def count_pronouns_per_sentence(tagged_corpus):
num_of_pronouns = 0
num_of_sents = 0
for sent in tagged_corpus:
num_of_sents += 1.0
for (word, tag) in sent:
if (brown_pron_tags.__contains__(tag)) or (penn_pron_tags.__contains__(tag)):
num_of_pronouns += 1.0
if not num_of_pronouns == 0:
pron_per_sent = num_of_pronouns/num_of_sents
else:
pron_per_sent = 0
return pron_per_sent
def count_pronouns_per_words(tagged_corpus):
num_of_pronouns = 0
num_of_words = 0
for sent in tagged_corpus:
for (word, tag) in sent:
if not (penn_punct.__contains__(tag) or brown_punct.__contains__(tag)):
num_of_words += 1.0
if (brown_pron_tags.__contains__(tag)) or (penn_pron_tags.__contains__(tag)):
num_of_pronouns += 1.0
if not num_of_pronouns == 0:
pron_per_word_total = num_of_words/num_of_pronouns
else:
pron_per_word_total = 0.0
return pron_per_word_total
def count_pronouns_sent(tagged_sent):
num_of_pronouns = 0
for (word, tag) in tagged_sent:
if (brown_pron_tags.__contains__(tag)) or (penn_pron_tags.__contains__(tag)):
num_of_pronouns += 1
return num_of_pronouns
