def cosine_sim(str1, str2):
str_ = [
cleanText(str1, True, False, True, False),
cleanText(str2, True, False, True, False)
]
vectorizer = CountVectorizer()
vect_ = vectorizer.fit_transform(str_).toarray()
return cosine_similarity([vect_[0]], [vect_[1]])[0][0] * 100
def spam_clf(text):
text = [cleanText(str(text), True, False, False, False)]
vector = spam_vectorizer.transform(text)
return 'Spam' if spam_classifier.predict(vector)[0] == 1 else 'Ham'
def jaccard_sim(str1, str2):
a = set(cleanText(str1, True, False, True, False).split())
b = set(cleanText(str2, True, False, True, False).split())
c = a & b
d = a | b
return float(len(c)) / (len(d)) * 100
def fuzz_token_set_ratio(str1, str2):
return fuzz.token_sort_ratio(cleanText(str1, True, False, True, False),
cleanText(str2, True, False, True, False))
def fuzz_partial_ratio(str1, str2):
return fuzz.partial_ratio(cleanText(str1, True, False, True, False),
cleanText(str2, True, False, True, False))