<html>
  <head>
    <meta content="text/html; charset=ISO-8859-1"
      http-equiv="Content-Type">
  </head>
  <body text="#000000" bgcolor="#FFFFFF">
    <div class="moz-cite-prefix">On 12/10/2013 10:20 PM, Laatar Rim
      wrote:<br>
    </div>
    <blockquote
      cite="mid:1386742826.82557.YahooMailNeo@web173205.mail.ir2.yahoo.com"
      type="cite">
      <div style="color:#000; background-color:#fff; font-family:times
        new roman, new york, times, serif;font-size:12pt">
        <div><span>Dear Andreas,</span></div>
        <div style="background-color: transparent;"><span>First i'm so
            sorry for disturbing you, </span></div>
        <div style="background-color: transparent;"><br>
        </div>
        <div style="background-color: transparent;"><span><span>from
              many day </span>I want to train a class-based model ,So I
            use (1), (2) and (3) to create the class model. </span></div>
        <div style="background-color: transparent;"><br>
        </div>
        <div style="background-color: transparent;"><span><span
              style="font-weight: bold;">(1)</span>  ngram-class -vocab
            '/home/hp/Documents/SRILM/tata.txt' -text
            '/home/hp/Documents/SRILM/trainingData.txt' -numclasses 37
            -class-counts output.class-counts -classes
            '/home/hp/Documents/SRILM/Replace_word_with_class_SRILM' <br>
            <br>
            <br>
          </span></div>
        <div style="background-color: transparent;"><span><span
              style="font-weight: bold;">(2)</span>  
            replace-words-with-classes 
            classes='/home/hp/Documents/SRILM/Replace_word_with_class_SRILM'
            '/home/hp/Documents/SRILM/trainingData.txt' >
            output_text_with_classes_2<br>
            <br>
          </span></div>
        <div style="background-color: transparent;"><span><span
              style="font-weight: bold;">(3)</span> ngram-count -tolower
            -text '/home/hp/Documents/SRILM/trainingData.txt' -lm
            class_based_model_2</span></div>
      </div>
    </blockquote>
    <br>
    For step 3 you need to use<br>
    <br>
        ngram-count -text output_text_with_classes_2 -lm <span>class_based_model_2<br>
      <br>
      To evaluate the LM you would then use <br>
      <br>
          ngram -lm </span><span>class_based_model_2 -classes
      '/home/hp/Documents/SRILM/Replace_word_with_class_SRILM' -ppl ...
      (or other options that use the LM)<br>
      <br>
    </span><br>
    Andreas <br>
    <br>
  </body>
</html>