Login| Sign Up| Help| Contact|

Patent Searching and Data


Title:
RISK RECOGNITION AND UPDATE OF RISK RECOGNITION SYSTEM
Document Type and Number:
WIPO Patent Application WO/2022/156701
Kind Code:
A1
Abstract:
Provided in the embodiments of the present description is a risk recognition method. The method comprises: acquiring a first event sample to be subjected to recognition; and inputting the first event sample into a risk recognition system for a plurality of risk domains. The risk recognition system comprises a first representation layer, a second representation layer and an output layer; the second representation layer comprises a shared representation sub-layer and a plurality of specific representation sub-layers corresponding to the plurality of risk domains; the output layer comprises a plurality of output sub-layers corresponding to the plurality of risk domains, wherein the first representation layer determines a first representation vector of the first event sample on the basis of an event feature of the first event sample; each representation sub-layer in the second representation layer individually determines a representation sub-vector of the first event sample on the basis of the first representation vector; and each output sub-layer in the output layer determines a risk prediction result on the basis of the representation sub-vector of the corresponding risk domain and a representation sub-vector determined by the shared representation sub-layer.

Inventors:
ZHOU JING (CN)
LV LE (CN)
FU XING (CN)
BAO PENGQING (CN)
FU XINYI (CN)
ZHANG CHANGHAO (CN)
WANG WEIQIANG (CN)
Application Number:
PCT/CN2022/072704
Publication Date:
July 28, 2022
Filing Date:
January 19, 2022
Export Citation:
Click for automatic bibliography generation   Help
Assignee:
ALIPAY HANGZHOU INF TECH CO LTD (CN)
International Classes:
G06Q10/00
Foreign References:
CN111291900A2020-06-16
CN111401914A2020-07-10
CN112785157A2021-05-11
CN111967807A2020-11-20
Attorney, Agent or Firm:
BEIJING BESTIPR INTELLECTUAL PROPERTY LAW CORPORATION (CN)
Download PDF: