def __init__(self):
Controller.__init__(self)
texts = ''
for path in constants.PROCESSED_SPEECHES_PATHS:
with open(path, 'r') as corpus:
texts += re.sub(pattern='\n+', repl='\n\n', string=corpus.read().lower()).strip() + '\n\n'
self.mcc_speech_generator = MccSpeechGenerator().preprocess(texts)
def __init__(self):
Controller.__init__(self)
self._qr = QueryResponder()
with open(constants.PROCESSED_QNA_PATH) as f:
qna_list = pickle.load(f)
with open(constants.PROCESSED_QUOTES_PATH) as f:
text = f.read().lower().decode('utf-8').strip()
with open(constants.PROCESSED_FB_POSTS_PATH) as f:
text = text + '\n\n' + f.read().lower().decode('utf-8').strip()
for path in constants.PROCESSED_SPEECHES_PATHS:
with open(path, 'r') as corpus:
text += re.sub(
pattern='\n+', repl='\n\n', string=corpus.read().lower(
)).decode('utf-8').strip() + '\n\n'
tokenizer = nltk.data.load('tokenizers/punkt/english.pickle')
sentences = tokenizer.tokenize(text)
if not os.path.exists(constants.CHAT_INFERSENT_MODEL_PATH):
self._engine = self._qr.preprocess(qna_list, sentences)
def __init__(self):
Controller.__init__(self)
self._api = tweeter_utils.TweeterUtils()
os.chdir(DEEP_BB_PREFIX_PATH)
with open(constants.PROCESSED_TWEETS_PATH, 'r') as corpus:
tweets_text = re.sub(pattern='\n+',
repl='\n',
string=corpus.read().lower())
with open(constants.PROCESSED_QUOTES_PATH, 'r') as corpus:
quotes_text = re.sub(pattern='\n+',
repl='\n',
string=corpus.read().lower())
with open(constants.PROCESSED_FB_POSTS_PATH, 'r') as corpus:
fb_text = re.sub(pattern='.',
repl='.\n',
string=re.sub(pattern='\n+',
repl='\n',
string=corpus.read().lower()))
self.mcc_tweets_generator = MccTweetGenerator().preprocess(
[tweets_text, fb_text, quotes_text], weights=[0.4, 2, 1])
def __init__(self, server):
Controller.__init__(self, server)