def build_keras_input():
filename_data, filename_w = './tmp/indexed_data.p', './tmp/Weight.p'
if os.path.isfile(filename_data) and os.path.isfile(filename_w):
data = load_pickle(filename_data)
W = load_pickle(filename_w)
print('Load OK.')
return (data, W)
# load data from pickle
texts, valence, arousal = load_CVAT_2('./resources/CVAT2.0(sigma=1.5).csv')
vocab = get_vocab(texts)
# using word2vec vectors
# word_vecs = load_embeddings('google_news', '/home/hs/Data/Word_Embeddings/google_news.bin')
word_vecs = load_embeddings(
'zh',
'/home/hs/Data/wikipedia/word2vec_word/traditional_wordvecs/wiki.zh.text.traditional_wordvecs.txt'
)
# load glove vectors
# word_vecs = load_embeddings(arg='glove')
word_vecs = add_unknown_words(word_vecs, vocab)
W, word_idx_map = build_embedding_matrix(word_vecs, vocab)
idx_data = make_idx_data(texts, word_idx_map)
data = (idx_data, valence, arousal)
dump_picle(data, filename_data)
dump_picle(W, filename_w)
return (data, W)
def train_doc2vec():
# def isEnglish(s):
# try:
# s.encode('ascii')
# except UnicodeEncodeError:
# return False
# else:
# return True
labeled_data, _ = load_vader('./resource/tweets.txt')
# for i,d in enumerate(labeled_data):
# print(i)
# if not isEnglish(d):
# print('*'*111)
# print(i,d)
# exit()
unlabeled_data, _ = load_sentiment140('/home/hs/Data/Corpus/training.csv')
labeled_data = preprocess(labeled_data, replace=True)
dump_picle(labeled_data, './data/acc/labeled_data.p')
unlabeled_data = preprocess(unlabeled_data, replace=True)
dump_picle(unlabeled_data, './data/acc/unlabeled_data.p')
# labeled_data = load_pickle('./data/acc/labeled_data.p')
# unlabeled_data = load_pickle('./data/acc/unlabeled_data.p')
sentence = TaggedLineSentence(labeled_data, unlabeled_data)
train_docvecs(sentence)
def build_keras_input():
filename_data, filename_w = './tmp/indexed_data.p', './tmp/Weight.p'
if os.path.isfile(filename_data) and os.path.isfile(filename_w):
data = load_pickle(filename_data)
W = load_pickle(filename_w)
print('Load OK.')
return (data, W)
# load data from pickle
texts, valence, arousal = load_CVAT_2('./resources/CVAT2.0(sigma=1.5).csv')
vocab = get_vocab(texts)
# using word2vec vectors
# word_vecs = load_embeddings('google_news', '/home/hs/Data/Word_Embeddings/google_news.bin')
word_vecs = load_embeddings('zh',
'/home/hs/Data/wikipedia/word2vec_word/traditional_wordvecs/wiki.zh.text.traditional_wordvecs.txt')
# load glove vectors
# word_vecs = load_embeddings(arg='glove')
word_vecs = add_unknown_words(word_vecs, vocab)
W, word_idx_map = build_embedding_matrix(word_vecs, vocab)
idx_data = make_idx_data(texts, word_idx_map)
data = (idx_data, valence, arousal)
dump_picle(data, filename_data)
dump_picle(W, filename_w)
return (data, W)
def train_doc2vec():
# def isEnglish(s):
# try:
# s.encode('ascii')
# except UnicodeEncodeError:
# return False
# else:
# return True
labeled_data, _ = load_vader('./resource/tweets.txt')
# for i,d in enumerate(labeled_data):
# print(i)
# if not isEnglish(d):
# print('*'*111)
# print(i,d)
# exit()
unlabeled_data, _ = load_sentiment140('/home/hs/Data/Corpus/training.csv')
labeled_data = preprocess(labeled_data, replace=True)
dump_picle(labeled_data, './data/acc/labeled_data.p')
unlabeled_data = preprocess(unlabeled_data, replace=True)
dump_picle(unlabeled_data, './data/acc/unlabeled_data.p')
# labeled_data = load_pickle('./data/acc/labeled_data.p')
# unlabeled_data = load_pickle('./data/acc/unlabeled_data.p')
sentence = TaggedLineSentence(labeled_data, unlabeled_data)
train_docvecs(sentence)
def kNN(train_data, train_labels, test):
log_state('Use kNN classifier')
clf = KNeighborsClassifier(n_neighbors=5)
clf.fit(train_data, train_labels)
predict_labels = clf.predict(test)
dump_picle(predict_labels, './data/predict_labels/predict_labels.p')
logger.info('kNN classifier training complete, saved predict labels to pickle')
return
def logit(train_data, train_labels, test):
log_state('Use logistic regression classifier')
clf = linear_model.LogisticRegression(C=1e5)
clf.fit(train_data, train_labels)
predict_labels = clf.predict(test)
dump_picle(predict_labels, './data/predict_labels/predict_labels.p')
logger.info('MaxEnt classifier training complete, saved predict labels to pickle')
return
def svm_classify(train_data, train_labels, test):
log_state('Use SVM classifier')
clf = svm.SVC(C=5.0, kernel='linear')
clf.fit(train_data, train_labels)
predict_labels = clf.predict(test)
dump_picle(predict_labels, './data/predict_labels/predict_labels.p')
logger.info('SVM classifier training complete, saved predict labels to pickle')
return
def build_docvecs(model, ratings):
nb_text = len(ratings) # 4200
size = len(model.docvecs['L_SENT_0']) # 50
vecs = [
model.docvecs['L_SENT_%s' % id].reshape((1, size))
for id in range(nb_text)
]
dump_picle((np.concatenate(vecs), ratings), './data/acc/twitter_docvecs.p')
def logit(train_data, train_labels, test):
log_state('Use logistic regression classifier')
clf = linear_model.LogisticRegression(C=1e5)
clf.fit(train_data, train_labels)
predict_labels = clf.predict(test)
dump_picle(predict_labels, './data/predict_labels/predict_labels.p')
logger.info(
'MaxEnt classifier training complete, saved predict labels to pickle')
return
def svm_classify(train_data, train_labels, test):
log_state('Use SVM classifier')
clf = svm.SVC(C=5.0, kernel='linear')
clf.fit(train_data, train_labels)
predict_labels = clf.predict(test)
dump_picle(predict_labels, './data/predict_labels/predict_labels.p')
logger.info(
'SVM classifier training complete, saved predict labels to pickle')
return
def kNN(train_data, train_labels, test):
log_state('Use kNN classifier')
clf = KNeighborsClassifier(n_neighbors=5)
clf.fit(train_data, train_labels)
predict_labels = clf.predict(test)
dump_picle(predict_labels, './data/predict_labels/predict_labels.p')
logger.info(
'kNN classifier training complete, saved predict labels to pickle')
return
def build_ori_anew_vectors(words):
filename = "./tmp/anew_vectors.p"
if os.path.isfile(filename):
return load_pickle(filename)
model = load_embeddings("google_news", "/home/hs/Data/Word_Embeddings/google_news.bin")
vecs = []
for w in words:
vecs.append(model[w])
vecs = np.array(vecs)
dump_picle(vecs, filename)
return vecs
def gNB(train_data, train_labels, test, save_result=False):
log_state('Use Gaussian Naive Bayes classifier')
clf = GaussianNB()
clf.fit(train_data, train_labels)
predict_labels = clf.predict(test)
predict_proba = clf.predict_proba(test)
if save_result == True:
dump_picle(predict_labels, './data/predict_labels/predict_labels.p')
dump_picle(predict_proba, './data/predict_labels/predict_proba.p')
logger.info('Classifier training complete, saved predict labels to pickle')
return predict_labels
def gNB(train_data, train_labels, test, save_result=False):
log_state('Use Gaussian Naive Bayes classifier')
clf = GaussianNB()
clf.fit(train_data, train_labels)
predict_labels = clf.predict(test)
predict_proba = clf.predict_proba(test)
if save_result == True:
dump_picle(predict_labels, './data/predict_labels/predict_labels.p')
dump_picle(predict_proba, './data/predict_labels/predict_proba.p')
logger.info('Classifier training complete, saved predict labels to pickle')
return predict_labels
def build_ori_anew_vectors(words):
filename = './tmp/anew_vectors_glove.p'
if os.path.isfile(filename):
return load_pickle(filename)
model = load_embeddings(arg='zh_tw', filename="D:\Word_Embeddings\English\glove.6B\glove.6B.300d.txt")
vecs = []
for w in words:
vecs.append(model[w])
vecs = np.array(vecs)
dump_picle(vecs, filename)
return vecs
def build_ori_anew_vectors(words):
filename = './tmp/anew_vectors.p'
if os.path.isfile(filename):
return load_pickle(filename)
model = load_embeddings('google_news',
'/home/hs/Data/Word_Embeddings/google_news.bin')
vecs = []
for w in words:
vecs.append(model[w])
vecs = np.array(vecs)
dump_picle(vecs, filename)
return vecs
def build_ori_anew_vectors(words):
filename = './tmp/anew_vectors.p'
if os.path.isfile(filename):
return load_pickle(filename)
model = load_embeddings(
'google_news',
'D:\Word_Embeddings\English\GoogleNews-vectors-negative300.bin')
vecs = []
for w in words:
vecs.append(model[w])
vecs = np.array(vecs)
dump_picle(vecs, filename)
return vecs
def build_amended_vectors(arg='word2vec'):
prefix = None if arg == 'word2vec' else 'GloVe_'
pos_vectors = load_pickle('./tmp/'+prefix+'common_positive_words.p')
neg_vectors = load_pickle('./tmp/'+prefix+'common_negative_words.p')
size = len(pos_vectors[list(pos_vectors.keys())[0]])
print('The dimension of word vectors: %s.' % size)
for k in pos_vectors:
pos_vectors[k]=np.array(pos_vectors[k]).reshape((1, size))
for k in neg_vectors:
neg_vectors[k]=np.array(neg_vectors[k]).reshape((1, size))
amended_pos, amended_neg = amend(pos_vectors, neg_vectors)
dump_picle(amended_pos, './tmp/amended_'+prefix+'pos.p')
dump_picle(amended_neg, './tmp/amended_'+prefix+'neg.p')
def build_ori_anew_vectors(words):
filename = './tmp/anew_vectors_retrofitted_glove.p'
if os.path.isfile(filename):
return load_pickle(filename)
model = load_embeddings(
arg='zh_tw',
filename="D:\Word_Embeddings\English\glove.6B\GloVe_out_vec_file.txt")
vecs = []
for w in words:
vecs.append(model[w])
vecs = np.array(vecs)
dump_picle(vecs, filename)
return vecs
def common_words(word_vectors, words_list):
####################################
# for word2vec model:
# full_words = word_vectors.vocab.keys()
# for GloVe model:
full_words = word_vectors.keys()
####################################
same_words = set(words_list).intersection(full_words)
print('Total Number: %s, same word number: %s.'%(len(words_list), len(same_words)))
vector_dict=dict()
for w in same_words:
vector_dict[w]=word_vectors[w]
dump_picle(vector_dict, './tmp/GloVe_common_negative_words.p')
def common_words(word_vectors, words_list):
####################################
# for word2vec model:
# full_words = word_vectors.vocab.keys()
# for GloVe model:
full_words = word_vectors.keys()
####################################
same_words = set(words_list).intersection(full_words)
print('Total Number: %s, same word number: %s.' %
(len(words_list), len(same_words)))
vector_dict = dict()
for w in same_words:
vector_dict[w] = word_vectors[w]
dump_picle(vector_dict, './tmp/GloVe_common_negative_words.p')
def build_amended_anew_vectors(words):
filename = './tmp/amended_anew_vectors_glove.p'
if os.path.isfile(filename):
return load_pickle(filename)
model = load_embeddings(arg='zh_tw', filename="D:\Word_Embeddings\English\glove.6B\glove.6B.300d.txt")
amended_pos = load_pickle('./tmp/amended_GloVe_pos.p')
amended_neg = load_pickle('./tmp/amended_GloVe_neg.p')
vecs = []
for w in words:
vec = amended_neg.get(w)
if vec is None:
vec = amended_pos.get(w)
if vec is None:
vec = model[w]
vecs.append(vec)
vecs = np.array(vecs)
dump_picle(vecs, filename)
return vecs
def build_amended_anew_vectors(words):
filename = './tmp/retrofitted_anew_vectors_word2vec.p'
if os.path.isfile(filename):
return load_pickle(filename)
model = load_embeddings(arg='zh_tw', filename="D:\Word_Embeddings\English\word2vec_out_vec_file.txt")
amended_pos = load_pickle('./tmp/retrofitted_word2vec_pos.p')
amended_neg = load_pickle('./tmp/retrofitted_word2vec_neg.p')
vecs = []
for w in words:
vec = amended_neg.get(w)
if vec is None:
vec = amended_pos.get(w)
if vec is None:
vec = model[w]
vecs.append(vec)
vecs = np.array(vecs)
dump_picle(vecs, filename)
return vecs
def build_amended_anew_vectors(words):
filename = "./tmp/amended_anew_vectors.p"
if os.path.isfile(filename):
return load_pickle(filename)
model = load_embeddings("google_news", "/home/hs/Data/Word_Embeddings/google_news.bin")
amended_pos = load_pickle("./tmp/amended_pos.p")
amended_neg = load_pickle("./tmp/amended_neg.p")
vecs = []
for w in words:
vec = amended_neg.get(w)
if vec is None:
vec = amended_pos.get(w)
if vec is None:
vec = model[w]
vecs.append(vec)
vecs = np.array(vecs)
dump_picle(vecs, filename)
return vecs
def build_keras_input_amended():
filename_data, filename_w = './tmp/amended_indexed_data.p', './tmp/amended_Weight.p'
if os.path.isfile(filename_data) and os.path.isfile(filename_w):
data = load_pickle(filename_data)
W = load_pickle(filename_w)
print('Load OK.')
return (data, W)
# load data from pickle
(x_train, y_train_valence, y_train_labels,
x_test, y_test_valence, y_test_labels,
x_valid, y_valid_valence, y_valid_labels,
x_train_polarity, y_train_polarity,
x_test_polarity, y_test_polarity,
x_valid_polarity, y_valid_polarity) = load_sst(path='./resources/stanfordSentimentTreebank/')
vocab = get_vocab(x_train)
# word_vecs = load_embeddings('google_news', '/home/hs/Data/Word_Embeddings/google_news.bin')
# word_vecs = load_embeddings('glove')
# load amended word vectors
word_vecs = load_embeddings('amended_word2vec')
word_vecs = add_unknown_words(word_vecs, vocab)
W, word_idx_map = build_embedding_matrix(word_vecs, vocab)
x_train_idx_data = make_idx_data(x_train, word_idx_map)
x_test_idx_data = make_idx_data(x_test, word_idx_map)
x_valid_idx_data = make_idx_data(x_valid, word_idx_map)
x_train_polarity_idx_data = make_idx_data(x_train_polarity, word_idx_map)
x_test_polarity_idx_data = make_idx_data(x_test_polarity, word_idx_map)
x_valid_polarity_idx_data = make_idx_data(x_valid_polarity, word_idx_map)
data = (x_train_idx_data, y_train_valence, y_train_labels,
x_test_idx_data, y_test_valence, y_test_labels,
x_valid_idx_data, y_valid_valence, y_valid_labels,
x_train_polarity_idx_data, y_train_polarity,
x_test_polarity_idx_data, y_test_polarity,
x_valid_polarity_idx_data, y_valid_polarity)
dump_picle(data, filename_data)
dump_picle(W, filename_w)
return (data, W)
def build_keras_input_amended():
filename_data, filename_w = './tmp/amended_indexed_data.p', './tmp/amended_Weight.p'
if os.path.isfile(filename_data) and os.path.isfile(filename_w):
data = load_pickle(filename_data)
W = load_pickle(filename_w)
print('Load OK.')
return (data, W)
# load data from pickle
(x_train, y_train_valence, y_train_labels,
x_test, y_test_valence, y_test_labels,
x_valid, y_valid_valence, y_valid_labels,
x_train_polarity, y_train_polarity,
x_test_polarity, y_test_polarity,
x_valid_polarity, y_valid_polarity) = load_sst(path='./resources/stanfordSentimentTreebank/')
vocab = get_vocab(x_train)
# word_vecs = load_embeddings('google_news', '/home/hs/Data/Word_Embeddings/google_news.bin')
# word_vecs = load_embeddings('glove')
# load amended word vectors
word_vecs = load_embeddings('amended_word2vec')
word_vecs = add_unknown_words(word_vecs, vocab)
W, word_idx_map = build_embedding_matrix(word_vecs, vocab)
x_train_idx_data = make_idx_data(x_train, word_idx_map)
x_test_idx_data = make_idx_data(x_test, word_idx_map)
x_valid_idx_data = make_idx_data(x_valid, word_idx_map)
x_train_polarity_idx_data = make_idx_data(x_train_polarity, word_idx_map)
x_test_polarity_idx_data = make_idx_data(x_test_polarity, word_idx_map)
x_valid_polarity_idx_data = make_idx_data(x_valid_polarity, word_idx_map)
data = (x_train_idx_data, y_train_valence, y_train_labels,
x_test_idx_data, y_test_valence, y_test_labels,
x_valid_idx_data, y_valid_valence, y_valid_labels,
x_train_polarity_idx_data, y_train_polarity,
x_test_polarity_idx_data, y_test_polarity,
x_valid_polarity_idx_data, y_valid_polarity)
dump_picle(data, filename_data)
dump_picle(W, filename_w)
return (data, W)
def build_amended_anew_vectors(words):
filename = './tmp/amended_anew_vectors.p'
if os.path.isfile(filename):
return load_pickle(filename)
model = load_embeddings('google_news',
'/home/hs/Data/Word_Embeddings/google_news.bin')
amended_pos = load_pickle('./tmp/amended_pos.p')
amended_neg = load_pickle('./tmp/amended_neg.p')
vecs = []
for w in words:
vec = amended_neg.get(w)
if vec is None:
vec = amended_pos.get(w)
if vec is None:
vec = model[w]
vecs.append(vec)
vecs = np.array(vecs)
dump_picle(vecs, filename)
return vecs
def build_amended_anew_vectors(words):
filename = './tmp/retrofitted_anew_vectors_word2vec.p'
if os.path.isfile(filename):
return load_pickle(filename)
model = load_embeddings(
arg='zh_tw',
filename="D:\Word_Embeddings\English\word2vec_out_vec_file.txt")
amended_pos = load_pickle('./tmp/retrofitted_word2vec_pos.p')
amended_neg = load_pickle('./tmp/retrofitted_word2vec_neg.p')
vecs = []
for w in words:
vec = amended_neg.get(w)
if vec is None:
vec = amended_pos.get(w)
if vec is None:
vec = model[w]
vecs.append(vec)
vecs = np.array(vecs)
dump_picle(vecs, filename)
return vecs
def build_amended_anew_vectors(words):
filename = './tmp/amended_anew_vectors_glove.p'
if os.path.isfile(filename):
return load_pickle(filename)
model = load_embeddings(
arg='zh_tw',
filename="D:\Word_Embeddings\English\glove.6B\glove.6B.300d.txt")
amended_pos = load_pickle('./tmp/amended_GloVe_pos.p')
amended_neg = load_pickle('./tmp/amended_GloVe_neg.p')
vecs = []
for w in words:
vec = amended_neg.get(w)
if vec is None:
vec = amended_pos.get(w)
if vec is None:
vec = model[w]
vecs.append(vec)
vecs = np.array(vecs)
dump_picle(vecs, filename)
return vecs
def build_keras_input(texts, scores, test, new=True):
dims = 300
# texts, scores are dict type, key: train, dev, devtest.
keys = ["train", "dev", "devtest"]
train, train_scores = texts[keys[0]], scores[keys[0]]
dev, dev_scores = texts[keys[1]], scores[keys[1]]
devtest, devtest_scores = texts[keys[2]], scores[keys[2]]
filename_data, filename_w = './tmp/indexed_data.p', './tmp/Weight.p'
test_filename = './tmp/test_data.p'
if os.path.isfile(filename_data) and os.path.isfile(filename_w) and new == False:
data = load_pickle(filename_data)
W = load_pickle(filename_w)
test_data = load_pickle(test_filename)
print('Use existing data. Load OK.')
return (data, W, test_data)
print("Construct new data.")
# load data from pickle
vocab = get_vocab(train)
# using word2vec vectors
# word_vecs = load_embeddings('google_news', '/home/hs/Data/Word_Embeddings/google_news.bin')
# word_vecs = load_embeddings('D:/Word_Embeddings/glove.840B.300d.txt.w2v')
word_vecs = load_embeddings('/home/hs/Data/Word_Embeddings/glove.840B.300d.txt.w2v')
# word_vecs = load_embeddings('/home/hs/Data/Word_Embeddings/word2vec_twitter_model/word2vec_twitter_model.bin',
# binary=True)
word_vecs = add_unknown_words(word_vecs, vocab, k=dims)
W, word_idx_map = build_embedding_matrix(word_vecs, vocab, k=dims)
idx_data_train = make_idx_data(train, word_idx_map)
idx_data_dev = make_idx_data(dev, word_idx_map)
idx_data_devtest = make_idx_data(devtest, word_idx_map)
idx_data_test = make_idx_data(test[2], word_idx_map)
data = (idx_data_train, idx_data_dev, idx_data_devtest, train_scores, dev_scores, devtest_scores)
test_data = (test[0], test[1], idx_data_test)
dump_picle(data, filename_data)
dump_picle(W, filename_w)
dump_picle(test_data, test_filename)
print("Saved: data and W are saved into: %s, and %s." % (filename_data, filename_w))
return (data, W, test_data)
def load_sst(path=None, level=None):
filename = './tmp/SST.p'
if os.path.isfile(filename):
print('Load OK.')
return load_pickle(filename)
def cleanStr(string):
string = re.sub(r'^A-Za-z0-9(),!?\'\`', ' ', string)
string = re.sub(r'\s{2,}', ' ', string)
string = string.replace('á', 'á').replace('é', 'é').replace('ñ', 'ñ').replace('Â', '').replace('ï', 'ï')
string = string.replace('ü', 'ü').replace('â', 'â').replace('è', 'è').replace('ö', 'ö').replace('æ', 'æ')
string = string.replace('ó', 'ó').replace('û', 'û').replace('ô', 'ô').replace('ã', 'ã').replace('ç', 'ç')
string = string.replace('Ã ', 'à ').replace('Ã', 'í').replace('í', 'í')
return string
# sentiment label
sentiment_file = open(path + 'sentiment_labels.txt', 'r')
sentiment_label = {}
n = 0
for line in sentiment_file:
lines = line.strip().split('|')
if n > 0:
sentiment_label[int(lines[0])] = float(lines[1])
n += 1
sentiment_file.close()
# phrase dict
dict_file = open(path + 'dictionary.txt', 'r')
phrase_dict = {}
for line in dict_file:
# line = line.decode('utf-8')
lines = line.strip().split('|')
phrase_dict[lines[0]] = int(lines[1])
dict_file.close()
# sentence dict
sentence_file = open(path + 'datasetSentences.txt', 'r')
sentence_dict = {}
n = 0
for line in sentence_file:
# line = line.decode('utf-8')
line = line.replace('-LRB-', '(')
line = line.replace('-RRB-', ')')
lines = line.strip().split('\t')
if n > 0:
sentence_dict[int(lines[0])] = lines[1]
n += 1
sentence_file.close()
# datasplit
datasplit_file = open(path + 'datasetSplit.txt', 'r')
split_dict = {}
n = 0
for line in datasplit_file:
lines = line.strip().split(',')
if n > 0:
split_dict[int(lines[0])] = int(lines[1])
n += 1
datasplit_file.close()
size = len(sentence_dict) # size = 11855
# for i in range(1000):
# senti = sentiment_label[phrase_dict[cleanStr(sentence_dict[i + 1])]]
# print(i, senti, cleanStr(sentence_dict[i + 1]))
# exit()
x_train, y_train_valence, y_train_labels = [], [], []
x_test, y_test_valence, y_test_labels = [], [], []
x_valid, y_valid_valence, y_valid_labels = [], [], []
x_train_polarity, y_train_polarity = [], []
x_test_polarity, y_test_polarity = [], []
x_valid_polarity, y_valid_polarity = [], []
for i in range(size):
# print sentence_dict[i+1].encode('utf-8')
sentence = cleanStr(sentence_dict[i + 1])
senti = sentiment_label[phrase_dict[sentence]]
# print(senti, sentence)
labels, polarity = None, None
if 0 <= senti <= 0.2:
labels = 1
polarity = 0
if 0.2 < senti <= 0.4:
labels = 2
polarity = 0
if 0.4 < senti <= 0.6:
labels = 3
if 0.6 < senti <= 0.8:
labels = 4
polarity = 1
if 0.8 < senti <= 1:
labels = 5
polarity = 1
if labels is None:
raise Exception('Sentiment Error !')
if split_dict[i + 1] == 1:
x_train.append(sentence)
y_train_valence.append(senti)
y_train_labels.append(labels)
if polarity is not None:
x_train_polarity.append(sentence)
y_train_polarity.append(polarity)
elif split_dict[i + 1] == 2:
x_test.append(sentence)
y_test_valence.append(senti)
y_test_labels.append(labels)
if polarity is not None:
x_test_polarity.append(sentence)
y_test_polarity.append(polarity)
else:
x_valid.append(sentence)
y_valid_valence.append(senti)
y_valid_labels.append(labels)
if polarity is not None:
x_valid_polarity.append(sentence)
y_valid_polarity.append(polarity)
print("Fine-grained: #training: %s, #valid: %s, #test: %s" % (len(x_train), len(x_valid), len(x_test)))
print("Binary classification: #train: %s, #valid: %s, #test: %s" % (
len(x_train_polarity), len(x_valid_polarity), len(x_test_polarity)))
# t = zip(x_train, y_train)
# random.shuffle(t)
# x_train, y_train = zip(*t)
output = (x_train, y_train_valence, y_train_labels,
x_test, y_test_valence, y_test_labels,
x_valid, y_valid_valence, y_valid_labels,
x_train_polarity, y_train_polarity,
x_test_polarity, y_test_polarity,
x_valid_polarity, y_valid_polarity)
dump_picle(output, filename)
print('Data saved and load successfully.')
return output
def build_docvecs(model, ratings):
nb_text = len(ratings) # 4200
size = len(model.docvecs['L_SENT_0']) # 50
vecs = [model.docvecs['L_SENT_%s' % id].reshape((1, size)) for id in range(nb_text)]
dump_picle((np.concatenate(vecs), ratings), './data/acc/twitter_docvecs.p')
# corpus, ratings = load_vader(['movie_reviews'])
# # name: tweets, movie_reviews, product_reviews, news_articles
lexicon_name = get_file_path('anew')
words, valences, _ = load_anew(lexicon_name)
corpus, ratings = screen_data(corpus, ratings, words)
ratings = np.array(ratings) + np.ones(len(ratings), dtype=float) * 5
print(len(corpus), len(corpus))
vocab = get_vocab(corpus)
# dump_picle(vocab, './data/corpus/vader/vocab_moview_reviews.p')
# vocab = load_pickle('./data/corpus/vader/vocab_moview_reviews.p')
print(len(vocab))
W, word_idx_map = build_embedding_matrix(load_embeddings('google_news'), vocab, k=300)
# dump_picle(word_idx_map, './data/corpus/vader/word_idx_map_movie_reviews.p')
# print('dump word_idx_map successful')
dump_picle(W, './data/corpus/vader/embedding_matrix_all.p')
print('dump embedding matrix file OK')
c
idx_data = make_idx_data(corpus, word_idx_map, max_len=200, kernel_size=5)
dump_picle([idx_data, ratings], './data/corpus/vader/vader_processed_data_all.p')
print(idx_data[0])
print(ratings[0])
############################################## tweets ###############################################
# corpus, ratings = load_vader(['tweets', 'movie_reviews', 'product_reviews', 'news_articles'])
corpus, ratings = load_vader(['tweets'])
# # name: tweets, movie_reviews, product_reviews, news_articles
lexicon_name = get_file_path('anew')
words, valences, _ = load_anew(lexicon_name)
corpus, ratings = screen_data(corpus, ratings, words)
ratings = np.array(ratings) + np.ones(len(ratings), dtype=float) * 5
from load_data import load_CVAT_2
filename = './resources/CVAT (utf-8).csv'
texts, valence, arousal = load_CVAT_2(filename, categorical="all")
len_text = []
from CKIP_tokenizer import segsentence
out = []
for idx, i in enumerate(texts):
# print(list(i))
print(idx)
out.append(" ".join(segsentence(i)))
# len_text.append(len(.split()))
from save_data import dump_picle
dump_picle(out, "tokenized_texts_(newest3.31).p")
print("The tokenized text is saved.")
vectorizer_param = {'preprocessor': preprocessor, 'ngram_range': parameters['ngram_range'], 'analyzer': 'word',
'min_df': parameters['min_df'], 'max_df': parameters['max_df'],
'binary': parameters['TF_binary'], 'norm': parameters['norm'],
'sublinear_tf': parameters['sublinear_tf'], 'max_features': parameters['max_features']}
if __name__ == "__main__":
unigram = StemmedTfidfVectorizer(**vectorizer_param)
anew = anew_vectorizer()
pct = punctuation_estimator()
strength = strength_vectorizer()
avg_strength = avg_affective_vectorizer()
log_state('combine unigram and avg strength features')
combined_features = FeatureUnion([('unigram', unigram), ('avg_strength', avg_strength)])
# log_state('combine unigram and strength features')
# combined_features =FeatureUnion([('unigram',unigram),('strength',strength)])
# log_state('combine unigram and anew features')
# combined_features =FeatureUnion([('unigram',unigram),('anew',anew)])
# log_state('combine unigram and punctuation features')
# combined_features =FeatureUnion([('unigram',unigram),('pct',pct)])
texts, _ = load_train_data('Sentiment140')
transformed_train = combined_features.fit_transform(texts)
testdata, _ = load_test_data()
transformed_test = combined_features.transform(testdata)
dump_picle(combined_features.get_feature_names(), './data/features/feature_names.p')
dump_picle(transformed_train, "./data/transformed_data/transformed_train.p")
dump_picle(transformed_test, "./data/transformed_data/transformed_test.p")
neg_idx_data, neg_length = load_data(file_dir + 'neg/')
print(neg_idx_data.shape, neg_length)
data = np.concatenate((pos_idx_data, neg_idx_data), axis=0)
print(data.shape)
return data, pos_length, neg_length
if __name__ == '__main__':
########################################## config ########################################
file_dir = 'E:/研究/Data/IMDB/aclImdb/train/' if os.name == 'nt' else '/home/hs/Data/imdb/aclImdb/train/'
vec_dim = 300
##########################################################################################
# get vocab and save to pickle
vocab = get_vocab(file_dir)
dump_picle(vocab, './data/tmp/vocab.p')
print('OK')
a = load_pickle('./data/tmp/vocab.p')
for i in a:
print(i)
print(len(a))
exit()
# end
# make word index map
vocab = load_pickle('./data/tmp/vocab.p')
W, word_idx_map = build_embedding_matrix(load_embeddings('google_news'), vocab, k=300)
dump_picle(word_idx_map, get_file_path('word_idx_map'))
print('dump word_idx_map successful')
dump_picle(W, '/home/hs/Data/embedding_matrix.p')
print('OK')
def keras_nn_input(word_vectors_model, amending):
if word_vectors_model == 'word2vec':
if amending == True:
filename_data, filename_w = './tmp/amended_w2v_indexed_data.p', './tmp/amended_w2v_Weight.p'
elif amending == False:
filename_data, filename_w = './tmp/w2v_indexed_data.p', './tmp/w2v_Weight.p'
else:
raise Exception('Wrong!')
elif word_vectors_model == 'GloVe':
if amending == True:
filename_data, filename_w = './tmp/amended_GloVe_indexed_data.p', './tmp/amended_GloVe_Weight.p'
elif amending == False:
filename_data, filename_w = './tmp/GloVe_indexed_data.p', './tmp/GloVe_Weight.p'
else:
raise Exception('Wrong!')
elif word_vectors_model == 'retrofitted_GloVe':
filename_data, filename_w = './tmp/retrofitted_GloVe_indexed_data.p', './tmp/retrofitted_GloVe_Weight.p'
elif word_vectors_model == 'retrofitted_word2vec':
filename_data, filename_w = './tmp/retrofitted_word2vec_indexed_data.p', './tmp/retrofitted_word2vec_Weight.p'
else:
raise Exception('Wrong parameter!')
if os.path.isfile(filename_data) and os.path.isfile(filename_w):
data = load_pickle(filename_data)
W = load_pickle(filename_w)
print('Load OK, parameters: word_vectors_model = %s, amending = %s'%(word_vectors_model, amending))
return (data, W)
# load data from pickle
(x_train, y_train_valence, y_train_labels,
x_test, y_test_valence, y_test_labels,
x_valid, y_valid_valence, y_valid_labels,
x_train_polarity, y_train_polarity,
x_test_polarity, y_test_polarity,
x_valid_polarity, y_valid_polarity) = load_sst(path='./resources/stanfordSentimentTreebank/')
vocab = get_vocab(x_train)
if word_vectors_model == 'word2vec':
if amending == True:
word_vecs = load_embeddings('amended_word2vec')
elif amending == False:
word_vecs = load_embeddings('google_news', '/home/hs/Data/Word_Embeddings/google_news.bin')
else:
raise Exception('Wrong!')
elif word_vectors_model == 'GloVe':
if amending == True:
word_vecs = load_embeddings('amended_glove')
elif amending == False:
word_vecs = load_embeddings('glove')
else:
raise Exception('Wrong!')
elif word_vectors_model == 'retrofitted_GloVe':
word_vecs = load_embeddings('zh_tw', 'D:\Word_Embeddings\English\glove.6B\GloVe_out_vec_file.txt')
# convert gensim model to dict type
w2v = dict()
for key in word_vecs.vocab.keys():
w2v[key] = word_vecs[key]
word_vecs = w2v
elif word_vectors_model == 'retrofitted_word2vec':
word_vecs = load_embeddings('zh_tw', 'D:\Word_Embeddings\English\word2vec_out_vec_file.txt')
# convert gensim model to dict type
w2v = dict()
for key in word_vecs.vocab.keys():
w2v[key] = word_vecs[key]
word_vecs = w2v
else:
raise Exception('Wrong parameter!')
word_vecs = add_unknown_words(word_vecs, vocab)
W, word_idx_map = build_embedding_matrix(word_vecs, vocab)
x_train_idx_data = make_idx_data(x_train, word_idx_map)
x_test_idx_data = make_idx_data(x_test, word_idx_map)
x_valid_idx_data = make_idx_data(x_valid, word_idx_map)
x_train_polarity_idx_data = make_idx_data(x_train_polarity, word_idx_map)
x_test_polarity_idx_data = make_idx_data(x_test_polarity, word_idx_map)
x_valid_polarity_idx_data = make_idx_data(x_valid_polarity, word_idx_map)
data = (x_train_idx_data, y_train_valence, y_train_labels,
x_test_idx_data, y_test_valence, y_test_labels,
x_valid_idx_data, y_valid_valence, y_valid_labels,
x_train_polarity_idx_data, y_train_polarity,
x_test_polarity_idx_data, y_test_polarity,
x_valid_polarity_idx_data, y_valid_polarity)
dump_picle(data, filename_data)
dump_picle(W, filename_w)
print('Load OK, parameters: word_vectors_model = %s, amending = %s'%(word_vectors_model, amending))
return (data, W)
def convert(source_file):
s = load_pickle(source_file)
dump_picle(s, str(source_file)[:-2] + '_v2.7.p', protocol=2)
vocab = defaultdict(float)
for sent in corpus:
for word in sent:
vocab[word] += 1
print(len(vocab))
return vocab
# 注意： 这个文件就是CVAT构造cnn输入数据的代码
########################################## config ########################################
vec_dim = 400
##########################################################################################
corpus = load_corpus(get_file_path('cn_corpus'))
print(corpus[:2])
vocab = get_vocab(corpus)
dump_picle(vocab, get_file_path('CVAT_Vocab'))
print('Dump CVAT vocab OK')
# vocab = load_pickle(get_file_path('CVAT_Vocab'))
for i in vocab:
print(i)
print(len(vocab))
W, word_idx_map = build_embedding_matrix(load_embeddings('zh_tw'),
vocab,
k=400)
dump_picle(word_idx_map, get_file_path('word_idx_map_CVAT'))
print('dump word_idx_map successful')
dump_picle(W, './data/tmp/embedding_matrix_CVAT.p')
print('OK')
# word_idx_map = load_pickle(get_file_path('word_idx_map_CVAT'))
neg_idx_data, neg_length = load_data(file_dir + 'neg/')
print(neg_idx_data.shape, neg_length)
data = np.concatenate((pos_idx_data, neg_idx_data), axis=0)
print(data.shape)
return data, pos_length, neg_length
if __name__ == '__main__':
########################################## config ########################################
file_dir = 'E:/研究/Data/IMDB/aclImdb/train/' if os.name == 'nt' else '/home/hs/Data/imdb/aclImdb/train/'
vec_dim = 300
##########################################################################################
# get vocab and save to pickle
vocab = get_vocab(file_dir)
dump_picle(vocab, './data/tmp/vocab.p')
print('OK')
a = load_pickle('./data/tmp/vocab.p')
for i in a:
print(i)
print(len(a))
exit()
# end
# make word index map
vocab = load_pickle('./data/tmp/vocab.p')
W, word_idx_map = build_embedding_matrix(load_embeddings('google_news'),
vocab,
k=300)
dump_picle(word_idx_map, get_file_path('word_idx_map'))
print('dump word_idx_map successful')
lexicon_name = get_file_path('anew')
words, valences, _ = load_anew(lexicon_name)
corpus, ratings = screen_data(corpus, ratings, words)
ratings = np.array(ratings) + np.ones(len(ratings), dtype=float) * 5
print(len(corpus), len(corpus))
vocab = get_vocab(corpus)
# dump_picle(vocab, './data/corpus/vader/vocab_moview_reviews.p')
# vocab = load_pickle('./data/corpus/vader/vocab_moview_reviews.p')
print(len(vocab))
W, word_idx_map = build_embedding_matrix(load_embeddings('google_news'),
vocab,
k=300)
# dump_picle(word_idx_map, './data/corpus/vader/word_idx_map_movie_reviews.p')
# print('dump word_idx_map successful')
dump_picle(W, './data/corpus/vader/embedding_matrix_all.p')
print('dump embedding matrix file OK')
# word_idx_map = load_pickle('./data/corpus/vader/word_idx_map_movie_reviews.p')
idx_data = make_idx_data(corpus, word_idx_map, max_len=200, kernel_size=5)
dump_picle([idx_data, ratings],
'./data/corpus/vader/vader_processed_data_all.p')
print(idx_data[0])
print(ratings[0])
############################################## tweets ###############################################
# corpus, ratings = load_vader(['tweets', 'movie_reviews', 'product_reviews', 'news_articles'])
corpus, ratings = load_vader(['tweets'])
# # name: tweets, movie_reviews, product_reviews, news_articles
lexicon_name = get_file_path('anew')
words, valences, _ = load_anew(lexicon_name)
corpus, ratings = screen_data(corpus, ratings, words)
def load_sst(path=None, level=None):
filename = './tmp/SST.p'
if os.path.isfile(filename):
print('Load OK.')
return load_pickle(filename)
def cleanStr(string):
string = re.sub(r'^A-Za-z0-9(),!?\'\`', ' ', string)
string = re.sub(r'\s{2,}', ' ', string)
string = string.replace('á', 'á').replace('é', 'é').replace(
'ñ', 'ñ').replace('Â', '').replace('ï', 'ï')
string = string.replace('ü', 'ü').replace('â', 'â').replace(
'è', 'è').replace('ö', 'ö').replace('æ', 'æ')
string = string.replace('ó', 'ó').replace('û', 'û').replace(
'ô', 'ô').replace('ã', 'ã').replace('ç', 'ç')
string = string.replace('Ã ', 'à ').replace('Ã',
'í').replace('í', 'í')
return string
# sentiment label
sentiment_file = open(path + 'sentiment_labels.txt', 'r')
sentiment_label = {}
n = 0
for line in sentiment_file:
lines = line.strip().split('|')
if n > 0:
sentiment_label[int(lines[0])] = float(lines[1])
n += 1
sentiment_file.close()
# phrase dict
dict_file = open(path + 'dictionary.txt', 'r')
phrase_dict = {}
for line in dict_file:
# line = line.decode('utf-8')
lines = line.strip().split('|')
phrase_dict[lines[0]] = int(lines[1])
dict_file.close()
# sentence dict
sentence_file = open(path + 'datasetSentences.txt', 'r')
sentence_dict = {}
n = 0
for line in sentence_file:
# line = line.decode('utf-8')
line = line.replace('-LRB-', '(')
line = line.replace('-RRB-', ')')
lines = line.strip().split('\t')
if n > 0:
sentence_dict[int(lines[0])] = lines[1]
n += 1
sentence_file.close()
# datasplit
datasplit_file = open(path + 'datasetSplit.txt', 'r')
split_dict = {}
n = 0
for line in datasplit_file:
lines = line.strip().split(',')
if n > 0:
split_dict[int(lines[0])] = int(lines[1])
n += 1
datasplit_file.close()
size = len(sentence_dict) # size = 11855
# for i in range(1000):
# senti = sentiment_label[phrase_dict[cleanStr(sentence_dict[i + 1])]]
# print(i, senti, cleanStr(sentence_dict[i + 1]))
# exit()
x_train, y_train_valence, y_train_labels = [], [], []
x_test, y_test_valence, y_test_labels = [], [], []
x_valid, y_valid_valence, y_valid_labels = [], [], []
x_train_polarity, y_train_polarity = [], []
x_test_polarity, y_test_polarity = [], []
x_valid_polarity, y_valid_polarity = [], []
for i in range(size):
# print sentence_dict[i+1].encode('utf-8')
sentence = cleanStr(sentence_dict[i + 1])
senti = sentiment_label[phrase_dict[sentence]]
# print(senti, sentence)
labels, polarity = None, None
if 0 <= senti <= 0.2:
labels = 1
polarity = 0
if 0.2 < senti <= 0.4:
labels = 2
polarity = 0
if 0.4 < senti <= 0.6:
labels = 3
if 0.6 < senti <= 0.8:
labels = 4
polarity = 1
if 0.8 < senti <= 1:
labels = 5
polarity = 1
if labels is None:
raise Exception('Sentiment Error !')
if split_dict[i + 1] == 1:
x_train.append(sentence)
y_train_valence.append(senti)
y_train_labels.append(labels)
if polarity is not None:
x_train_polarity.append(sentence)
y_train_polarity.append(polarity)
elif split_dict[i + 1] == 2:
x_test.append(sentence)
y_test_valence.append(senti)
y_test_labels.append(labels)
if polarity is not None:
x_test_polarity.append(sentence)
y_test_polarity.append(polarity)
else:
x_valid.append(sentence)
y_valid_valence.append(senti)
y_valid_labels.append(labels)
if polarity is not None:
x_valid_polarity.append(sentence)
y_valid_polarity.append(polarity)
print("Fine-grained: #training: %s, #valid: %s, #test: %s" %
(len(x_train), len(x_valid), len(x_test)))
print("Binary classification: #train: %s, #valid: %s, #test: %s" %
(len(x_train_polarity), len(x_valid_polarity), len(x_test_polarity)))
# t = zip(x_train, y_train)
# random.shuffle(t)
# x_train, y_train = zip(*t)
output = (x_train, y_train_valence, y_train_labels, x_test, y_test_valence,
y_test_labels, x_valid, y_valid_valence, y_valid_labels,
x_train_polarity, y_train_polarity, x_test_polarity,
y_test_polarity, x_valid_polarity, y_valid_polarity)
dump_picle(output, filename)
print('Data saved and load successfully.')
return output
from vectorizers import TFIDF_estimator, anew_estimator
# class StemmedTfidfVectorizer(TfidfVectorizer):
# def build_analyzer(self):
# # 利用NLTK进行词干化处理
# english_stemmer = nltk.stem.SnowballStemmer('english')
# analyzer = super(TfidfVectorizer, self).build_analyzer()
# return lambda doc: (english_stemmer.stem(w) for w in analyzer(doc))
#
#
# vectorizer_param = {'preprocessor': preprocessor, 'ngram_range': parameters['ngram_range'], 'analyzer': 'word',
# 'min_df': parameters['min_df'], 'max_df': parameters['max_df'],
# 'binary': parameters['TF_binary'], 'norm': parameters['norm'],
# 'sublinear_tf': parameters['sublinear_tf'], 'max_features': parameters['max_features']}
# log_state('Start generating features')
# log_state((sorted(list(vectorizer_param.items()))))
# log_state('Training data size: ' + str(parameters['test_data_size']))
if __name__ == "__main__":
vectorizer = anew_estimator()
train_type = 'Sentiment140'
texts, _ = load_train_data(train_type)
transformed_train = vectorizer.fit_transform(texts)
testdata, _ = load_test_data()
transformed_test = vectorizer.transform(testdata)
dump_picle(vectorizer.get_feature_names(),
'./data/features/feature_names.p')
dump_picle(transformed_train,
"./data/transformed_data/transformed_train.p")
dump_picle(transformed_test, "./data/transformed_data/transformed_test.p")
print(embedding_matrix[1])
print(idx_map['我們'])
print(len(word_vecs['我們']))
print(word_vecs['我們'].shape)
print(build_sentence_matrix(model=word_vecs, sententces=corpus[:2], dim=dim))
print('Result')
sentence_embedding_matrix = build_sentence_matrix(word_vecs, corpus, dim=dim)
print(sentence_embedding_matrix.shape)
print(sentence_embedding_matrix[3], valence[3], arousal[3])
from save_data import dump_picle
dump_picle((sentence_embedding_matrix, valence),
get_file_path('CVAT_sentence_matrix_400'))
exit()
'''
X_train, X_test, Y_train, Y_test = cross_validation.train_test_split(sentence_embedding_matrix, valence, test_size=0.2,
random_state=0)
print(X_train.shape)
print(len(Y_test))
maxlen = 200
size = 50
X_train = X_train.reshape(X_train.shape[0], 1, maxlen, size)
X_test = X_test.reshape(X_test.shape[0], 1, maxlen, size)
print(X_train.shape)
batch_size = 128
def keras_nn_input(word_vectors_model, amending):
if word_vectors_model == 'word2vec':
if amending == True:
filename_data, filename_w = './tmp/amended_w2v_indexed_data.p', './tmp/amended_w2v_Weight.p'
elif amending == False:
filename_data, filename_w = './tmp/w2v_indexed_data.p', './tmp/w2v_Weight.p'
else:
raise Exception('Wrong!')
elif word_vectors_model == 'GloVe':
if amending == True:
filename_data, filename_w = './tmp/amended_GloVe_indexed_data.p', './tmp/amended_GloVe_Weight.p'
elif amending == False:
filename_data, filename_w = './tmp/GloVe_indexed_data.p', './tmp/GloVe_Weight.p'
else:
raise Exception('Wrong!')
else:
raise Exception('Wrong parameter!')
if os.path.isfile(filename_data) and os.path.isfile(filename_w):
data = load_pickle(filename_data)
W = load_pickle(filename_w)
print('Load OK, parameters: word_vectors_model = %s, amending = %s'%(word_vectors_model, amending))
return (data, W)
# load data from pickle
(x_train, y_train_valence, y_train_labels,
x_test, y_test_valence, y_test_labels,
x_valid, y_valid_valence, y_valid_labels,
x_train_polarity, y_train_polarity,
x_test_polarity, y_test_polarity,
x_valid_polarity, y_valid_polarity) = load_sst(path='./resources/stanfordSentimentTreebank/')
vocab = get_vocab(x_train)
if word_vectors_model == 'word2vec':
if amending == True:
word_vecs = load_embeddings('amended_word2vec')
elif amending == False:
word_vecs = load_embeddings('google_news', '/home/hs/Data/Word_Embeddings/google_news.bin')
else:
raise Exception('Wrong!')
elif word_vectors_model == 'GloVe':
if amending == True:
word_vecs = load_embeddings('amended_glove')
elif amending == False:
word_vecs = load_embeddings('glove')
else:
raise Exception('Wrong!')
else:
raise Exception('Wrong parameter!')
word_vecs = add_unknown_words(word_vecs, vocab)
W, word_idx_map = build_embedding_matrix(word_vecs, vocab)
x_train_idx_data = make_idx_data(x_train, word_idx_map)
x_test_idx_data = make_idx_data(x_test, word_idx_map)
x_valid_idx_data = make_idx_data(x_valid, word_idx_map)
x_train_polarity_idx_data = make_idx_data(x_train_polarity, word_idx_map)
x_test_polarity_idx_data = make_idx_data(x_test_polarity, word_idx_map)
x_valid_polarity_idx_data = make_idx_data(x_valid_polarity, word_idx_map)
data = (x_train_idx_data, y_train_valence, y_train_labels,
x_test_idx_data, y_test_valence, y_test_labels,
x_valid_idx_data, y_valid_valence, y_valid_labels,
x_train_polarity_idx_data, y_train_polarity,
x_test_polarity_idx_data, y_test_polarity,
x_valid_polarity_idx_data, y_valid_polarity)
dump_picle(data, filename_data)
dump_picle(W, filename_w)
print('Load OK, parameters: word_vectors_model = %s, amending = %s'%(word_vectors_model, amending))
return (data, W)
print(embedding_matrix[1])
print(idx_map['我們'])
print(len(word_vecs['我們']))
print(word_vecs['我們'].shape)
print(build_sentence_matrix(model=word_vecs, sententces=corpus[:2], dim=dim))
print('Result')
sentence_embedding_matrix = build_sentence_matrix(word_vecs, corpus, dim=dim)
print(sentence_embedding_matrix.shape)
print(sentence_embedding_matrix[3], valence[3], arousal[3])
from save_data import dump_picle
dump_picle((sentence_embedding_matrix, valence), get_file_path('CVAT_sentence_matrix_400'))
exit()
'''
X_train, X_test, Y_train, Y_test = cross_validation.train_test_split(sentence_embedding_matrix, valence, test_size=0.2,
random_state=0)
print(X_train.shape)
print(len(Y_test))
maxlen = 200
size = 50
X_train = X_train.reshape(X_train.shape[0], 1, maxlen, size)
X_test = X_test.reshape(X_test.shape[0], 1, maxlen, size)
print(X_train.shape)
lexicon_name = get_file_path('anew')
logger.info(r"loading lexicon form : " + lexicon_name)
words, valences, _ = load_anew(lexicon_name)
corpus, ratings = screen_data(corpus, ratings, words)
ratings = np.array(ratings) + np.ones(len(ratings), dtype=float) * 5
print(len(corpus))
# for i in corpus[:100]:
# print(i)
lexicon = dict()
for i, word in enumerate(words):
lexicon[word] = valences[i]
mean_ratings, tf_means, tfidf_means, geos, tf_geos, tfidf_geos = calculate_ratings(
corpus, ratings, lexicon)
dump_picle([
mean_ratings, tf_means, tfidf_means, geos, tf_geos, tfidf_geos, ratings
], './data/vader_out.p')
exit()
from collections import defaultdict
# idf = sp.log(float(len(D)) / (len([doc.split() for doc in D if t in doc.split()])))
# vocab = load_pickle('./data/corpus/vader/vocab_moview_reviews.p')
# length = len(vocab)
##################################### IDF ####################################################
# idf=dict()
# for i, word in enumerate(words):
# denominator = sum(1 for doc in corpus if word in doc.split())
# if denominator != 0:
# idf[word] = sp.log(float(len(corpus)) / denominator)
# if i%50 == 0:
for i in top_n_ind:
print('Parameter setting: %s, acc: %s' % (str(list(grid)[i]), param_fitness[i]))
if __name__=='__main__':
result_analysis('./tmp/grid_search_result.p')
print('OK')
exit()
scope = [0., 0.1, 0.2, 0.3, 0.4, 0.5, 0.6, 0.7, 0.8, 0.9]
param_grid = {'a':scope, 'b': scope, 'c': scope}
param_fitness = []
grid = ParameterGrid(param_grid)
for params in grid:
print('calculating... parameter: %s' % str(params))
score = my_function(params['a'], params['b'], params['c'])
print('Score: %s' % score)
param_fitness.append(score)
print('grid search complete.')
# return the best fitness value and its settings
best_fitness = np.min(np.array(param_fitness))
best_ind = np.where(np.array(param_fitness)==best_fitness)[0]
print('best fitness: %s' % best_fitness)
print('best setting: %s' % str(list(grid)[best_ind]))
dump_picle((param_grid, param_fitness), './tmp/grid_search_result.p')
def get_vocab(corpus):
vocab = defaultdict(float)
for sent in corpus:
for word in sent:
vocab[word] += 1
print(len(vocab))
return vocab
# 注意： 这个文件就是CVAT构造cnn输入数据的代码
########################################## config ########################################
vec_dim = 400
##########################################################################################
corpus = load_corpus(get_file_path('cn_corpus'))
print(corpus[:2])
vocab = get_vocab(corpus)
dump_picle(vocab, get_file_path('CVAT_Vocab'))
print('Dump CVAT vocab OK')
# vocab = load_pickle(get_file_path('CVAT_Vocab'))
for i in vocab:
print(i)
print(len(vocab))
W, word_idx_map = build_embedding_matrix(load_embeddings('zh_tw'), vocab, k=400)
dump_picle(word_idx_map, get_file_path('word_idx_map_CVAT'))
print('dump word_idx_map successful')
dump_picle(W, './data/tmp/embedding_matrix_CVAT.p')
print('OK')
# word_idx_map = load_pickle(get_file_path('word_idx_map_CVAT'))
mark = load_mark(get_file_path('mark'))
valence, arousal = gold_valence_arousal(corpus, mark)
return [clean_str(sent) for sent in corpus]
vec_dim = 300
############################################## tweets ###############################################
# corpus, ratings = load_vader(['tweets', 'movie_reviews', 'product_reviews', 'news_articles'])
corpus, ratings = load_vader(['tweets'])
# # name: tweets, movie_reviews, product_reviews, news_articles
lexicon_name = get_file_path('anew')
words, valences, _ = load_anew(lexicon_name)
corpus, ratings = screen_data(corpus, ratings, words)
ratings = np.array(ratings) + np.ones(len(ratings), dtype=float) * 5
print(len(corpus), len(corpus))
vocab = get_vocab(corpus)
dump_picle(vocab, './data/corpus/vader/vocab_moview_tweets.p')
# vocab = load_pickle('./data/corpus/vader/vocab_moview_reviews.p')
print('词汇数量：%s' % str(len(vocab)))
W, word_idx_map = build_embedding_matrix(load_embeddings('google_news'), vocab, k=300)
dump_picle(word_idx_map, './data/corpus/vader/word_idx_map_tweets.p')
print('dump word_idx_map successful')
dump_picle(W, './data/corpus/vader/embedding_matrix_tweets.p')
print('dump embedding matrix file OK')
# word_idx_map = load_pickle('./data/corpus/vader/word_idx_map_movie_reviews.p')
idx_data = make_idx_data(corpus, word_idx_map, max_len=200, kernel_size=5)
dump_picle([idx_data, ratings], './data/corpus/vader/vader_processed_data_tweets.p')
print(idx_data[0])
print(ratings[0])
print('OK')
