信息检索——倒排索引的合并，两个词的查询

• 代码实现功能：
  1.构建文档中每个词的倒排索引并输出
  2.输入两个查询词，通过合并倒排索引，输出它们共同的文档ID

• 实现如下：
  1.在工程内添加doc1.txt~doc10.txt文档
  2.代码如下

#include <iostream>
#include <string>
#include <vector>
#include <fstream>
#include <map>
#include <stdlib.h>
#include <sstream>
#include <utility>
using namespace std;
//存储倒排列表信息包括:TF,pos
struct DocMes{
    int TF=0;      //词频
    string pos = "";     //在文档出现的位置
};

typedef map<int,DocMes> invertMap;//倒排列表
typedef map<string,map<int,DocMes> > IndexMap;  //定义multiMap类型
//把单词转为小写
//void judge(string &word)
//把int转为String
string intToString(int a){
    string res;
    stringstream ss;
    ss<<a;
    ss>>res;
    return res;
}
//索引建立，在与代码同文件夹的1.txt 2.txt 3.txt读入文档，分词，记录词语与词语对应倒排列表信息
void CreateIndex(IndexMap &index){
    //读本地txt文档
    ifstream fileIn;
    for(int i=1;i<=10;i++){
        string filepath = "doc"+intToString(i)+".txt";
        fileIn.open(filepath.c_str());
        string temp;
        int posNum = 0;
        while(fileIn>>temp){
            posNum++;
            index[temp][i].TF++;
            index[temp][i].pos +=intToString(posNum)+" , ";
        }


        fileIn.close();
    }
}
//查询处理
void queryProcess(IndexMap &index){
    string query;
    while(true){
        cout<<"请输入查询词：";
        pair <string , string> query;
        cin>>query.first>>query.second;
        if(index.find(query.first)==index.end()){
            cout<<"没有找到"<<query.first<<"!请确认您输入的查询词是否有误"<<endl<<endl;
            continue;
        }
        else if(index.find(query.second)==index.end()){
            cout<<"没有找到"<<query.second<<"!请确认您输入的查询词是否有误"<<endl<<endl;
            continue;
        }
        vector<int> docID;
        invertMap::iterator riter1;
        invertMap::iterator riter2;
        riter1 = index[query.first].begin();
        riter2 = index[query.second].begin();
        while(riter1!=index[query.first].end() && riter2 != index[query.second].end()){      //合并倒排索引
            if (riter1->first == riter2->first){
                docID.push_back(riter1->first);
                riter1++;
                riter2++;
            }else if(riter1->first < riter2->first){
                riter1++;
            }else {
                riter2++;
            }

 }
         if(docID.empty() == 1){          //查询词不在同一文档，输出各自的倒排索引
             cout<<"这两个词在不同的文档，其倒排索引如下"<<endl;
             invertMap::iterator riter1;
             riter1 = index[query.first].begin();
             while (riter1 != index[query.first].end()){
             cout<<query.first<<"  doc"<<riter1->first<<".txt   Tf="<<riter1->second.TF<<"    pos = ( "<<riter1->second.pos<<" )"<<endl;
             riter1++;
             }
             invertMap::iterator riter2;
             riter2 = index[query.second].begin();
             while (riter2 != index[query.second].end()){
             cout<<query.second<<"  doc"<<riter2->first<<".txt   Tf="<<riter2->second.TF<<"    pos = ( "<<riter2->second.pos<<" )"<<endl;
             riter2++;
             }
         }else {
         for (int i = 0;i<docID.size();i++){
                cout<<"查询词所在文档为：doc"<<docID[i]<<".txt "<<endl<<endl;
            }
         }
    }

}
//输出倒排索引
int printinverse(IndexMap &index){

    IndexMap::iterator iter;//索引表迭代器
    invertMap::iterator riter;//倒排记录表迭代器
    iter = index.begin();

    while (iter != index.end()){
        string term = iter->first;
        cout<<endl<<term<<endl;
        riter = index[term].begin();
    while (riter != index[term].end()){
        cout<<"doc"<<riter->first<<".txt   Tf="<<riter->second.TF<<"    pos = ( "<<riter->second.pos<<" )"<<endl;
            riter++;
    }
    iter++;
}
}
int main(){
    IndexMap invertIndex;
    CreateIndex(invertIndex);
    cout<<"建立倒排索引如下:"<<endl;
    printinverse(invertIndex);
    cout<<endl;
    queryProcess(invertIndex);
    return 0;
}

实现主要通过c++模板类map和pair。

• 实现效果如下