// 将用户购买过的商品添加到list集合中
 36                 String[] item_mark = data.split(":");
 37                 itemList.add(item_mark[0]);
 38             }
 39 
 40             for (int i = 0; i < itemList.size(); i++) {
 41                 for (int j = 0; j < itemList.size(); j++) {
 42                     k.set(itemList.get(i) + "-" + itemList.get(j));
 43                     v.set(1);
 44                     context.write(k, v);
 45                 }
 46 
 47             }
 48         }
 49     }
 50 
 51     static class MyReducer extends Reducer<Text, IntWritable, Text, IntWritable> {
 52         Text k = new Text();
 53         IntWritable v = new IntWritable();
 54 
 55         @Override
 56         protected void reduce(Text key, Iterable<IntWritable> value,
 57                 Reducer<Text, IntWritable, Text, IntWritable>.Context context)
 58                 throws IOException, InterruptedException {
 59             int sum = 0;
 60             for (IntWritable val : value) {
 61                 sum += val.get();
 62             }
 63             k.set(key.toString());
 64             v.set(sum);
 65             context.write(k, v);
 66         }
 67     }
 68 
 69     public static void main(String[] args) throws ClassNotFoundException, InterruptedException {
 70         Configuration conf = new Configuration();
 71         try {
 72             Job job = Job.getInstance(conf);
 73 
 74             job.setJarByClass(Step03.class);
 75             job.setMapperClass(MyMapper.class);
 76             job.setReducerClass(MyReducer.class);
 77 
 78             job.setMapOutputKeyClass(Text.class);
 79             job.setMapOutputValueClass(IntWritable.class);
 80 
 81             job.setOutputKeyClass(Text.class);
 82             job.setOutputValueClass(IntWritable.class);
 83 
 84             job.setInputFormatClass(KeyValueTextInputFormat.class);
 85 
 86             // 判断output文件夹是否存在，如果存在则删除
 87             Path outPath = new Path("hdfs://192.168.9.13:8020/implyCount");// 输出路径
 88             FileSystem fs = outPath.getFileSystem(conf);// 根据输出路径找到文件,参数为配置文件
 89             if (fs.exists(outPath)) {
 90                 fs.delete(outPath);
 91                 // fs.delete(outPath, true);true的意思是，就算output有东西，也一带删除,默认为true
 92 
 93             }
 94             FileInputFormat.setInputPaths(job, new Path("hdfs://192.168.9.13:8020/gradeMarking"));
 95             FileOutputFormat.setOutputPath(job, outPath);
 96             job.waitForCompletion(true);
 97         } catch (IOException e) {
 98             // TODO Auto-generated catch block
 99             e.printStackTrace();
100         }
101     }
102 }

3、权重矩阵(用户对同一件商品的不同行为操作得到的评分矩阵)

  1 package com.oracle.www.TianChi_compition;
  2 
  3 import java.io.IOException;
  4 import java.util.HashMap;
  5 import java.util.Iterator;
  6 
  7 import org.apache.hadoop.conf.Configuration;
  8 import org.apache.hadoop.fs.FileSystem;
  9 import org.apache.hadoop.fs.Path;
 10 /*
 11  * 生成评分矩阵
 12  * map端拆分,发送<用户    商品+":"+操作>
 13  * reduce端统计生成<用户    商品1+":"+评分,商品2+":"+评分,...>
 14  */
 15 import org.apache.hadoop.io.LongWritable;
 16 import org.apache.hadoop.io.Text;
 17 import org.apache.hadoop.mapreduce.Job;
 18 import org.apache.hadoop.mapreduce.Mapper;
 19 import org.apache.hadoop.mapreduce.Reducer;
 20 import org.apache.hadoop.mapreduce.lib.input.FileInputFormat;
 21 import org.apache.hadoop.mapreduce.lib.output.FileOutputFormat;
 22 
 23 public class Step02 {
 24     static Text userId = new Text();
 25     static Text shopping_operate = new Text();
 26 
 27     static class MyMapper extends Mapper<LongWritable, Text, Text, Text> {
 28         @Override
 29         protected void map(LongWritable key, Text value, Mapper<LongWritable, Text, Text, Text>.Context context)
 30                 throws IOException, InterruptedException {
 31 
 32             String line = value.toString();
 33             String[] datas = line.split("\t");
 34             userId.set(datas[1]);
 35             shopping_operate.set(datas[0] + ":" + datas[2]);
 36             context.write(userId, shopping_operate);
 37         }
 38     }
 39 
 40     static class MyReducer extends Reducer<Text, Text, Text, Text> {
 41         Text v = new Text();
 42         double click = 0;
 43         double collect = 0;
 44         double cart = 0;
 45         double alipay = 0;
 46 
 47         @Override
 48         protected void reduce(Text key, Iterable<Text> value, Reducer<Text, Text, Text, Text>.Context context)
 49                 throws IOException, InterruptedException {
 50             // shoppingOperate_counter<商品,<商品操作,操作次数>>
 51             HashMap<String, HashMap<String, Integer>> shoppingOperate_counter = new HashMap<>();
 52             String[] temp_str = null;
 53             String shoppingName = null;
 54             String shoppingOperate = null;
 55             HashMap<String, Integer> operate_counter = null;// 内层map,记录对商品的操作和操作次数
 56             for (Text val : value) {
 57                 temp_str = val.toString().split(":");
 58                 shoppingName = temp_str[0];
 59                 shoppingOperate = temp_str[1];
 60                 if (!shoppingOperate_counter.containsKey(shoppingName)) {// map中不存在此商品信息,添加并给予初始值
 61                     operate_counter = new HashMap<>();
 62                     operate_counter.put(shoppingOperate, 1);
 63                     shoppingOperate_counter.put(shoppingName, operate_counter);
 64                 } else {// map中包含此商品
 65                     operate_counter = shoppingOperate_counter.get(shoppingName);
 66                     if (!operate_counter.containsKey(shoppingOperate)) {// 包含此商品不包含此操作
 67                         operate_counter.put(shoppingOperate, 1);
 68                     } else {
 69                         operate_counter.put(shoppingOperate, operate_counter.get(shoppingOperate) + 1);
 70                     }
 71                 }
 72             }
 73             // 通过对shoppingOperate_counter循环遍历,统计算分
 74             Iterator<String> iter = shoppingOperate_counter.keySet().iterator();
 75             StringBuffer shopping_marking = new StringBuffer();
 76             while (iter.hasNext()) {
 77                 click = 0;
 78                 collect = 0;
 79                 cart = 0;
 80                 alipay = 0;
 81                 shoppingName = iter.next();
 82                 operate_counter = shoppingOperate_counter.get(shoppingName);
 83                 Iterator<String> operateIter = operate_counter.keySet().iterator();
 84                 int counter = 0;// 记录用户对单个商品操作过的次数
 85                 while (operateIter.hasNext()) {
 86                     counter++;
 87                     shoppingOperate = operateIter.next();
 88                     if ("click".equals(shoppingOperate)) {
 89                         click = operate_counter.get(shoppingOperate);
 90                     } else if ("collect".equals(shoppingOperate)) {
 91                         collect = operate_counter.get(shoppingOperate);
 92                     } else if ("cart".equals(shoppingOperate)) {
 93                         cart = operate_counter.get(shoppingOperate);
 94                     } else {
 95                         alipay = operate_counter.get(shoppingOperate);
 96                     }
 97                 }
 98                 double sum = click / counter * 1.0 + collect / counter * 2.0 + cart / counter * 3.0
 99                         + alipay / counter * 4.0;
100                 shopping_marking.append(shoppingName + ":" + sum + "\t");
101             }
102             v.set(shopping_marking.toString());
103             context.write(key, v);
104         }
105     }
106 
107     public static void main(String[] args) throws ClassNotFoundException, InterruptedException {
108         Configuration conf = new Configuration();
109         try {
110             Job job = Job.getInstance(conf);
111             job.setJarByClass(Step02.class);
112             job.setMapperClass(MyMapper.class);
113             job.setReducerClass(MyReducer.class);
114 
115             job.setMapOutputKeyClass(Text.class);
116             job.setMapOutputValueClass(Text.class);
117 
118             job.setOutputKeyClass(Text.class);
119             job.setOutputKeyClass(Text.class);
120 
121             Path outPath = new Path("hdfs://192.168.9.13:8020/deweight");
122             FileSystem fs = outPath.getFileSystem(conf);
123             if (fs.exists(outPath)) {
124                 fs.delete(outPath);
125             }
126             FileInputFormat.setInputPaths(job, new Path("hdfs://192.168.9.13:8020/deweight"));
127             FileOutputFormat.setOutputPath(job, outPath);
128             job.waitForCompletion(true);
129         } catch (IOException e) {
130             // TODO Auto-generated catch block
131             e.printStackTrace();
132         }
133 
134     }
135 
136 }

4、两个矩阵相乘得到三维矩阵

  1 package com.oracle.www.TianChi_compition;
  2 
  3 import java.io.IOException;
  4 import java.util.HashMap;
  5 import java.util.Map.Entry;
  6 
  7 import org.apache.hadoop.conf.Configuration;
  8 import org.apache.hadoop.fs.FileSystem;
  9 import org.apache.hadoop.fs.Path;
 10 import org.apache.hadoop.io.DoubleWritable;
 11 import org.apache.hadoop.io.Text;
 12 import org.apache.hadoop.mapreduce.Job;
 13 import org.apache.hadoop.mapreduce.Mapper;
 14 import org.apache.hadoop.mapreduce.Reducer;
 15 import org.apache.hadoop.mapreduce.lib.input.FileInputFormat;
 16 import org.apache.hadoop.mapreduce.lib.input.FileSplit;
 17 import org.apache.hadoop.mapreduce.lib.input.KeyValueTextInputFormat;
 18 import org.apache.hadoop.mapreduce.lib.output.FileOutputFormat;
 19 
 20 public class Step04 {
 21     static class MyMapper extends Mapper<Text, Text, Text, Text> {
 22         String parentName = null;
 23         Text k = new Text();
 24         Text v = new Text();
 25 
 26         @Override
 27         protected void setup(Mapper<Text, Text, Text, Text>.Context context) throws IOException, InterruptedException {
 28             FileSplit fs = (FileSplit) context.getInputSplit();
 29             parentName = fs.getPath().getParent().getName();
 30         }
 31 
 32         @Override
 33         protected void map(Text key, Text value, Mapper<Text, Text, Text, Text>.Context context)
 34                 throws IOException, InterruptedException {
 35             String line = value.toString();
 36             String[] datas = null;
 37             // 判断输入目录
 38             if (parentName.equals("gradeMarking")) {// 评分
 39                 datas = line.split("\t");
 40                 for (String data : datas) {
 41                     String[] item_mark = data.split(":");
 42                     k.set(item_mark[0]);
 43                     v.set(key.toString() + ":" + item_mark[1]);
 44                     context.write(k, v);
 45                 }
 46             } else {
 47                 datas = key.toString().split("-");
 48                 k.set(datas[1]);
 49                 v.set(datas[0] + ":" + line);
 50                 context.write(k, v);
 51             }
 52         }
 53     }
 54 
 55     static class MyReducer extends Reducer<Text, Text, Text, DoubleWritable> {
 56         Text k = new Text();
 57         DoubleWritable v = new DoubleWritable();
 58         // <商品x 用户1:评分1,用户2:评分2,...,商品1:频次1,商品2:频次2,...>(频次值为两件商品同时出现的次数)
 59 
 60         @Override
 61         protected void reduce(Text key, Iterable<Text> value, Reducer<Text, Text, Text, DoubleWritable>.Context context)
 62                 throws IOException, InterruptedException {
 63             HashMap<String, Double> user_mark = new HashMap<>();
 64             HashMap<String, Double> item_counter = new HashMap<>();
 65             // 将 用户:评分 , 商品:频次 添加到对应的map中
 66             String[] datas = null;
 67             for (Text val : value) {
 68                 datas = val.toString().split(":");
 69                 if (datas[0].startsWith("u")) {
 70                     user_mark.put(datas[0], Double.parseDouble(datas[1]));
 71                 } else {
 72                     item_counter.put(datas[0], Double.parseDouble(datas[1]));
 73                 }
 74             }
 75 
 76             // 遍历循环相乘
 77             String userName = null;
 78             double userMark = 0.0;
 79             String itemName = null;
 80             double iterCounter = 0;
 81             for (Entry<String, Double> entry1 : user_mark.entrySet()) {
 82                 userName = entry1.getKey();
 83                 userMark = entry1.getValue();
 84                 for (Entry<String, Double> entry2 : item_counter.entrySet()) {
 85                     itemName = entry2.getKey();
 86                     iterCounter = entry2.getValue();
 87                     k.set(userName + ":" + itemName);
 88                     v.set(userMark * iterCounter);
 89                     context.write(k, v);
 90                 }
 91             }
 92 
 93         }
 94     }
 95 
 96     public static void main(String[] args) throws ClassNotFoundException, InterruptedException {
 97         Configuration conf = new Configuration();
 98         try {
 99             Job job = Job.getInstance(conf);
100 
101             job.setJarByClass(Step03.class);
102             job.setMapperClass(MyMapper.class);
103             job.setReducerClass(MyReducer.class);
104 
105             job.setMapOutputKeyClass(Text.class);
106             job.setMapOutputValueClass(Text.class);
107 
108             job.setOutputKeyClass(Text.class);
109             job.setOutputValueClass(DoubleWritable.class);
110 
111             job.setInputFormatClass(KeyValueTextInputFormat.class);
112 
113             // 判断output文件夹是否存在，如果存在则删除
114             Path outPath = new Path("hdfs://192.168.9.13:8020/mark&implyCount_multiply");// 输出路径
115             FileSystem fs = outPath.getFileSystem(conf);// 根据输出路径找到文件,参数为配置文件
116             if (fs.exists(outPath)) {
117                 fs.delete(outPath);
118                 // fs.delete(outPath, true);true的意思是，就算output有东西，也一带删除,默认为true
119 
120             }
121             FileInputFormat.setInputPaths(job, new Path[] { new Path("hdfs://192.168.9.13:8020/gradeMarking"),
122                     new Path("hdfs://192.168.9.13:8020/implyCount") });
123             FileOutputFormat.setOutputPath(job, outPath);
124             job.waitForCompletion(true);
125         } catch (IOException e) {
126             // TODO Auto-generated catch block
127             e.printStackTrace();
128         }
129     }
130 
131 }

5、三维矩阵的数据相加获得所有用户对所有物品的推荐值（二维矩阵）

  1 package com.oracle.www.TianChi_compition;
  2 
  3 /*
  4  * 筛选掉用户购买过的商品,并求和
  5  */
  6 import java.io.BufferedReader;
  7 import java.io.FileReader;
  8 import java.io.IOException;
  9 import java.net.URI;
 10 import java.net.URISyntaxException;
 11 import java.util.ArrayList;
 12 
 13 import org.apache.hadoop.conf.Configuration;
 14 import org.apache.hadoop.fs.FileSystem;
 15 import org.apache.hadoop.fs.Path;
 16 import org.apache.hadoop.io.Text;
 17 import org.apache.hadoop.mapreduce.Job;
 18 import org.apache.hadoop.mapreduce.Mapper;
 19 import org.apache.hadoop.mapreduce.Reducer;
 20 import org.apache.hadoop.mapreduce.lib.input.FileInputFormat;
 21 import org.apache.hadoop.mapreduce.lib.input.KeyValueTextInputFormat;
 22 import org.apache.hadoop.mapreduce.lib.output.FileOutputFormat;
 23 
 24 public class Step05 {
 25     static class MyMapper extends Mapper<Text, Text, Text, Text> {
 26         // boughtList集合用于存放哪些用户买过哪些商品,不能使用map集合存放,
 27         // 同一个用户可能买过多件商品,同一件商品也有可能同时被好多人买过;
 28         ArrayList<String> boughtList = new ArrayList<>();
 29         BufferedReader br = null;
 30 
 31         // setup方法初始化boughtList集合
 32         @Override
 33         protected void setup(Mapper<Text, Text, Text, Text>.Context context) throws IOException, InterruptedException {
 34             br = new BufferedReader(new FileReader("part-r-00000"));
 35             String line = null;
 36             String[] datas = null;
 37             while ((line = br.readLine()) != null) {
 38                 datas = line.split("\t");
 39                 if ("alipay".equals(datas[2])) {
 40                     boughtList.add(datas[1] + ":" + datas[0]);
 41                 }
 42             }
 43         }
 44 
 45         // map方法排除掉用户购买过的商品,使其不推荐
 46         @Override
 47         protected void map(Text key, Text value, Mapper<Text, Text, Text, Text>.Context context)
 48                 throws IOException, InterruptedException {
 49             // 判断向该用户推荐的商品是否被该用户购买过,如果购买过,则不推荐(即不向reduce端发送)
 50             if (!boughtList.contains(key.toString())) {
 51                 context.write(key, value);
 52             }
 53         }
 54     }
 55 
 56     static class MyReducer extends Reducer<Text, Text, Text, Text> {
 57         Text k = new Text();
 58         Text v = new Text();
 59 
 60         @Override
 61         protected void reduce(Text key, Iterable<Text> value, Reducer<Text, Text, Text, Text>.Context context)
 62                 throws IOException, InterruptedException {
 63             double rank = 0.0;
 64             for (Text val : value) {
 65                 rank += Double.parseDouble(val.toString());
 66             }
 67             k.set(key.toString().split(":")[0]);
 68             v.set(key.toString().split(":")[1] + ":" + rank);
 69             context.write(k, v);
 70         }
 71     }
 72 
 73     public static void main(String[] args)
 74             throws IOException, ClassNotFoundException, InterruptedException, URISyntaxException {
 75         Configuration conf = new Configuration();
 76         Job job = Job.getInstance(conf);
 77         job.setJarByClass(Step05.class);
 78         job.setMapperClass(MyMapper.class);
 79         job.setReducerClass(MyReducer.class);
 80 
 81         job.setMapOutputKeyClass(Text.class);
 82         job.setMapOutputValueClass(Text.class);
 83 
 84         job.setOutputKeyClass(Text.class);
 85         job.setOutputValueClass(Text.class);
 86 
 87         job.setInputFormatClass(KeyValueTextInputFormat.class);
 88 
 89         job.addCacheFile(new URI("hdfs://192.168.9.13:8020/deweight/part-r-00000"));
 90 
 91         Path outPath = new Path("hdfs://192.168.9.13:8020/shoppingRecommend");
 92         FileSystem fs = outPath.getFileSystem(conf);
 93         if (fs.exists(outPath)) {
 94             fs.delete(outPath, true);
 95         }
 96         FileInputFormat.setInputPaths(job, new Path("hdfs://192.168.9.13:8020/mark&implyCount_multiply"));
 97         FileOutputFormat.setOutputPath(job, outPath);
 98 
 99         job.waitForCompletion(true);
100 
101     }
102 
103 }

6、按照推荐值降序排序(筛选权重高的前十件商品)。

  1 package com.oracle.www.TianChi_compition;
  2 
  3 import java.io.DataInput;
  4 import java.io.DataOutput;
  5 import java.io.IOException;
  6 import java.lang.reflect.InvocationTargetException;
  7 import java.util.ArrayList;
  8 import java.util.Collections;
  9 
 10 import org.apache.commons.beanutils.BeanUtils;
 11 import org.apache.hadoop.conf.Configuration;
 12 import org.apache.hadoop.fs.FileSystem;
 13 import org.apache.hadoop.fs.Path;
 14 import org.apache.hadoop.io.Text;
 15 import org.apache.hadoop.io.WritableComparable;
 16 import org.apache.hadoop.mapreduce.Job;
 17 import org.apache.hadoop.mapreduce.Mapper;
 18 import org.apache.hadoop.mapreduce.Reducer;
 19 import org.apache.hadoop.mapreduce.lib.input.FileInputFormat;
 20 import org.apache.hadoop.mapreduce.lib.input.KeyValueTextInputFormat;
 21 import org.apache.hadoop.mapreduce.lib.output.FileOutputFormat;
 22 
 23 /*
 24  * 排序,选出向用户推荐的前十个商品
 25  */
 26 public class Step06 {
 27     // 将取到一行的内容拆分,以 <用户,商品j:权重r>发送到reduce端进行处理
 28     static class MyMapper extends Mapper<Text, Text, Text, Sort> {
 29         Sort sort = null;
 30 
 31         @Override
 32         protected void map(Text key, Text value, Mapper<Text, Text, Text, Sort>.Context context)
 33                 throws IOException, InterruptedException {
 34 
 35             sort = new Sort(value.toString().split(":")[0], Double.parseDouble(value.toString().split(":")[1]));
 36             context.write(key, sort);
 37         }
 38     }
 39 
 40     // reduce端将同一用户的推荐商品按权值大小排序,将前十个拼接输出
 41     static class MyReducer extends Reducer<Text, Sort, Text, Text> {
 42         ArrayList<Sort> list = new ArrayList<>();
 43         Text v = new Text();
 44 
 45         @Override
 46         protected void reduce(Text key, Iterable<Sort> value, Reducer<Text, Sort, Text, Text>.Context context)
 47                 throws IOException, InterruptedException {
 48             StringBuffer sb = new StringBuffer();
 49             list.clear();
 50             // map端如果将自定义对象作为value发送到reduce端进行迭代时,需要将迭代器中的每个对象使用BeanUtils.copyProperties(dest,org)将属性拷贝到另外一个对象中;
 51             for (Sort sort : value) {
 52                 Sort tempSort = new Sort();
 53                 try {
 54                     BeanUtils.copyProperties(tempSort, sort);
 55                     list.add(tempSort);
 56                 } catch (IllegalAccessException e) {
 57                     // TODO Auto-generated catch block
 58                     e.printStackTrace();
 59                 } catch (InvocationTargetException e) {
 60                     // TODO Auto-generated catch block
 61                     e.printStackTrace();
 62                 }
 63             }
 64 
 65             Collections.sort(list);
 66             for (int i = 0; i < list.size() && i < 10; i++) {
 67                 sb.append(list.get(i));
 68             }
 69             v.set(sb.toString());
 70             context.write(key, v);
 71         }
 72     }
 73 
 74     static public class Sort implements WritableComparable<Sort> {
 75         private String shoppingName;
 76         private double shoppingRank;
 77 
 78         public Sort() {
 79         }
 80 
 81         public Sort(String shoppingName, double shoppingRank) {
 82             this.shoppingName = shoppingName;
 83             this.shoppingRank = shoppingRank;
 84         }
 85 
 86         public String getShoppingName() {
 87             return shoppingName;
 88         }
 89 
 90         public void setShoppingName(String shoppingName) {
 91             this.shoppingName = shoppingName;
 92         }
 93 
 94         public double getShoppingRank() {
 95             return shoppingRank;
 96         }
 97 
 98         public void setShoppingRank(double shoppingRank) {
 99             this.shoppingRank = shoppingRank;
100         }
101 
102         @Override
103         public String toString() {
104             return shoppingName + ":" + shoppingRank + "\t";
105         }
106 
107         @Override
108         public void write(DataOutput out) throws IOException {
109             out.writeDouble(shoppingRank);
110             out.writeUTF(shoppingName);
111         }
112 
113         @Override
114         public void readFields(DataInput in) throws IOException {
115             this.shoppingRank = in.readDouble();
116             this.shoppingName = in.readUTF();
117         }
118 
119         @Override
120         public int compareTo(Sort o) {
121             int temp = 0;
122             if (this.getShoppingRank() - o.getShoppingRank() < 0) {
123                 return 1;
124             } else if (this.getShoppingRank() - o.getShoppingRank() > 0) {
125                 return -1;
126             }
127             return temp;
128         }
129     }
130 
131     public static void main(String[] args) throws ClassNotFoundException, InterruptedException {
132         Configuration conf = new Configuration();
133         try {
134             Job job = Job.getInstance();
135 
136             job.setJarByClass(Step06.class);
137             job.setMapperClass(MyMapper.class);
138             job.setReducerClass(MyReducer.class);
139 
140             job.setMapOutputKeyClass(Text.class);
141             job.setMapOutputValueClass(Sort.class);
142 
143             job.setOutputKeyClass(Text.class);
144             job.setOutputValueClass(Text.class);
145 
146             job.setInputFormatClass(KeyValueTextInputFormat.class);
147 
148             Path outPath = new Path("hdfs://192.168.9.13:8020/ShoppingRecommend_Sort");
149             FileSystem fs = outPath.getFileSystem(conf);
150             if (fs.exists(outPath)) {
151                 fs.delete(outPath);
152             }
153 
154             FileInputFormat.setInputPaths(job, new Path("hdfs://192.168.9.13:8020/shoppingRecommend"));
155             FileOutputFormat.setOutputPath(job, outPath);
156 
157             job.waitForCompletion(true);
158         } catch (IOException e) {
159             // TODO Auto-generated catch block
160             e.printStackTrace();
161         }
162 
163     }
164 
165 }

posted @ 2017-11-04 17:40 TheCompass 阅读(1151) 评论(1) 编辑收藏举报

会员力量，点亮园子希望

刷新页面返回顶部

吾生也有涯,而知也无涯

Hello World!

协同过滤算法(天池竞赛试题)

一:推荐算法分类：

二:实现思路及代码

公告