Method and device for pushing information

• US 11,074,623 B2
• Filed: 12/19/2019
• Issued: 07/27/2021
• Est. Priority Date: 06/06/2016
• Status: Active Grant

• Current Assignee: Advanced New Technologies Company Limited (Ant Group Co., Ltd.)
• Original Assignee: Advanced New Technologies Company Limited (Ant Group Co., Ltd.)
• Inventors: Song, Nian
• Primary Examiner(s): Summers, Kiersten V
• Application Number:
  US16/721,857
• Publication Number:
  US 20200134674A1
• Time in Patent Office: 586 Days
• Field of Search: None
• CPC Class Codes: G06Q 30/02 Marketing; Price estimation or determination; Fundraising
• CPC Class Codes: G06Q 30/0242 Determining effectiveness of advertisements
• CPC Class Codes: G06Q 30/0255 based on user history
• CPC Class Codes: G06Q 30/0269 based on user profile or attribute
• CPC Class Codes: G06Q 50/01 Social networking
• CPC Class Codes: H04L 67/55 Push-based network services
• CPC Class Codes: H04L 67/63 Routing a service request depending on the request content or context

Abstract

The present disclosure discloses a method and device for pushing information to a target user. One example method includes identifying a plurality of users that meet a predetermined condition; selecting a target user from the identified users to be provided with information, where in the target user is selected based on a relationship strength and an influence of the identified users; and transmitting the information to the determined target user over a network, and relates to the field of information technologies.

Claims

1. A computer-implemented method, comprising:

identifying, by one or more processing devices, a plurality of users, wherein identifying the plurality of users comprises;

obtaining asset data in one or more asset accounts of each sample user of a set of sample users, and filling in missing data in the asset data according to one or more predetermined rules,

identifying anomalous data in the asset data, and replacing the anomalous data with other data according to one or more additional predetermined rules,

based on the asset data in the one or more asset accounts of each sample user, obtaining a total asset amount of each sample user, wherein attribute data of each sample user comprises the total asset amount of the sample user,

converting attribute data of the set of sample users into a sparse matrix format,

training multiple scoring models using the converted attribute data of the set of sample users,

measuring scoring effects of each scoring model based on different combinations of data of the attribute data of the set of sample users,

selecting, as a predetermined scoring model, a scoring model of the multiple scoring models having scoring effects with a particular characteristic,

scoring attribute data, the attribute data comprising asset data, of potential users using the predetermined scoring model, to obtain a score for each potential user, and

selecting, as the identified plurality of users, potential users having a score that satisfies a predetermined score condition;

obtaining initial relation community labels of the identified plurality of users;

determining an updated relation community label for each user of the identified plurality of users, comprising;

determining that a relationship strength between a first user of the identified plurality of users and a second user of the identified plurality of users is larger than a threshold, and

responsive to determining that the relationship strength between the first user and the second user is larger than the threshold, assigning matching updated relation community labels to the first user and the second user;

dividing the identified plurality of users into a plurality of relation communities, wherein users of the identified plurality of users having matching updated relation community labels are assigned to a particular relation community, wherein each user of the identified plurality of users is assigned to a single relation community based on the updated relation community label of the user, and wherein at least one of the plurality of relation communities corresponds to membership in an online social network community;

selecting a plurality of target users from identified users of each particular relation community to be provided with information, wherein the plurality of target users are selected based on respective influences of the plurality of target users compared to respective influences of other identified users of the particular relation community, wherein an influence of each user comprises a strength with which the user affects other users; and

transmitting the information to the selected plurality of target users over a network.

2. The computer-implemented method of claim 1, wherein the updated relation community labels are final updated relation community labels, and wherein determining the final updated relation community label for each user of the identified plurality of users further comprises:

updating, in a predetermined number of iterations, comparisons of relationship strength between the identified plurality of users, wherein the predetermined number of iterations is at least two, and wherein each iteration after a first iteration of the predetermined number of iterations comprises;

obtaining, for the first user, a previous relation community label based on a previous iteration to the iteration;

determining whether a relationship strength between the first user and another user of the identified plurality of users is larger than the threshold; and

in response to determining that the relationship strength between the first user and the other user is larger than the threshold, assigning matching updated relation community labels to the first user and the other user, comprising one of

changing the previous relation community label of the first user to match a previous relation community label of the other user, or

changing the previous relation community label of the other user to match the previous relation community label of the first user,

wherein an updated relationship community label of the first user based on a final iteration of the predetermined number of iterations is the final updated relation community label of the first user.

3. The computer-implemented method of claim 1, further comprising:

determining an influence level for each user in the identified plurality of users based on a predetermined influence level table.

4. The computer-implemented method of claim 3, wherein the predetermined influence level table comprises influence threshold intervals that correspond to different influence levels, the different influence levels divided into high, medium, sub-medium, and low.

5. The computer-implemented method of claim 1, wherein selecting the plurality of target users from each particular relation community to be provided with the information comprises:

selecting, as the plurality of target users, a matching proportion of the identified users of each relation community.

6. A non-transitory computer-readable storage medium coupled to one or more computers and configured with instructions executable by the one or more computers to perform operations comprising:

identifying, by the one or more computers, a plurality of users, wherein identifying the plurality of users comprises;

obtaining asset data in one or more asset accounts of each sample user of a set of sample users, and filling in missing data in the asset data according to one or more predetermined rules,

identifying anomalous data in the asset data, and replacing the anomalous data with other data according to one or more additional predetermined rules,

based on the asset data in the one or more asset accounts of each sample user, obtaining a total asset amount of each sample user, wherein attribute data of each sample user comprises the total asset amount of the sample user,

converting attribute data of the set of sample users into a sparse matrix format,

training multiple scoring models using the converted attribute data of the set of sample users,

measuring scoring effects of each scoring model based on different combinations of data of the attribute data of the set of sample users,

selecting, as a predetermined scoring model, a scoring model of the multiple scoring models having scoring effects with a particular characteristic,

scoring attribute data, the attribute data comprising asset data, of potential users using the predetermined scoring model, to obtain a score for each potential user, and

selecting, as the identified plurality of users, potential users having a score that satisfies a predetermined score condition;

obtaining initial relation community labels of the identified plurality of users;

determining an updated relation community label for each user of the identified plurality of users, comprising;

determining that a relationship strength between a first user of the identified plurality of users and a second user of the identified plurality of users is larger than a threshold, and

responsive to determining that the relationship strength between the first user and the second user is larger than the threshold, assigning matching updated relation community labels to the first user and the second user;

dividing the identified plurality of users into a plurality of relation communities, users of the identified plurality of users having matching updated relation community labels are assigned to a particular relation community, wherein each user of the identified plurality of users is assigned to a single relation community based on the updated relation community label of the user, and wherein at least one of the plurality of relation communities corresponds to membership in an online social network community;

selecting a plurality of target users from identified users of each particular relation community to be provided with information, wherein the plurality of target users are selected based on respective influences of the plurality of target users compared to respective influences of other identified users of the particular relation community, wherein an influence of each user comprises a strength with which the user affects other users; and

transmitting the information to the selected plurality of target users over a network.

7. The non-transitory computer-readable storage medium of claim 6, wherein the updated relation community labels are final updated relation community labels, and wherein determining the final updated relation community label for each user of the identified plurality of users further comprises:

updating, in a predetermined number of iterations, comparisons of relationship strength between the identified plurality of users, wherein the predetermined number of iterations is at least two, and wherein each iteration after a first iteration of the predetermined number of iterations comprises;

obtaining, for the first user, a previous relation community label based on a previous iteration to the iteration;

determining whether a relationship strength between the first user and another user of the identified plurality of users is larger than the threshold; and

in response to determining that the relationship strength between the first user and the other user is larger than the threshold, assigning matching updated relation community labels to the first user and the other user, comprising one of

changing the previous relation community label of the first user to match a previous relation community label of the other user, or

changing the previous relation community label of the other user to match the previous relation community label of the first user,

wherein an updated relationship community label of the first user based on a final iteration of the predetermined number of iterations is the final updated relation community label of the first user.

8. The non-transitory computer-readable storage medium of claim 6, configured with instructions executable by the one or more computers to:

determining an influence level for each user in the identified plurality of users based on a predetermined influence level table.

9. The non-transitory computer-readable storage medium of claim 8, wherein the predetermined influence level table comprises influence threshold intervals that correspond to different influence levels, the different influence levels divided into high, medium, sub-medium, and low.

10. The non-transitory computer-readable storage medium of claim 6, wherein selecting the plurality of target users from each particular relation community to be provided with the information comprises:

selecting, as the plurality of target users, a matching proportion of the identified users of each relation community.

11. A system, comprising:

one or more computers; and

one or more computer-readable memories coupled to the one or more computers and configured with instructions executable by the one or more computers to perform operations comprising;

identifying, by the one or more computers, a plurality of users, wherein identifying the plurality of users comprises;

obtaining asset data in one or more asset accounts of each sample user of a set of sample users, and filling in missing data in the asset data according to one or more predetermined rules,

identifying anomalous data in the asset data, and replacing the anomalous data with other data according to one or more additional predetermined rules,

based on the asset data in the one or more asset accounts of each sample user, obtaining a total asset amount of each sample user, wherein attribute data of each sample user comprises the total asset amount of the sample user,

converting attribute data of the set of sample users into a sparse matrix format,

training multiple scoring models using the converted attribute data of the set of sample users,

measuring scoring effects of each scoring model based on different combinations of data of the attribute data of the set of sample users,

selecting, as a predetermined scoring model, a scoring model of the multiple scoring models having scoring effects with a particular characteristic,

scoring attribute data, the attribute data comprising asset data, of potential users using the predetermined scoring model, to obtain a score for each potential user, and

selecting, as the identified plurality of users, potential users having a score that satisfies a predetermined score condition;

obtaining initial relation community labels of the identified plurality of users;

determining an updated relation community label for each user of the identified plurality of users, comprising;

determining that a relationship strength between a first user of the identified plurality of users and a second user of the identified plurality of users is larger than a threshold, and

responsive to determining that the relationship strength between the first user and the second user is larger than the threshold, assigning matching updated relation community labels to the first user and the second user;

dividing the identified plurality of users into a plurality of relation communities, wherein users of the identified plurality of users having matching updated relation community labels are assigned to a particular relation community, wherein each user of the identified plurality of users is assigned to a single relation community based on the updated relation community label of the user, and wherein at least one of the plurality of relation communities corresponds to membership in an online social network community;

selecting a plurality of target users from identified users of each particular relation community to be provided with information, wherein the plurality of target users are selected based on respective influences of the plurality of target users compared to respective influences of other identified users of the particular relation community, wherein an influence of each user comprises a strength with which the user affects other users; and

transmitting the information to the selected plurality of target users over a network.

12. The system of claim 11, wherein the updated relation community labels are final updated relation community labels, and wherein determining the final updated relation community label for each user of the identified plurality of users further comprises:

updating, in a predetermined number of iterations, comparisons of relationship strength between the identified plurality of users, wherein the predetermined number of iterations is at least two, and wherein each iteration after a first iteration of the predetermined number of iterations comprises;

obtaining, for the first user, a previous relation community label based on a previous iteration to the iteration;

determining whether a relationship strength between the first user and another user of the identified plurality of users is larger than the threshold; and

in response to determining that the relationship strength between the first user and the other user is larger than the threshold, assigning matching updated relation community labels to the first user and the other user, comprising one of

changing the previous relation community label of the first user to match a previous relation community label of the other user, or

changing the previous relation community label of the other user to match the previous relation community label of the first user,

wherein an updated relationship community label of the first user based on a final iteration of the predetermined number of iterations is the final updated relation community label of the first user.

13. The system of claim 11, wherein the operations comprise:

determining an influence level for each user in the identified plurality of users based on a predetermined influence level table.

14. The system of claim 13, wherein the predetermined influence level table comprises influence threshold intervals that correspond to different influence levels, the different influence levels divided into high, medium, sub-medium, and low.

15. The system of claim 11, wherein selecting the plurality of target users from each particular relation community to be provided with the information comprises:

selecting, as the plurality of target users, a matching proportion of the identified users of each relation community.