def jac():
l = []
df = word_shear()
q1 = dt.getQues1(df)
q2 = dt.getQues2(df)
lemmatizer = WordNetLemmatizer()
for i in range(len(q1)):
word_list = nltk.word_tokenize(q1[i])
word_list2 = nltk.word_tokenize(q2[i])
lemmatized_output = [lemmatizer.lemmatize(w) for w in word_list]
lemmatized_output2 = [lemmatizer.lemmatize(w) for w in word_list2]
count = 0
if len(lemmatized_output) >= len(lemmatized_output2):
for j in range(len(lemmatized_output2)):
if lemmatized_output2[j] == lemmatized_output[j]:
count += 1
else:
for j in range(len(lemmatized_output)):
if lemmatized_output[j] == lemmatized_output2[j]:
count += 1
z = round(
float(count) /
(len(lemmatized_output) + len(lemmatized_output2) - count), 2)
l.append(z)
df.insert(14, 'jakard', " ")
x = dt.setjakard(df, l)
return df
def compareDiffWords(): l = [] tokenizer = RegexpTokenizer(r'\w+') df = compareSameWords() q1=dt.getQues1(df) q2=dt.getQues2(df) for i in range(len(q1)): x = tokenizer.tokenize(q1[i]) y = tokenizer.tokenize(q2[i]) count = 0 if len(x) >= len(y): for j in range(len(y)): if y[j] != x[j]: count += 1 count += (len(x) - len(y)) l.append(count) else: for j in range(len(x)): if x[j] != y[j]: count +=1 count += (len(y) - len(x)) l.append(count) df.insert(10, 'diffwords', " ") dt.setcdw(df,l) return df
def punc():
l = []
m = []
df= lower()
x = dt.getQues1(df)
y = dt.getQues2(df)
tokenizer = RegexpTokenizer(r'\w+')
for i in range(len(x)):
x[i] = tokenizer.tokenize(x[i])
y[i] = tokenizer.tokenize(y[i])
if x[i] == l or y[i] == l:
x[i] = " ".join(x[i])
y[i] = " ".join(y[i])
m.append(i)
else:
x[i] = " ".join(x[i])
y[i] = " ".join(y[i])
x = list(x)
for index in sorted(m, reverse=True):
del x[index]
df = df.drop(df.index[m])
dt.setQues1(df,x)
dt.setQues2(df,y)
print('punc')
return df
def lower():
df = dt.readfile()
x = dt.getQues1(df)
y = dt.getQues2(df)
for i in range(len(x)):
x[i] = x[i].lower()
y[i] = y[i].lower()
dt.setQues1(df,x)
dt.setQues2(df,y)
print('lower')
return df
def Fuz_sort():
l = []
df = Fuz()
q1 = dt.getQues1(df)
q2 = dt.getQues2(df)
for i in range(len(q1)):
a = fuzz.token_sort_ratio(q1[i], q2[i])
a = a / 100
l.append(a)
df.insert(12, 'sortedfuzzy', " ")
x = dt.setfuzzsort(df, l)
return df
def Fuz():
df = bf.compareDiffWords()
q1 = dt.getQues1(df)
q2 = dt.getQues2(df)
l = []
for i in range(len(q1)):
a = fuzz.ratio(q1[i], q2[i])
a = a / 100
l.append(a)
df.insert(11, 'fuzzywuzzy', " ")
x = dt.setfuzz(df, l)
return df
def leng(): l = [] m = [] df = pre.punc() x = dt.getQues1(df) y = dt.getQues2(df) for i in range(len(x)): l.append(len(x[i])) m.append(len(y[i])) df.insert(5, 'len q1', " ") df.insert(6, 'len q2', " ") dt.setlen1(df,l) dt.setlen2(df,m) return df
def levention():
l = []
df = jac()
q1 = dt.getQues1(df)
q2 = dt.getQues2(df)
levenshtein = Levenshtein()
for i in range(len(q1)):
w1 = q1[i]
w2 = q2[i]
z = levenshtein.distance(w1, w2)
l.append(z)
df.insert(15, 'levenshtein', " ")
x = dt.setleven(df, l)
return df
def word_shear():
l = []
df = Fuz_sort()
q1 = dt.getQues1(df)
q2 = dt.getQues2(df)
tokenizer = RegexpTokenizer(r'\w+')
for i in range(len(q1)):
w1 = tokenizer.tokenize(q1[i])
w2 = tokenizer.tokenize(q2[i])
f = ((1.0) * len(w1 and w2)) / (len(w1) + len(w2))
l.append(round(f, 2))
df.insert(13, 'word_shear', " ")
x = dt.setwordshear(df, l)
return df
def compare(): l =[] tokenizer = RegexpTokenizer(r'\w+') df = leng_diff() q1=dt.getQues1(df) q2=dt.getQues2(df) for i in range(len(q1)): x = tokenizer.tokenize(q1[i]) y = tokenizer.tokenize(q2[i]) if x[0] == y[0]: l.append(1) else: l.append(0) df.insert(8, 'fcw', " ") dt.setfcw(df,l) return df
def tf_idf():
l = []
df = levention()
q1 = dt.getQues1(df)
q2 = dt.getQues2(df)
for i in range(len(q1)):
l.append(q1[i])
l.append(q2[i])
vectorizer = TfidfVectorizer()
vectorizer.fit(l)
m = []
n = []
for i in range(len(q1)):
m.append(vectorizer.transform([q1[i]]))
n.append(vectorizer.transform([q2[i]]))
p = cs(m, n)
q = eq(m, n)
r = pwm(m, n)
s = pwl(m, n)
t = mink(m, n)
u = canb(m, n)
df.insert(16, 'cs', " ")
df.insert(17, 'eq', " ")
df.insert(18, 'pwm', " ")
df.insert(19, 'pwl', " ")
df.insert(20, 'mink', " ")
df.insert(21, 'canb', " ")
x = dt.setcs(df, p)
y = dt.seteq(df, q)
z = dt.setpwm(df, r)
a = dt.setpwl(df, s)
b = dt.setmink(df, t)
c = dt.setcanb(df, u)
df[[
'id', 'qid1', 'qid2', 'question1', 'question2', 'len q1', 'len q2',
'len_diff', 'fcw', 'samewords', 'diffwords', 'fuzzywuzzy',
'sortedfuzzy', 'word_shear', 'jakard', 'levenshtein', 'cs', 'eq',
'pwm', 'pwl', 'mink', 'canb'
]]
return df
