def process(self):
log.info("Commencing execution")
# Get tagged articles from Veriday
log.info("Getting tagged Veriday articles ... ")
veriday_articles_raw = file_helper.get_articles_list(self.articles_source_file_path)
veriday_tagged_articles = doc2vec_helper.get_tagged_articles_veriday(veriday_articles_raw)
# Convert articles file into a Tagged documents for doc2vec
log.info("Getting tagged Semeval articles ... ")
articles = file_helper.get_articles_list(self.labeled_articles_file_path)
tagged_articles, sentiment_scores_dict = doc2vec_helper.get_tagged_articles_scores(articles)
# combine both article sets
tagged_articles.extend(veriday_tagged_articles)
# model initialization and vocab building
log.info("Initializing the doc2vec model ...")
doc2vec_model = doc2vec_helper.init_model(tagged_articles)
# shuffling and training the model
log.info("Training the doc2vec model ...")
for i in range(self.shuffle_count):
log.info("Shuffles remaining: " + str(self.shuffle_count - i))
doc2vec_helper.shuffle_and_train_articles(doc2vec_model, tagged_articles)
# saving the doc2vec model to disk
doc2vec_model.save(self.doc2vec_model_file_path)
# Extracting parameters for and training the ML model
x_docvecs, y_scores = scikit_ml_helper.extract_training_parameters(doc2vec_model, sentiment_scores_dict)
log.info("Training the ML model ...")
ml_model = scikit_ml_helper.train_linear_model(x_docvecs, y_scores)
# saving the ml model to disk
scikit_ml_helper.persist_model_to_disk(ml_model, self.ml_model_file_path)
log.info("Completed execution")
def process(self):
log.info("Commencing execution")
# Get tagged articles from Semeval
log.info("Getting Semeval articles ... ")
semeval_articles_raw = file_helper.get_articles_list(
self.labeled_articles_file_path)
semeval_tagged_articles, document_sentiment_classes = \
doc2vec_helper.get_tagged_semeval_articles(semeval_articles_raw)
# model initialization and vocab building
log.info("Initializing the doc2vec model ...")
doc2vec_model = doc2vec_helper.init_model(semeval_tagged_articles)
# shuffling and training the model
log.info("Training the doc2vec model ...")
for i in range(self.shuffle_count):
log.info("Shuffles remaining: " + str(self.shuffle_count - i))
doc2vec_helper.shuffle_and_train_articles(doc2vec_model,
semeval_tagged_articles)
# saving the doc2vec model to disk
doc2vec_model.save(self.doc2vec_model_file_path)
# Extracting parameters for and training the ML model
x_docvecs, y_scores = scikit_ml_helper.extract_training_parameters(
doc2vec_model, document_sentiment_classes)
log.info("Training the ML model ...")
# ml_model = scikit_ml_helper.train_linear_model(x_docvecs, y_scores)
ml_model = scikit_ml_helper.train_gnb_classifier(x_docvecs, y_scores)
# saving the ml model to disk
scikit_ml_helper.persist_model_to_disk(ml_model,
self.ml_model_file_path)
log.info("Completed execution")
def __init__(self, source):
super(SemevalTaggedLineDocument, self).__init__(source)
self.semeval_articles = file_helper.get_articles_list(source)
self.counter = -1
def process(self):
log.info("Commencing execution")
with open(self.classification_sources_file_path) as source_cfg:
sources_dict = json.load(source_cfg)
x_documents = list()
y_scores = list()
log.info("Parsing documents")
for sentiment in sources_dict:
review_file_path = sources_dict[sentiment]
with open(review_file_path) as review_file:
for line in review_file:
x_document = line
y_score = sentiment
x_documents.append(x_document)
y_scores.append(y_score)
semeval_classified_articles_file = self.labeled_articles_file_path
semeval_classified_articles = file_helper.get_articles_list(
semeval_classified_articles_file)
y_true = list()
semeval_count = 0
for semeval_classified_article in semeval_classified_articles:
# article_text = semeval_classified_article['articleText']
article_text = semeval_classified_article['title']
if not article_text:
continue
semeval_count += 1
x_documents.append(article_text)
y_true.append(semeval_classified_article['label'])
log.info("Initiating training for document vectors")
vectorizer = TfidfVectorizer(sublinear_tf=True,
max_df=0.5,
stop_words='english')
x_vectors = vectorizer.fit_transform(x_documents)
labelled_docs = x_vectors[(-1 * semeval_count):]
log.info("Vectors have been trained")
log.info("Training the ML models")
ml_model_logreg = scikit_ml_helper.train_logistic_reg_classifier(
x_vectors[:(-1 * semeval_count)], y_scores)
ml_model_nb = scikit_ml_helper.train_gnb_classifier_dense(
x_vectors[:(-1 * semeval_count)], y_scores)
ml_model_svm_linear = scikit_ml_helper.train_svm_classifier(
x_vectors[:(-1 * semeval_count)], y_scores)
log.info("Saving the ML models to disk")
scikit_ml_helper.persist_model_to_disk(
ml_model_logreg, self.ml_model_file_path + ".tfidf.log_reg")
scikit_ml_helper.persist_model_to_disk(
ml_model_nb, self.ml_model_file_path + ".tfidf.nb")
scikit_ml_helper.persist_model_to_disk(
ml_model_svm_linear, self.ml_model_file_path + ".tfidf.svm_linear")
predictions_logreg = ml_model_logreg.predict(labelled_docs)
predictions_linearsvm = ml_model_svm_linear.predict(labelled_docs)
predictions_nb = ml_model_nb.predict(labelled_docs)
accuracy_logreg = \
sklearn.metrics.accuracy_score(y_true=y_true, y_pred=predictions_logreg, normalize=True, sample_weight=None)
accuracy_linearsvm = \
sklearn.metrics.accuracy_score(y_true=y_true, y_pred=predictions_linearsvm, normalize=True, sample_weight=None)
accuracy_nb = \
sklearn.metrics.accuracy_score(y_true=y_true, y_pred=predictions_nb, normalize=True, sample_weight=None)
log.info("accuracy_logreg: " + str(accuracy_logreg))
log.info("accuracy_linearsvm: " + str(accuracy_linearsvm))
log.info("accuracy_nb: " + str(accuracy_nb))
log.info("\ncm_logreg\n" + str(
scikit_ml_helper.get_confusion_matrix(y_true, predictions_logreg)))
log.info("\ncm_linearsvm\n" + str(
scikit_ml_helper.get_confusion_matrix(y_true,
predictions_linearsvm)))
log.info(
"\ncm_nb\n" +
str(scikit_ml_helper.get_confusion_matrix(y_true, predictions_nb)))
log.info("Completed execution")
nb_ml_model_file_path = \
"/home/v2john/Documents/amazon/models/books_ml_docvec.model.docvec.nb"
doc2vec_model = Doc2Vec.load(doc2vec_model_file_path)
logreg_model = scikit_ml_helper.get_model_from_disk(logreg_ml_model_file_path)
svm_model = scikit_ml_helper.get_model_from_disk(svm_linear_ml_model_file_path)
nb_model = scikit_ml_helper.get_model_from_disk(nb_ml_model_file_path)
semeval_classified_articles_file = \
"/home/v2john/Dropbox/Personal/Academic/Masters/UWaterloo/Academics/ResearchProject/semeval_task/" + \
"semeval-2017-task-5-subtask-2/semeval_combined_fulltext_classified.json"
semeval_classified_articles = file_helper.get_articles_list(semeval_classified_articles_file)
article_vectors = list()
count = 0
y_true = list()
for semeval_classified_article in semeval_classified_articles:
print 'working on article ' + str(count)
count += 1
article_text = semeval_classified_article['articleText']
# article_text = semeval_classified_article['title']
if not article_text:
continue