def tokenize(self):
print "Tokenizing users.\n"
for user, docs in self.users.iteritems():
doc_lower = " ".join(docs).lower().replace("w/", "").replace("\n", "")
doc_final = anticontract.expand_contractions(doc_lower)
tokens = self.tokenizer.tokenize(doc_final)
clean_tokens = [token for token in tokens if token not in self.stop_en]
self.tokenized_docs[user] = clean_tokens
print "Done tokenizing users.\n"
def tokenize(self):
"""
Tokenize (extract unique tokens) all reviews given in self.reviews
"""
print 'Tokenizing reviews.\n'
for doc in self.reviews:
raw_doc = doc['text'].replace("w/", "")
raw_doc = raw_doc.replace("\n", "")
doc_lower = raw_doc.lower()
doc_final = anticontract.expand_contractions(doc_lower)
tokens = self.tokenizer.tokenize(doc_final)
clean_tokens = [token for token in tokens if token not in self.stop_en]
self.tokenized_docs[doc['review_id']] = {'tokens': clean_tokens, 'user': doc['user_id']}
print 'Done tokenizing reviews.\n'