多线程程序花费的时间比单线程（Java）长

我必须编写一个程序来读取文件中的所有单词，并确定每个单词使用了多少次。我的任务是使用多线程来加快运行时间，但是单线程程序的运行速度比多线程程序快。我曾尝试研究此问题的解决方案，但很多解释只会让我更加困惑。我对使用线程很陌生，想知道是否有人可以指出正确的方向修复我的代码，以便创建线程的开销不会导致程序运行速度比单个线程慢。

public class Main 
{
final static int THREADS = 4;
static HashMap<String, Integer> map = new HashMap<>();
static List<String> file = new ArrayList<String>();
static String filename = "D:\\yes.txt";
static int count;




public static void main(String args[]) throws Exception
{
    long startTime = System.nanoTime();
    Monitor m = new Monitor();
    final Queue<String> dataQueue = new ConcurrentLinkedQueue<>();

    try ( Scanner in = new Scanner(new File(filename))) 
    {
                while ( in.hasNext() ) 
                {
                    dataQueue.add( in.next() );
                }
    }
    catch ( IOException e ) 
    {
        e.printStackTrace();
    }


    Thread T1 = new Thread( new WordCount(m, map, dataQueue ));
    Thread T2 = new Thread( new WordCount(m, map, dataQueue ));
    Thread T3 = new Thread( new WordCount(m, map, dataQueue ));
    Thread T4 = new Thread( new WordCount(m, map, dataQueue ));

    T1.start();
    T2.start();
    T3.start();
    T4.start();


     //wait for threads to end
       try {
       T1.join();
       T2.join();
       T3.join();
       T4.join();
    } catch ( Exception e) {
       System.out.println("Interrupted");
    }   

    Set<String> keys = map.keySet();
    for (String key : keys) 
    {
        System.out.println(key);
        System.out.println(map.get(key));
    }
    long endTime = System.nanoTime();
    System.out.println("Thread Took "+((endTime - startTime)/100000) + " ms");


}
}

public class WordCount implements Runnable
{

    private Monitor m;
    private Queue<String> dataQueue;
    private HashMap<String, Integer> map;

    public WordCount(Monitor m, HashMap<String, Integer> map,Queue<String> dataQueue)
    {
        this.m = m;
        this.dataQueue = dataQueue;
        this.map = map;
    }

    @Override public void run()
    {
        while ( !dataQueue.isEmpty() ) 
        {
            String line = dataQueue.poll();
            m.keySet(map, line);
        }
    }   
}

public class Monitor 
{
    public synchronized void keySet(HashMap<String, Integer> map, String word) 
    {
        String[] words = filterIllegalTokens(word );
        String[] lowerCaseWords = mapToLowerCase( words );
         for ( String s : lowerCaseWords ) {


        if (map.containsKey(s)) 
        {
            map.put(s, map.get(s) + 1);

        } 
        else 
        {
            map.put(s, 1);
        }
         }
    }
    public  String[] filterIllegalTokens(String words)
    {
        List<String> filteredList = new ArrayList<>();

        if ( words.matches( "[a-zA-Z]+" ) ) {
                filteredList.add( words );
            }

        return filteredList.toArray( new String[filteredList.size()] );
    }
    public  String[] mapToLowerCase( String[] words )
    {
        String[] filteredList = new String[words.length];
        for ( int i = 0; i < words.length; i++ ) {
            filteredList[i] = words[i].toLowerCase();
        }
        return filteredList;
    }
}

这是我的三个班级。有任何提示或建议吗？