def count_noun(s1, s2):
try:
s1 = utils.valid_words(w2v_google, convert(s1))
s2 = utils.valid_words(w2v_google, convert(s2))
return len(s1) == len(s2) * 1
except:
return -1
def diff_w2v_set_diff(s1, s2):
try:
if 'differ' not in s1 and 'differ' not in s2:
return -1
s1 = set(utils.valid_words(w2v_google, convert(s1)))
s2 = set(utils.valid_words(w2v_google, convert(s2)))
return w2v_google.n_similarity(s1 - s2, s2 - s1)
except:
return -1
def diff_w2v_quora(s1, s2):
try:
if 'differ' not in s1 and 'differ' not in s2:
return -1
s1 = utils.valid_words(w2v_quora, convert(s1))
s2 = utils.valid_words(w2v_quora, convert(s2))
return w2v_quora.n_similarity(s1, s2)
except:
return -1
def len_valid(s1, s2):
try:
if 'differ' not in s1 and 'differ' not in s2:
return -1
s1 = set(utils.valid_words(w2v_google, convert(s1)))
s2 = set(utils.valid_words(w2v_google, convert(s2)))
return (len(s1) == len(s2)) * 1
except:
return -1
def most_sim(s1, s2):
"""
"""
s1 = utils.valid_words(w2v_google, s1 - stops)
s2 = utils.valid_words(w2v_google, s2 - stops)
words_sets = product(s1, s2)
sims = [
sorted([w1, w2]) for w1, w2 in words_sets
if w2v_google.similarity(w1, w2) > sim_value
]
# sims = [(w1,w2) for w1,w2 in words_sets if w2v_google.similarity(w1,w2)>sim_value or w2v_quora.similarity(w1,w2)>sim_value]
return sims