Method for determining risk preference of user, information recommendation method, and apparatus

• US 10,783,457 B2
• Filed: 11/21/2019
• Issued: 09/22/2020
• Est. Priority Date: 05/26/2017
• Status: Active Grant

• Current Assignee: Advanced New Technologies Company Limited (Ant Group Co., Ltd.)
• Original Assignee: Alibaba Group Holding Ltd.
• Inventors: Yang, Fan, Fu, Xin
• Primary Examiner(s): Afshar, Kamran
• Assistant Examiner(s): Baldwin, Randall K.
• Application Number:
  US16/691,161
• Publication Number:
  US 20200090077A1
• Time in Patent Office: 306 Days
• CPC Class Codes: G06N 20/00 Machine learning
• CPC Class Codes: G06N 5/025 Extracting rules from data
• CPC Class Codes: G06Q 30/02 Marketing; Price estimation or determination; Fundraising
• CPC Class Codes: G06Q 30/0255 based on user history
• CPC Class Codes: G06Q 40/06 Asset management; Financial planning or analysis

Abstract

Methods, systems, and apparatus, including computer programs encoded on computer storage media, for determining risk preference are provided. One of the methods includes: obtaining user data generated during a risk-related transaction of a user; determining a characteristic value of the user under each variable in a plurality of variables according to the user data, the plurality of variables comprising at least one variable affecting a risk preference of the user; and inputting the characteristic value of the user under each variable into a risk preference model to determine an output of the risk preference model as a risk preference index indicating a level of the risk preference of the user.

Claims

1. A method, comprising:

training a risk preference model by;

selecting a plurality of sample users according to actual behavior of the plurality of sample users when facing a loss or when interacting with an experimental application for testing risk preferences, the plurality of sample users comprising a plurality of sample users of a high risk preference type and a plurality of sample users of a low risk preference type;

obtaining a characteristic value of each sample user in the plurality of sample users under each variable in a plurality of variables; and

training the risk preference model according to the obtained characteristic values and the risk preference type corresponding to each of the plurality of sample users, an input of the risk preference model being the characteristic value under the each variable in the plurality of variables, and an output of the risk preference model being a possibility that the user is classified into the high risk preference type, and

determining a risk preference index of a user by;

obtaining user data generated during a risk-related transaction of the user;

determining a characteristic value of the user under each variable in the plurality of variables according to the user data, the plurality of variables comprising at least one variable affecting a risk preference of the user;

inputting the characteristic value of the user under each variable in the plurality of variables into the risk preference model; and

outputting the risk preference index indicating a level of the risk preference of the user from the risk preference model.

2. The method of claim 1, wherein the selecting a plurality of sample users comprises:

based on a high risk preference rule and a low risk preference rule, determining users whose user data conforms to the high risk preference rule as sample users of the high risk preference type, and determining users whose user data conforms to the low risk preference rule as sample users of the low risk preference type; and

/or

based on an experimental application for testing risk preferences of users as well as a high risk preference rule and a low risk preference rule, determining users whose behaviors in the experimental application conform to the high risk preference rule as sample users of the high risk preference type, and determining users whose behaviors in the experimental application conform to the low risk preference rule as sample users of the low risk preference type.

3. The method of claim 1, before determining the characteristic value of the user under the each variable in the plurality of variables according to the user data, further comprising determining the each variable affecting the risk preference of the user, wherein determining the each variable affecting the risk preference of the user comprises:

selecting a plurality of sample users, the plurality of sample users comprising a plurality of sample users of a high risk preference type and a plurality of sample users of a low risk preference type;

for a variable to be verified, obtaining a characteristic value of each sample user in the plurality of sample users under the variable to be verified; and

determining that the variable to be verified is the variable affecting the risk preference of the user if the obtained characteristic values conform to a rule.

4. The method of claim 3, wherein the determining that the variable to be verified is the variable affecting the risk preference of the user if the obtained characteristic values conform to a rule comprises:

determining, based on the obtained characteristic values, a first characteristic value rule corresponding to the plurality of sample users of the high risk preference type and a second characteristic value rule corresponding to the plurality of sample users of the low risk preference type;

determining a difference between the first characteristic value rule and the second characteristic value rule; and

determining that the variable to be verified is the variable affecting the risk preference of the user if the difference conforms to a rule.

5. The method of claim 1, wherein the risk-related transaction comprises a transaction with a capital loss risk, and/or a transaction associated with a risky event.

6. The method of claim 1, further comprising:

recommending, according to the risk preference index of the user, financing product information matching the risk preference index of the user.

7. The method of claim 1, wherein:

the risk-related transaction comprises an economic life service transaction or a financial transaction.

8. A system, comprising one or more processors and one or more non-transitory computer-readable memories coupled to the one or more processors and configured with instructions executable by the one or more processors to cause the system to perform operations comprising:

training a risk preference model by;

selecting a plurality of sample users according to actual behavior of the plurality of sample users when facing a loss or when interacting with an experimental application for testing risk preferences, the plurality of sample users comprising a plurality of sample users of a high risk preference type and a plurality of sample users of a low risk preference type;

obtaining a characteristic value of each sample user in the plurality of sample users under each variable in a plurality of variables; and

training the risk preference model according to the obtained characteristic values and the risk preference type corresponding to each of the plurality of sample users, an input of the risk preference model being the characteristic value under the each variable in the plurality of variables, and an output of the risk preference model being a possibility that the user is classified into the high risk preference type, and

determining a risk preference index of a user by;

obtaining user data generated during a risk-related transaction of the user;

determining a characteristic value of the user under each variable in the plurality of variables according to the user data, the plurality of variables comprising at least one variable affecting a risk preference of the user;

inputting the characteristic value of the user under each variable in the plurality of variables into the risk preference model; and

outputting the risk preference index indicating a level of the risk preference of the user from the risk preference model.

9. The system of claim 8, wherein the selecting a plurality of sample users comprises:

based on a high risk preference rule and a low risk preference rule, determining users whose user data conforms to the high risk preference rule as sample users of the high risk preference type, and determining users whose user data conforms to the low risk preference rule as sample users of the low risk preference type; and

/or

based on an experimental application for testing risk preferences of users as well as a high risk preference rule and a low risk preference rule, determining users whose behaviors in the experimental application conform to the high risk preference rule as sample users of the high risk preference type, and determining users whose behaviors in the experimental application conform to the low risk preference rule as sample users of the low risk preference type.

10. The system of claim 8, wherein before determining the characteristic value of the user under the each variable in the plurality of variables according to the user data, the operations further comprise determining the each variable affecting the risk preference of the user, wherein determining the each variable affecting the risk preference of the user comprises:

selecting a plurality of sample users, the plurality of sample users comprising a plurality of sample users of a high risk preference type and a plurality of sample users of a low risk preference type;

for a variable to be verified, obtaining a characteristic value of each sample user in the plurality of sample users under the variable to be verified; and

determining that the variable to be verified is the variable affecting the risk preference of the user if the obtained characteristic values conform to a rule.

11. The system of claim 10, wherein the determining that the variable to be verified is the variable affecting the risk preference of the user if the obtained characteristic values conform to a rule comprises:

determining, based on the obtained characteristic values, a first characteristic value rule corresponding to the plurality of sample users of the high risk preference type and a second characteristic value rule corresponding to the plurality of sample users of the low risk preference type;

determining a difference between the first characteristic value rule and the second characteristic value rule; and

determining that the variable to be verified is the variable affecting the risk preference of the user if the difference conforms to a rule.

12. The system of claim 8, wherein the risk-related transaction comprises a transaction with a capital loss risk, and/or a transaction associated with a risky event.

13. The system of claim 8, further comprising:

recommending, according to the risk preference index of the user, financing product information matching the risk preference index of the user.

14. The system of claim 8, wherein:

the risk-related transaction comprises an economic life service transaction or a financial transaction.

15. A non-transitory computer-readable storage medium, the storage medium configured with instructions executable by one or more processors to cause the one or more processors to perform operations comprising:

training a risk preference model by;

selecting a plurality of sample users according to actual behavior of the users when facing a loss or when interacting with an experimental application for testing risk preferences, the plurality of sample users comprising a plurality of sample users of a high risk preference type and a plurality of sample users of a low risk preference type;

obtaining a characteristic value of each sample user in the plurality of sample users under each variable in a plurality of variables; and

training the risk preference model according to the obtained characteristic values and the risk preference type corresponding to each of the plurality of sample users, an input of the risk preference model being the characteristic value under the each variable in the plurality of variables, and an output of the risk preference model being a possibility that the user is classified into the high risk preference type, and

determining a risk preference index of a user by;

obtaining user data generated during a risk-related transaction of the user;

determining a characteristic value of the user under each variable in the plurality of variables according to the user data, the plurality of variables comprising at least one variable affecting a risk preference of the user;

inputting the characteristic value of the user under each variable in the plurality of variables into the risk preference model; and

outputting the risk preference index indicating a level of the risk preference of the user from the risk preference model.

16. The storage medium of claim 15, wherein the selecting a plurality of sample users comprises:

based on a high risk preference rule and a low risk preference rule, determining users whose user data conforms to the high risk preference rule as sample users of the high risk preference type, and determining users whose user data conforms to the low risk preference rule as sample users of the low risk preference type; and

/or

based on an experimental application for testing risk preferences of users as well as a high risk preference rule and a low risk preference rule, determining users whose behaviors in the experimental application conform to the high risk preference rule as sample users of the high risk preference type, and determining users whose behaviors in the experimental application conform to the low risk preference rule as sample users of the low risk preference type.

17. The storage medium of claim 15, wherein before determining the characteristic value of the user under the each variable in the plurality of variables according to the user data, the operations further comprise determining the each variable affecting the risk preference of the user, wherein determining the each variable affecting the risk preference of the user comprises:

selecting a plurality of sample users, the plurality of sample users comprising a plurality of sample users of a high risk preference type and a plurality of sample users of a low risk preference type;

for a variable to be verified, obtaining a characteristic value of each sample user in the plurality of sample users under the variable to be verified; and

determining that the variable to be verified is the variable affecting the risk preference of the user if the obtained characteristic values conform to a rule.

18. The storage medium of claim 15, further comprising:

recommending, according to the risk preference index of the user, financing product information matching the risk preference index of the user.