def GET(self, request):
web.header('Access-Control-Allow-Origin', '*')
if not web.svm:
return 'error'
text = web.input(_unicode=False)['text'].decode('utf-8', 'ignore')
account_id = int(web.input(_unicode=False)['id'].decode('utf-8', 'ignore'))
account_id = account_id or 1
print(u"Classifying text [{0}]".format(text))
features_list = [features.tweet_to_words(text)]
if features_list and len(features_list) > 0:
try:
labels = web.svm[account_id].classify_many(features_list)
if labels and len(labels) > 0:
return labels[0]
else:
return 'empty'
except Exception, e:
traceback.print_exc()
return 'error'
def GET(self, request):
web.header('Access-Control-Allow-Origin', '*')
if not web.svm:
return 'error'
text = web.input(_unicode=False)['text'].decode('utf-8', 'ignore')
account_id = int(
web.input(_unicode=False)['id'].decode('utf-8', 'ignore'))
account_id = account_id or 1
print(u"Classifying text [{0}]".format(text))
features_list = [features.tweet_to_words(text)]
if features_list and len(features_list) > 0:
try:
labels = web.svm[account_id].classify_many(features_list)
if labels and len(labels) > 0:
return labels[0]
else:
return 'empty'
except Exception, e:
traceback.print_exc()
return 'error'
if not w.lower() in stopwords.words('english')
)
word_features = list(all_words)[:500]
nltk.FreqDist.pprint(all_words, 500)
return word_features
print(decisions.categories())
documents = [(list(decisions.words(fileid)), category)
for category in decisions.categories()
for fileid in decisions.fileids(category)]
#random.shuffle(documents)
print (documents)
pos_features = [(features.tweet_to_words(d), c) for (d, c) in documents if c == 'pos']
neg_features = [(features.tweet_to_words(d), c) for (d, c) in documents if c == 'neg']
random.shuffle(pos_features)
random.shuffle(neg_features)
chosen_features_200 = pos_features[:100] + neg_features[:100]
random.shuffle(chosen_features_200)
featuresets = chosen_features_200
size = 200
train_set, test_set = featuresets[size / 2:], featuresets[:size / 2]
classifier = nltk.DecisionTreeClassifier.train(train_set)
print("Decision trees accuracy = [{0}]".format(
home = os.path.expanduser("~")
path = os.path.join(
home,
'nltk_data{s}corpora{s}tweets_publish_choice_{account}{s}{version}'.format(
account=account_id,
version=version,
s=os.sep,
))
print(decisions.categories())
documents = [(list(decisions.words(fileid)), category)
for category in decisions.categories()
for fileid in decisions.fileids(category)
if os.path.getsize(os.path.join(path, fileid)) > 0]
pos_features = [(features.tweet_to_words(d), c) for (d, c) in documents
if c == 'pos']
neg_features = [(features.tweet_to_words(d), c) for (d, c) in documents
if c == 'neg']
chosen_features_limit = pos_features[:limit / 2] + neg_features[:limit / 2]
random.shuffle(chosen_features_limit)
featuresets = chosen_features_limit
train_set, test_set = featuresets[limit / 2:], featuresets[:limit / 2]
svm = SklearnClassifier(LinearSVC())
svm.train(train_set)
path = os.path.normpath('../model/svm/account_{0}/{1}/'.format(
