def api():
input1 = request.args['variable']
data = preprocess(str(input1))
output = {}
output['date'] = datetime.datetime.now().strftime("%d/%m/%Y")
output['response'] = data
output['message'] = "Thanks For Using Text Process"
return jsonify(**output)
def train(self,text,label):
data = preprocess(text)
self.training_data['total_data_set']+=1
if label in self.training_data['labels']:
self.training_data['labels'][label]['docs']+=1
else :
self.training_data['labels'][label] = {}
self.training_data['labels'][label]['docs'] = 1
self.training_data['labels'][label]['keywords'] = {}
self.training_data['labels'][label]['total_keywords'] = 0
for word in data :
self.training_data['labels'][label]['total_keywords']+=1
if word in self.training_data['labels'][label]['keywords']:
self.training_data['labels'][label]['keywords'][word]+=1
else:
self.training_data['labels'][label]['keywords'][word]=1
self.save()
def classify(self,text):
if self.training_data['total_data_set'] ==0:
print 'Training data empty: Please train the system before executing classify'
sys.exit(0)
data = preprocess(text)
label_weights = {}
for label in self.training_data['labels']:
prior = math.log10(1+(self.training_data['labels'][label]['docs'])/(self.training_data['total_data_set']))
term_weight = 0
for word in data:
if word in self.training_data['labels'][label]['keywords']:
n = self.training_data['labels'][label]['keywords'][word]
else :
n = 0
term_weight+= math.log10((n+1)/(self.training_data['labels'][label]['docs']+ len(data)))
label_weights[label] = prior + term_weight
print label_weights
return max(label_weights.iteritems(), key=operator.itemgetter(1))[0]
