def commit_pattern_cc_polarity():
import loacore.analysis.sentiment_analysis as sentiment_analysis
import loacore.load.file_load as file_load
import loacore.load.review_load as review_load
ids = file_load.get_id_files_by_file_paths([r'.*/uci/.+'])
reviews = review_load.load_reviews_by_id_files(id_files=ids,
load_sentences=True, load_words=True, load_deptrees=True)
sentiment_analysis.compute_pattern_reviews_polarity(reviews, commit_polarities=True, adj_pattern=False)
reviews = review_load.load_reviews_by_id_files(id_files=ids,
load_polarities=True, load_sentences=True, load_words=True)
for review in reviews:
print(review.review, " : ", review.polarities["pattern_cc"].pos_score, ", ",
review.polarities["pattern_cc"].neg_score, ", ",
review.polarities["pattern_cc"].obj_score)
def commit_pessimistic_polarity():
import loacore.analysis.sentiment_analysis as sentiment_analysis
import loacore.load.file_load as file_load
import loacore.load.review_load as review_load
ids = file_load.get_id_files_by_file_paths([r'.*/uci/.+'])
reviews = review_load.load_reviews_by_id_files(id_files=ids, load_sentences=True, load_words=True)
sentiment_analysis.compute_extreme_reviews_polarity(reviews,
commit_polarities=True, pessimistic=True, freeling_lang='en')
reviews = review_load.load_reviews_by_id_files(id_files=ids,
load_polarities=True, load_sentences=True, load_words=True)
for review in reviews:
print(review.review, " : ", review.polarities["pessimistic"].pos_score, ", ",
review.polarities["pessimistic"].neg_score, ", ",
review.polarities["pessimistic"].obj_score)
def test_lsi_leave_p_out():
import loacore.learning.lsi as lsi
import loacore.load.file_load as file_load
import loacore.load.review_load as review_load
import loacore.learning.validation as val
ids = file_load.get_id_files_by_file_path(r'.*/uci/yelp_labelled.txt')
reviews = review_load.load_reviews_by_id_files(id_files=ids,
load_polarities=True,
load_sentences=True,
load_words=True)
word2int_dict, model = lsi.lsi_model(reviews)
reviews_vectors = lsi.reviews_2_vec(reviews, model, word2int_dict)
print(val.leave_p_out_validation(reviews, reviews_vectors, 1))
def test_leave_p_out():
import loacore.load.file_load as file_load
import loacore.load.review_load as review_load
import loacore.learning.word2vec as w2v
import loacore.learning.validation as val
ids = file_load.get_id_files_by_file_path(r'.*/uci/yelp_labelled.txt')
reviews = review_load.load_reviews_by_id_files(id_files=ids,
load_polarities=True,
load_sentences=True,
load_words=True)
wv = w2v.word_2_vec(reviews)
reviews_vectors = w2v.reviews_2_vec(reviews, wv)
print(val.leave_p_out_validation(reviews, reviews_vectors, 1))
def test_pattern_polarity():
import loacore.load.file_load as file_load
import loacore.load.review_load as review_load
import loacore.analysis.sentiment_analysis as sentiment_analysis
ids = file_load.get_id_files_by_file_paths([r'.*/uci/yelp_labelled.txt'])
reviews = review_load.load_reviews_by_id_files(id_files=ids,
load_sentences=True, load_words=True, load_deptrees=True)
sentiment_analysis.compute_pattern_reviews_polarity(reviews)
for review in reviews:
print(review.review, " : ", review.polarities["pattern_adj_cc"].pos_score, ", ",
review.polarities["pattern_adj_cc"].neg_score, ", ",
review.polarities["pattern_adj_cc"].obj_score)
def test_lsi_k_fold():
import loacore.learning.lsi as lsi
import loacore.load.file_load as file_load
import loacore.load.review_load as review_load
import loacore.learning.validation as val
ids = file_load.get_id_files_by_file_path(r'.*/uci/yelp_labelled.txt')
reviews = review_load.load_reviews_by_id_files(id_files=ids,
load_polarities=True,
load_sentences=True,
load_words=True)[:100]
word2int_dict, model = lsi.lsi_model(reviews)
reviews_vectors = lsi.reviews_2_vec(reviews, model, word2int_dict)
# print(type(reviews_vectors))
# print(len(reviews_vectors))
# print(reviews_vectors)
# TODO: need to check consistency of lsi model
print(val.k_fold_validation(reviews, reviews_vectors, 3))
