java大作业 KShinglingAlgorithm

wiki上关于KShingling Algorithm（w-shingling）的说明：

http://en.wikipedia.org/wiki/W-shingling

摘要：

In natural language processing a w-shingling is a set of unique "shingles"—contiguous subsequences of tokens in a document—that can be used to gauge the similarity of two documents. The w denotes the number of tokens in each shingle in the set.

The document, "a rose is a rose is a rose" can be tokenized as follows:

(a,rose,is,a,rose,is,a,rose)

The set of all contiguous sequences of 4 tokens (N-grams, here: 4-grams) is

{ (a,rose,is,a), (rose,is,a,rose), (is,a,rose,is), (a,rose,is,a), (rose,is,a,rose) } = { (a,rose,is,a), (rose,is,a,rose), (is,a,rose,is) }

我理解的此算法，是把每段文本都像上述分解后，统计两段文本的合集b，再统计交集a，用a/b得到相似度。

写得有些复杂：

package bigproject2;

import javax.swing.JOptionPane;

public class union {
    //求子集
    public String[] ziji(String str)
    {
        char[] ch=str.toCharArray();
        int c=0;
        for(int i=0;i<ch.length;i++)
        {
            if(ch[i]==' ')
                c++;
        }
        //建立单词数组
        String[] strt=new String[c+1];
        for(int i=0;i<c+1;i++)
            strt[i]="";
        int h=0;
        for(int i=0;i<c+1;i++)
        {
            for(int j=h;j<ch.length;j++)
            {
                if(ch[j]==' ')
                {
                    h=j+1;
                    break;
                }
                else strt[i]+=ch[j];
            }
        }
        return strt;
    }
    //按k分，并去掉重复子集。
    public String[] cut(String[] str,int k) throws MyException{
        if(str.length<k)
                throw new MyException("单词数少于"+k+"，无法进行计算！");
        String[] t=new String[str.length-k+1];
        for(int i=0;i<str.length-k+1;i++)
            t[i]="";
        int h=0,m=0;
        for(;h<str.length-k+1;h++)
        {
            for(int i=m;i<m+k;i++)
                t[h]+=str[i];
            m++;
        }
        //去掉重复部分
        int merge=0;
        for(int i=0;i<t.length-1;i++)
        {
            if(t[i].equals("")) break;
            for(int j=i+1;j<t.length;j++)
            {
                if(t[i].equals(t[j]))
                {
                    merge++;
                    int y=j;
                    for(;y<t.length-1;y++)
                    {
                        t[y]=t[y+1];
                    }
                    t[y]="";
                }
            }
        }
        String[] fin=new String[t.length-merge];
        for(int i=0;i<t.length-merge;i++)
            fin[i]=t[i];
        return fin;
    }
    public class MyException extends Exception{
        public MyException(String str){
            JOptionPane.showMessageDialog(null, str,"警告", JOptionPane.INFORMATION_MESSAGE);
        }
    }
    //求两字符串数组合集个数。
    public int heji(String[] a,String[] b){
        int count=a.length+b.length;
        for(int i=0;i<a.length;i++)
        {
            for(int j=0;j<b.length;j++)
            {
                if(a[i].equals(b[j]))
                    count--;
            }
        }
        return count;
    }
    //求两字符串数组交集个数。
    public int jiaoji(String[] a,String[] b){
        int count=0;
        for(int i=0;i<a.length;i++)
        {
            for(int j=0;j<b.length;j++)
            {
                if(a[i].equals(b[j]))
                    count++;
            }
        }
        return count;
    }

}

package bigproject2;


public class KShinglingAlgorithm extends union{
    private String text1,text2;
    public String getText1()
    {
        return text1;
    }
    public String getText2()
    {
        return text2;
    }
    public void setText1(String text1)
    {
        this.text1=text1;
    }
    public void setText2(String text2)
    {
        this.text2=text2;
    }
    
    public float getSimilarity(int k)
    {
       union a=new union();
       String[] t1=a.ziji(this.text1);
       String[] t2=a.ziji(this.text2);
       String[] t1t,t2t;
       try{
           t1t=a.cut(t1, k);
           t2t=a.cut(t2, k);
           
       }catch(MyException e){
               return -1;
       }
       int he=a.heji(t1t, t2t);
       int jiao=a.jiaoji(t1t, t2t);
       return (float)jiao/he;
    }

}

面板设计部分：

package bigproject2;
import java.awt.*;
import java.awt.event.*;
import java.io.BufferedReader;
import java.io.File;
import java.io.FileNotFoundException;
import java.io.FileReader;
import java.io.IOException;
import java.io.InputStreamReader;

import javax.swing.*;
import javax.swing.event.*;
import javax.swing.filechooser.FileNameExtensionFilter;

public class Outlook extends JFrame{
    JFrame frm=new JFrame("相似度计算器");
    JPanel areabottom=new JPanel();
    JPanel areatop=new JPanel();
    JPanel areamiddle=new JPanel();
    static JTextArea tl=new JTextArea();
    static JTextArea tr=new JTextArea();
    JScrollPane left=new JScrollPane(tl,JScrollPane.VERTICAL_SCROLLBAR_ALWAYS,
            JScrollPane.HORIZONTAL_SCROLLBAR_AS_NEEDED);
    JScrollPane right=new JScrollPane(tr,JScrollPane.VERTICAL_SCROLLBAR_ALWAYS,
            JScrollPane.HORIZONTAL_SCROLLBAR_AS_NEEDED);
    JSplitPane sp=new JSplitPane(JSplitPane.HORIZONTAL_SPLIT,left,right);
    static JButton toBig=new JButton("全部大写");
    static JButton delbd=new JButton("去掉标点");
    static JButton count=new JButton("计算相似度");
    JLabel space=new JLabel("                                               ");
    JLabel t1=new JLabel("Text1");
    JLabel t2=new JLabel("Text2");

    JMenuBar mb=new JMenuBar();
    JMenu open=new JMenu("打开");
    JMenuItem opent1=new JMenuItem("打开到Text1");
    JMenuItem opent2=new JMenuItem("打开到Text2");
    
    private String str="";
    public Outlook()
    {
        judge();
        
        frm.setVisible(true);
        frm.setBounds(50, 50, 500, 400);
        frm.setLayout(new BorderLayout(5,5));
        
        frm.add("North",areatop);
        frm.add("Center",areamiddle);
        frm.add("South",areabottom);
        
        areatop.add(mb);
        mb.add(open);        
        open.add(opent1);
        open.add(opent2);
        open.setPreferredSize(new Dimension(40,18));
        mb.setBackground(frm.getBackground());
        areatop.setLayout(new FlowLayout(FlowLayout.LEFT));
        areamiddle.setLayout(new FlowLayout(FlowLayout.LEFT));
        
        areamiddle.add(t1);
        t1.setPreferredSize(new Dimension(frm.getWidth()/2-20,10));
        areamiddle.add(t2);
        t2.setPreferredSize(new Dimension(50,10));
        areamiddle.add(left);
        left.setPreferredSize(new Dimension(frm.getWidth()/2-20,frm.getHeight()/2));    
        areamiddle.add(right);
        right.setPreferredSize(new Dimension(frm.getWidth()/2-20,frm.getHeight()/2));
        tl.setLineWrap(true);
        tr.setLineWrap(true);
        
        areabottom.add(toBig);
        areabottom.add(delbd);
        areabottom.add(space);
        areabottom.add(count);
        
        opent1.addActionListener(new ActionListener(){
            public void actionPerformed(ActionEvent e) {
                try {
                    openfile();
                    tl.setText(str);
                } catch (IOException e1) {
                    e1.printStackTrace();
                }
                judge();
            }
        });
        opent2.addActionListener(new ActionListener(){
            public void actionPerformed(ActionEvent e) {
                try {
                    openfile();
                    tr.setText(str);
                } catch (IOException e1) {
                    e1.printStackTrace();
                }
                judge();
            }
        });
        toBig.addActionListener(new ActionListener(){
            public void actionPerformed(ActionEvent e){
                tl.setText(tobig(tl.getText()));
                tr.setText(tobig(tr.getText()));
            }
        });
        
        delbd.addActionListener(new ActionListener(){
            public void actionPerformed(ActionEvent e){
                tl.setText(del(tl.getText()));
                tr.setText(del(tr.getText()));
                judge();
            }
            
        });
        count.addActionListener(new ActionListener(){
            public void actionPerformed(ActionEvent e){
                KShinglingAlgorithm a=new KShinglingAlgorithm();
                a.setText1(tl.getText());
                a.setText2(tr.getText());
                float b=a.getSimilarity(4);
                if(b!=-1)
                    JOptionPane.showMessageDialog(null, Float.toString(b),"相似度", JOptionPane.INFORMATION_MESSAGE); 
            }
        });
        tr.addKeyListener(new KeyAdapter(){
            public void keyTyped(KeyEvent e){
                judge();
            }
        });
        tl.addKeyListener(new KeyAdapter(){
            public void keyTyped(KeyEvent e){
                judge();
            }
        });
    }
    public void judge(){
        if(tl.getText().length()!=0||tr.getText().length()!=0) {
            toBig.setEnabled(true);
            delbd.setEnabled(true);
            count.setEnabled(true);
        }
        else{
            toBig.setEnabled(false);
            delbd.setEnabled(false);
            count.setEnabled(false);
        }    
    }
    public void openfile() throws IOException{
        str="";
        JFileChooser choose=new JFileChooser();        
        int result = choose.showOpenDialog(this);
        File file = null; //注意初始化
        //加过滤器
        if (result == JFileChooser.APPROVE_OPTION) {
            file = choose.getSelectedFile();
            }
        else{
            return; //使点取消后不会抛出异常
        }
        FileReader fr=new FileReader(file);
        BufferedReader br=new BufferedReader(fr);
        char c[]=new char[512];
        String strline="";
        while(br.ready()){
            strline=br.readLine();
            str+=strline;
        };
        br.close();
        fr.close();
    }
    public String tobig(String str){
        String temp="";
        for(int i=0;i<str.length();i++)
        {
            if(str.charAt(i)>='a'&&str.charAt(i)<='z')
            {
                char t=str.charAt(i);
                t=(char)(str.charAt(i)-32);
                temp+=t;
            }
            else temp+=str.charAt(i);
        }
        return temp;
    }
    
    public String del(String str){
        String temp="";
        for(int i=0;i<str.length();i++)
        {
            char t=str.charAt(i);
            if(t>='!'&&t<='/'||t>=58&&t<=64||t>=91&&t<=96||t>=123&&t<=126);
            else temp+=t;
        }
        return temp;
    }
    public static void main(String[] args){
        new Outlook();
        
        
    }
}