def build_Classifier_Train_Test(self, feature_extractor):
self.q21 = q2_1()
classifier = self.q21.evaluate_features(feature_extractor, 2)
train = self.q21.maintrain
test = self.q21.maintest
testClaassify = self.q21.testClassify
return classifier, train, test, testClaassify
def build_Classifier_Train_Test(self, feature_extractor):
self.q21 = q2_1()
classifier = self.q21.evaluate_features(feature_extractor, 2)
train = self.q21.maintrain
test = self.q21.maintest
testClaassify = self.q21.testClassify
return classifier,train,test, testClaassify
def main():
q21 = q2_1()
x = []
y = []
pos = ['N', 'VG', 'ADJ', 'ADV']
print pos
extractor = make_pos_extractor(pos)
classifier = q21.evaluate_features(extractor, 10)
x.append(1)
acc = accuracy(q21.maintest, q21.testClassify)
y.append(acc)
pos = ['N', 'V', 'VG', 'VN', 'VN', 'ADJ', 'ADV']
print pos
extractor = make_pos_extractor(pos)
classifier = q21.evaluate_features(extractor, 10)
x.append(2)
acc = accuracy(q21.maintest, q21.testClassify)
y.append(acc)
pos = ['V', 'ADJ', 'ADV']
print pos
extractor = make_pos_extractor(pos)
classifier = q21.evaluate_features(extractor, 10)
x.append(3)
acc = accuracy(q21.maintest, q21.testClassify)
y.append(acc)
pos = ['ADJ', 'ADV']
print pos
extractor = make_pos_extractor(pos)
classifier = q21.evaluate_features(extractor, 10)
x.append(4)
acc = accuracy(q21.maintest, q21.testClassify)
y.append(acc)
pos = ['N', 'ADJ', 'ADV']
print pos
extractor = make_pos_extractor(pos)
classifier = q21.evaluate_features(extractor, 10)
x.append(5)
acc = accuracy(q21.maintest, q21.testClassify)
y.append(acc)
pylab.bar(x, y, width=0.02, facecolor='blue', align='center')
pylab.xlabel('POS')
pylab.ylabel("Accuracy")
pylab.title("Accuracy for each pos set")
pylab.grid(False)
pylab.show()
return
def main():
q21 = q2_1()
x = []
y = []
pos = ['N', 'VG', 'ADJ', 'ADV']
print pos
extractor = make_pos_extractor(pos)
classifier = q21.evaluate_features(extractor, 10)
x.append(1)
acc = accuracy(q21.maintest, q21.testClassify)
y.append(acc)
pos = ['N', 'V', 'VG', 'VN', 'VN', 'ADJ', 'ADV']
print pos
extractor = make_pos_extractor(pos)
classifier = q21.evaluate_features(extractor, 10)
x.append(2)
acc = accuracy(q21.maintest, q21.testClassify)
y.append(acc)
pos = ['V', 'ADJ', 'ADV']
print pos
extractor = make_pos_extractor(pos)
classifier = q21.evaluate_features(extractor, 10)
x.append(3)
acc = accuracy(q21.maintest, q21.testClassify)
y.append(acc)
pos = ['ADJ', 'ADV']
print pos
extractor = make_pos_extractor(pos)
classifier = q21.evaluate_features(extractor, 10)
x.append(4)
acc = accuracy(q21.maintest, q21.testClassify)
y.append(acc)
pos = ['N', 'ADJ', 'ADV']
print pos
extractor = make_pos_extractor(pos)
classifier = q21.evaluate_features(extractor, 10)
x.append(5)
acc = accuracy(q21.maintest, q21.testClassify)
y.append(acc)
pylab.bar(x, y, width=0.02, facecolor='blue', align='center')
pylab.xlabel('POS')
pylab.ylabel("Accuracy")
pylab.title("Accuracy for each pos set")
pylab.grid(False)
pylab.show()
return
def main():
q21 = q2_1()
print "bag of words extractor:"
firstClassifier = q21.evaluate_features(bag_of_words, 10)
oldClassify = q21.testClassify
print "top k frequent words without stop words extractor:"
extractor = make_topK_non_stopword_extractor(10000, stopset)
secondClassifier = q21.evaluate_features(extractor, 10)
newClassify = q21.testClassify
#identifying documents that classified differently and report new positive, new negative
noPos, noNeg = newTaggs(oldClassify, newClassify)
print "No of documents that the classifier classify them as pos in bag of words extractor and neg in top k extractor is:", noNeg
print "No of documents that the classifier classify them as neg in bag of words extractor and pos in top k extractor is:", noPos
#drawing plot of accuract vs. K/W
plotGraph(q21, 5000)
return
def main():
q21 = q2_1()
print "bag of words extractor:"
firstClassifier = q21.evaluate_features(bag_of_words, 10)
oldClassify = q21.testClassify
print "top k frequent words without stop words extractor:"
extractor = make_topK_non_stopword_extractor(10000, stopset)
secondClassifier = q21.evaluate_features(extractor, 10)
newClassify = q21.testClassify
#identifying documents that classified differently and report new positive, new negative
noPos, noNeg = newTaggs(oldClassify, newClassify)
print "No of documents that the classifier classify them as pos in bag of words extractor and neg in top k extractor is:", noNeg
print "No of documents that the classifier classify them as neg in bag of words extractor and pos in top k extractor is:", noPos
#drawing plot of accuract vs. K/W
plotGraph(q21, 5000)
return
def main():
q21 = q2_1()
print "bigram Extractor:"
#evaluate bigram extractor
extractor = make_bigram_extractor()
classifier = q21.evaluate_features(extractor, 4)
print "bag of words extractor:"
#evaluate bag of words extractor
extractor = bag_of_words
classifier = q21.evaluate_features(extractor, 4)
print "all bigram unigram extractor"
#evaluate all bigrams and all unigrams extractor
extractor = make_bigram_unigram_extractor()
classifier = q21.evaluate_features(extractor, 4)
print "good bigram unigram extractor:"
#evaluate good bigrams and all unigrams extractor
extractor = make_good_bigram_unigram_extractor(100)
classifier = q21.evaluate_features(extractor, 4)
return
def main():
q21 = q2_1()
print "bigram Extractor:"
#evaluate bigram extractor
extractor = make_bigram_extractor()
classifier = q21.evaluate_features(extractor, 4)
print "bag of words extractor:"
#evaluate bag of words extractor
extractor = bag_of_words
classifier = q21.evaluate_features(extractor, 4)
print "all bigram unigram extractor"
#evaluate all bigrams and all unigrams extractor
extractor = make_bigram_unigram_extractor()
classifier = q21.evaluate_features(extractor, 4)
print "good bigram unigram extractor:"
#evaluate good bigrams and all unigrams extractor
extractor = make_good_bigram_unigram_extractor(100)
classifier = q21.evaluate_features(extractor, 4)
return