Package com.aliasi.classify

Examples of com.aliasi.classify.ConfusionMatrix

44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
      ObjectInputStream oi = new ObjectInputStream( new FileInputStream(modelFile) );
      LMClassifier compiledClassifier = (LMClassifier) oi.readObject();
      oi.close();
     
      //*-- loop through the identical categories and test the classification of test documents
    ConfusionMatrix confMatrix = new ConfusionMatrix(CATEGORIES);
      NumberFormat nf = NumberFormat.getInstance();
      nf.setMaximumIntegerDigits(1); nf.setMaximumFractionDigits(3);
    for (int i=0; i < CATEGORIES.length; ++i)
       {
    File classDir = new File(TESTING_DIR, CATEGORIES[i]);
    String[] testingFiles = classDir.list();
       
        //*-- for each file, find the best category using the classifier and compare with the
        //*-- designated category
    for (int j = 0; j < testingFiles.length; ++j)
         {
      String text = Files.readFromFile( new File(classDir, testingFiles[j]) );
      logger.debug("Testing on " + CATEGORIES[i] + File.separator + testingFiles[j]);
      JointClassification jc =  compiledClassifier.classifyJoint(text);
          confMatrix.increment(CATEGORIES[i], jc.bestCategory());
          logger.debug("Best Category: " + jc.bestCategory() );
          StringBuffer sb = new StringBuffer();
          sb.append("Scores ");
          for (int k = 0; k < CATEGORIES.length; k++) sb.append(nf.format(jc.score(k)) + " ");
          logger.debug(sb);
     } //*-- end of inner for
    } //*-- end of outer for
     
      logger.info("--------------------------------------------");
      logger.info("- Results ");
      logger.info("--------------------------------------------");
      int[][] imatrix = confMatrix.matrix();
      StringBuffer sb = new StringBuffer();
      sb.append(StringTools.fillin("CATEGORY", 10, true, ' ') );
      for (int i = 0; i < CATEGORIES.length; i++) sb.append(StringTools.fillin(CATEGORIES[i], 12, true, ' ') );
      logger.info(sb.toString());
     
      for (int i = 0; i < imatrix.length; i++)
      { sb = new StringBuffer();
        sb.append(StringTools.fillin(CATEGORIES[i], 10, true, ' ', 10 - CATEGORIES[i].length() ) );
        for (int j = 0; j < imatrix.length; j++)
         {  String out = "" + imatrix[i][j];
          sb.append(StringTools.fillin(out, 10, false, ' ', 10 - out.length() ) );
         }
        logger.info(sb.toString());
      }
     
    logger.info("Total Accuracy: " + nf.format(confMatrix.totalAccuracy()) );
      logger.info("Total Correct : " + confMatrix.totalCorrect() + " out of " + confMatrix.totalCount() );
    }
View Full Code Here
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
      ObjectInputStream oi = new ObjectInputStream( new FileInputStream(modelFile) );
      LMClassifier compiledClassifier = (LMClassifier) oi.readObject();
      oi.close();
     
      //*-- loop through the identical categories and test the classification of test documents
    ConfusionMatrix confMatrix = new ConfusionMatrix(CATEGORIES);
      NumberFormat nf = NumberFormat.getInstance();
      nf.setMaximumIntegerDigits(1); nf.setMaximumFractionDigits(3);
    for (int i=0; i < CATEGORIES.length; ++i)
       {
    File classDir = new File(TESTING_DIR, CATEGORIES[i]);
    String[] testingFiles = classDir.list();
       
        //*-- for each file, find the best category using the classifier and compare with the
        //*-- designated category
    for (int j=0; j < testingFiles.length; ++j)
         {
      String text = Files.readFromFile( new File(classDir, testingFiles[j]) );
      logger.debug("Testing on " + CATEGORIES[i] + File.separator + testingFiles[j]);
      JointClassification jc =  compiledClassifier.classifyJoint(text);
          confMatrix.increment(CATEGORIES[i], jc.bestCategory());
          logger.debug("Best Category: " + jc.bestCategory() );
          StringBuffer sb = new StringBuffer();
          sb.append("Scores ");
          for (int k = 0; k < CATEGORIES.length; k++) sb.append(nf.format(jc.score(k)) + " ");
          logger.debug(sb);
     } //*-- end of inner for
    } //*-- end of outer for
     
      logger.info("--------------------------------------------");
      logger.info("- Results ");
      logger.info("--------------------------------------------");
      int[][] imatrix = confMatrix.matrix();
      StringBuffer sb = new StringBuffer();
      sb.append(StringTools.fillin("CATEGORY", 10, true, ' ') );
      for (int i = 0; i < CATEGORIES.length; i++) sb.append(StringTools.fillin(CATEGORIES[i], 8, false, ' ') );
      logger.info(sb.toString());
     
      for (int i = 0; i < imatrix.length; i++)
      { sb = new StringBuffer();
        sb.append(StringTools.fillin(CATEGORIES[i], 10, true, ' ', 10 - CATEGORIES[i].length() ) );
        for (int j = 0; j < imatrix.length; j++)
         {  String out = "" + imatrix[i][j];
          sb.append(StringTools.fillin(out, 8, false, ' ', 8 - out.length() ) );
         }
        logger.info(sb.toString());
      }
     
    logger.info("Total Accuracy: " + nf.format(confMatrix.totalAccuracy()) );
      logger.info("Total Correct : " + confMatrix.totalCorrect() + " out of " + confMatrix.totalCount() );
    }
View Full Code Here
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
      ObjectInputStream oi = new ObjectInputStream( new FileInputStream(modelFile) );
      LMClassifier compiledClassifier = (LMClassifier) oi.readObject();
      oi.close();
     
      //*-- loop through the identical categories and test the classification of test documents
    ConfusionMatrix confMatrix = new ConfusionMatrix(CATEGORIES);
      NumberFormat nf = NumberFormat.getInstance();
      nf.setMaximumIntegerDigits(1); nf.setMaximumFractionDigits(3);
    for (int i=0; i < CATEGORIES.length; ++i)
       {
    File classDir = new File(TESTING_DIR, CATEGORIES[i]);
    String[] testingFiles = classDir.list();
       
        //*-- for each file, find the best category using the classifier and compare with the
        //*-- designated category
    for (int j=0; j < testingFiles.length; ++j)
         {
      String text = Files.readFromFile( new File(classDir, testingFiles[j]) );
     
      //*-- limit the length of the text
        if (text.length() > 500)  text = text.substring(0, 500);
      logger.debug("Testing on " + CATEGORIES[i] + File.separator + testingFiles[j]);
      JointClassification jc =  compiledClassifier.classifyJoint(text);
     
      //*-- check if we have sufficient confidence in the decision
      String bestCategory = (jc.score(0) > -2.5) ? jc.bestCategory(): "text";
      confMatrix.increment(CATEGORIES[i], bestCategory); 
          logger.debug("Best Category: " + bestCategory );
          StringBuffer sb = new StringBuffer();
          sb.append("Scores ");
          for (int k = 0; k < CATEGORIES.length; k++)
            sb.append(nf.format(jc.score(k)) + " ");
          logger.debug(sb);
     } //*-- end of inner for
    } //*-- end of outer for
     
      logger.info("--------------------------------------------");
      logger.info("- Results ");
      logger.info("--------------------------------------------");
      int[][] imatrix = confMatrix.matrix();
      StringBuffer sb = new StringBuffer();
      sb.append(StringTools.fillin("CATEGORY", 10, true, ' ') );
      for (int i = 0; i < CATEGORIES.length; i++) sb.append(StringTools.fillin(CATEGORIES[i], 8, false, ' ') );
      logger.info(sb.toString());
     
      for (int i = 0; i < imatrix.length; i++)
      { sb = new StringBuffer();
        sb.append(StringTools.fillin(CATEGORIES[i], 10, true, ' ', 10 - CATEGORIES[i].length() ) );
        for (int j = 0; j < imatrix.length; j++)
         {  String out = "" + imatrix[i][j];
          sb.append(StringTools.fillin(out, 8, false, ' ', 8 - out.length() ) );
         }
        logger.info(sb.toString());
      }
     
    logger.info("Total Accuracy: " + nf.format(confMatrix.totalAccuracy()) );
      logger.info("Total Correct : " + confMatrix.totalCorrect() + " out of " + confMatrix.totalCount() );
    }
View Full Code Here
TOP

Related Classes of com.aliasi.classify.ConfusionMatrix

  • org.sf.mustru.test.TestFtypeClassifier
  • org.sf.mustru.test.TestMClassifier
  • org.sf.mustru.test.TestTtypeClassifier
    • Copyright © 2018 www.massapicom. All rights reserved.
    All source code are property of their respective owners. Java is a trademark of Sun Microsystems, Inc and owned by ORACLE Inc. Contact coftware#gmail.com.