def twss(sentence):
x = processSentence(str(sentence), vocabList)
old_stdout = sys.stdout
sys.stdout = StringIO()
p_label, p_acc, p_val = svm_predict([1], [x], model, '-b 1')
sys.stdout = old_stdout
return p_val[0][1]
def generateFeatures(filename, label, vocabList, X, y, Xtest, ytest):
input = open(filename)
sentences = pickle.load(input)
input.close()
# leave last 100 for test set
top = len(sentences) - 100
for i in range(top):
wi = processSentence(sentences[i], vocabList)
X = X + [wi]
y = y + [label]
bottom = top
top = len(sentences)
for i in range(bottom, top):
wi = processSentence(sentences[i], vocabList)
Xtest = Xtest + [wi]
ytest = ytest + [label]
return X, y, Xtest, ytest
def generateFeatures(filename,label,vocabList,X,y,Xtest,ytest):
input = open(filename);
sentences = pickle.load(input)
input.close()
# leave last 100 for test set
top = len(sentences)-100
for i in range(top):
wi = processSentence(sentences[i],vocabList)
X = X+[wi]
y = y+[label]
bottom = top
top = len(sentences)
for i in range(bottom,top):
wi = processSentence(sentences[i],vocabList)
Xtest = Xtest + [wi]
ytest = ytest + [label]
return X,y,Xtest,ytest
def twss(sentence,vocabList,model):
#print "you said: '"+sentence+"'\n"
# these should be moved to file
responses = ['Whatever ...','Okay','Yawn','What makes you think I care?','Yada yada','Uhuh','Yeah, yeah','figures',"I'm hungry",'give me a break','so ...']
x = processSentence(sentence, vocabList)
#print [x]
p_label, p_acc, p_val = svm_predict([1], [x], model, '-b 1 -q')
print p_label, p_acc, p_val
if p_label[0] == 1:
return "That's what she said!\n"
else:
return random.choice(responses) +'\n'
def twss(sentence, vocabList, model):
#print "you said: '"+sentence+"'\n"
# these should be moved to file
responses = [
'Whatever ...', 'Okay', 'Yawn', 'What makes you think I care?',
'Yada yada', 'Uhuh', 'Yeah, yeah', 'figures', "I'm hungry",
'give me a break', 'so ...'
]
x = processSentence(sentence, vocabList)
#print [x]
p_label, p_acc, p_val = svm_predict([1], [x], model, '-b 1 -q')
print p_label, p_acc, p_val
if p_label[0] == 1:
return "That's what she said!\n"
else:
return random.choice(responses) + '\n'
def twss_lite(sentence,vocabList,model):
x = processSentence(sentence, vocabList)
p_label, p_acc, p_val = svm_predict([1], [x], model, '-b 1 -q')
return p_label[0]
def twss_lite(sentence, vocabList, model):
x = processSentence(sentence, vocabList)
p_label, p_acc, p_val = svm_predict([1], [x], model, '-b 1 -q')
return p_label[0]
from processSentence import *
sent_brok = processSentence("The magician got so mad he pulled his hare out")
#print sent_brok
"""
def sentFeat(sent):
# Preprocessing the sentence
sent_brok = processSentence(sent)
# Looking whether sentence have homonym or not
for word in sent_brok:
for homWord in homonym:
if homWord[1] == word:
# search(ngram, word)
# how to get frequency
# searching ngram frequency for a word
def search(ngram_Corpus, searchFor):
for key in ngram_Corpus:
for value in ngram_Corpus[k]:
if searchFor in value:
return key
return None
"""