def tencent_classify(content):
"""
given a passage content, return its first level class
:param content:
:return:
"""
action = 'TextClassify'
module = 'wenzhi'
config = {
'Region': 'gz',
'secretId': 'AKIDi1ohL0zeMXuUaxEwGHMgd2aXeQYqnYJS',
'secretKey': 'oUEND16w54I8i6uragtks71EFK70hg8a',
'method': 'get'
}
content = content.replace('\n', '')
content = content.replace(' ', '')
params = {
u"content": content.replace('\n', '')
}
for i in xrange(3):
try:
qcloud_service = QcloudApi(module, config)
result = qcloud_service.call(action, params)
result_obj = json.loads(result)
if result_obj['code'] != 0:
continue
else:
return result_obj
except Exception, e:
return {'code': 1, 'msg': 'unknown error, detail=%s' % e}
def wenzhi_analysis(content):
"""
使用腾讯文智文本分类接口进行分析，得出
:param content:
:return:
"""
module = 'wenzhi'
action = 'TextClassify'
config = {
'Region': 'sz',
'secretId': 'AKIDi1ohL0zeMXuUaxEwGHMgd2aXeQYqnYJS',
'secretKey': 'oUEND16w54I8i6uragtks71EFK70hg8a',
'method': 'post'
}
refined_content = remove_illegal_characters(content)
params = {
'content': refined_content, 'query_encode': 0
}
try:
service = QcloudApi(module, config)
# req_url = service.generateUrl(action, params)
# print req_url
# result_str = requests.get(req_url)
result_str = service.call(action, params)
# print result_str
k = 0.001 # 平滑系数
result = json.loads(result_str)
prob_sum = 0
if result['code'] == 0:
for item in result['classes']:
prob_sum += item['conf'] + k
for item in result['classes']:
item['conf'] = (item['conf'] + k) / prob_sum
return result
except Exception, e:
print e
