基于用户评分的电影推荐 Hadoop/spark实现

Mapreduce框架

输入 user1 movie1 rating

mapreduce 1阶段

map(<user>,<movie>,<rating>){

       k2=movie;

        v2=Tuple2(user, rating);

         emit(ke,v2)

 }

//key =movie

// values=List<Tuple2<user,rating>>

reduce(key,values){

          numberofraters=values.size();

         for(Tuple2<user,rating> t2:values){

                  k3=t2.user;

                  v3=Tuple3(key,,t2.rating,numberofraters);

                 emit(k3,v3);

              }

 

 }

mapreduce 第二阶段

//key=user

//value=movie,rating,numberofraters

map(key, value){

Sting [] tokens=value.split(",");

movie=tokens[0];

rating=tokens[1];

numberofraters=tokens[2];

Tuple3<String,Integer,Integer> v3=new Tuple3<movie,rating,numberofraters>;

emit(key,v3)

 }

//key=user

//values=list(Tuple3(movie,rating,numberofraters))

 

reduce(key,values){

List(Tuple2(Tuple3<movie1>,<rating1>,<numberofraters1>,Tuple3<movie2>,<rating2>,<numberofraters2>)) list=

    generateUniqueCombinations(values);

for(Tuple2(Tuple3<movie1>,<rating1>,<numberofraters1>,Tuple3<movie2>,<rating2>,<numberofraters2>) pair :list){

     m1=pair._1;m2=pair._2;

     reducekey=Tuple2(m1.movie,m2.movie);

     int ratingproduct=m1.rating*m2.rating;

     int rating1squared= m1.rating*m1.rating;

     int rating2squared=m2.rating*m2.rating;

     reducevalue=Tupl7(m1.rating,m1.numberofratins,m2.rating,m2.numberofraters,ratingproduct,rating1squared,rating2squred);

    emit(reducekey,reducevalue);

 }

mapreduce阶段3

map(key,value){emit(key,value)}

 

reduce(key, values){

    int groupsize=value.size();

    int dotproduct=0;

   int