static void Main(string[] args)
{
List <Dictionary <string, double> > docWordDicList = new List <Dictionary <string, double> >();
Dictionary <string, int> dictionary = new Dictionary <string, int>();
List <int> trainingAnswer = new List <int>();
Dictionary <string, double> wordIDFDictionary = new Dictionary <string, double>();
Hashtable stopWordTable = genStopwordTable(@"D:\work\KPMG\learning\project1\stopword.txt");
List <string> testFileNameList = new List <string>();
int dicSize = 5000;
trainModel(@"D:\work\KPMG\learning\classification\project1_0422\test_data\1\Training",
@"D:\work\KPMG\learning\classification\project1_0422\log",
ref docWordDicList,
ref dictionary,
dicSize,
ref trainingAnswer,
ref wordIDFDictionary,
stopWordTable
);
KNN knn = new KNN();
knn.set(dicSize, docWordDicList.Count());
knn.initial(docWordDicList, dictionary, trainingAnswer);
knn.train(3, 20);
knn.getAveDistance();
//knn.genLog(@"D:\work\KPMG\learning\classification\project1_0422\log");
List <KeyValuePair <int, int> > testAnswer = runKnnTest(knn, @"D:\work\KPMG\learning\classification\project1_0422\test_data\1\Testing", @"D:\work\KPMG\learning\classification\project1_0422\test_data\log", dictionary, wordIDFDictionary, stopWordTable, ref testFileNameList);
genStatistic(testAnswer, testFileNameList, @"D:\work\KPMG\learning\classification\project1_0422\log");
}
static void Main(string[] args)
{
List <Dictionary <string, double> > docWordDicList = new List <Dictionary <string, double> >();
Dictionary <string, int> dictionary = new Dictionary <string, int>();
List <int> trainingAnswer = new List <int>();
Dictionary <string, double> wordIDFDictionary = new Dictionary <string, double>();
Hashtable stopWordTable = genStopwordTable(STOP_WORD_PATH);
List <string> testFileNameList = new List <string>();
int dicSize = 100;
Console.WriteLine("==> Starting prepare data...");
NLPAdapter nlpAdapter = new NLPAdapter(NLP_MODEL_PATH);
trainModel(TRAINING_DATA_DIR,
LOG_DIR,
ref docWordDicList,
ref dictionary,
dicSize,
ref trainingAnswer,
ref wordIDFDictionary,
stopWordTable,
nlpAdapter
);
#if KNN_MODE
KNN knn = new KNN();
knn.set(dicSize, docWordDicList.Count());
knn.initial(docWordDicList, dictionary, trainingAnswer);
knn.train(3, 20);
knn.getAveDistance();
//knn.genLog(@"D:\work\KPMG\learning\classification\project1_0422\log");
List <KeyValuePair <int, int> > testAnswer = runKnnTest(knn, TEST_DATA_DIR, TEST_LOG_DIR, dictionary, wordIDFDictionary, stopWordTable, ref testFileNameList, nlpAdapter);
#else
Console.WriteLine("==> Starting get model...");
SVMAdapter svmAdapter = new SVMAdapter();
svm_model model = svmAdapter.getSVMModel(docWordDicList, dictionary, trainingAnswer, SVMAdapter.SVM_C_DEFAULT, SVMAdapter.SVM_GAMMA_DEFAULT);
Console.WriteLine("==> Starting SVM test...");
List <KeyValuePair <int, int> > testAnswer = runSVMTest(svmAdapter, TEST_DATA_DIR, TEST_LOG_DIR, dictionary, wordIDFDictionary, stopWordTable, ref testFileNameList, model, nlpAdapter);
Console.WriteLine("==> Starting SVM test done!!");
#endif
Console.WriteLine("==> Starting saving result...");
genStatistic(testAnswer, testFileNameList, LOG_DIR);
}
private static List <KeyValuePair <int, int> > runKnnTest(KNN knn, string testPath, string logPath, Dictionary <string, int> dictionary, Dictionary <string, double> wordIDFDictionary, Hashtable stopWordTable, ref List <string> testFileNameList)
{
string[] categories = Directory.GetDirectories(testPath);
List <KeyValuePair <int, int> > testAnswer = new List <KeyValuePair <int, int> >();
for (int i = 0; i < categories.Length; i++) //traverse Categories
{
Console.WriteLine(categories[i]);
string[] files = Directory.GetFiles(categories[i]);
for (int j = 0; j < files.Length; j++)
{
int testResult = -1;
testFileNameList.Add(Path.GetFileName(files[j]));
testResult = knn.test(readDoc(files[j], stopWordTable), dictionary, wordIDFDictionary);
testAnswer.Add(new KeyValuePair <int, int>(testResult, i));
Console.WriteLine(testResult + "," + i);
}
}
return(testAnswer);
}