def count_ngram_score(sentence, vector, n):
score = 0
ngrams_array = ngrams.make_ngrams(sentence, n)
for ngram in ngrams_array:
if ngram in vector[n - 1]:
score += vector[n - 1][ngram]
return score
def count_ngram_score(sentence, vector, n):
score = 0
ngrams_array = ngrams.make_ngrams(sentence, n)
for ngram in ngrams_array:
if ngram in vector[n - 1]:
score += vector[n - 1][ngram]
return score
def count_ngram_score(sentence, vector, n): score = 0 ngrams_array = ngrams.make_ngrams(sentence, n) #smoothing - for ngrams which don't appear in vector of language #set the worst (max) value in lang vector smoothing = max(vector[n-1].items(), key=operator.itemgetter(1))[0] for ngram in ngrams_array: if ngram in vector[n-1]: score += vector[n-1][ngram] else: score += vector[n-1][smoothing]/smoothing_rate return score
def count_ngram_score(sentence, vector, n):
score = 0
ngrams_array = ngrams.make_ngrams(sentence, n)
#smoothing - for ngrams which don't appear in vector of language
#set the worst (max) value in lang vector
smoothing = max(vector[n - 1].items(), key=operator.itemgetter(1))[0]
for ngram in ngrams_array:
if ngram in vector[n - 1]:
score += vector[n - 1][ngram]
else:
score += vector[n - 1][smoothing] / smoothing_rate
return score
