Granted Patent B2

US 12,205,000 · App. 17/139,849 · Granted Jan 21, 2025

Inventors: Dennis Fong (Hillsborough, CA); Kun Gao (Hillsborough, CA); George Ng (Hillsborough, CA); Ling Xiao (Hillsborough, CA)

Assignee: GGWP, Inc.

G06N20/00G06F3/14G06F16/2379G06N5/04G06Q30/0201H04L63/102

Abstract

Claims (68)

1. A system for cross-platform user profiling based on disparate datasets using machine learning models, the system comprising:

one or more processors and memory storing instructions that, when executed by the one or more processors, cause operations comprising:

retrieving a cross-platform user profile comprising a profile linked to an account, for a user, that is used across multiple video games on different gaming platforms;

retrieving in-game user interaction data for a video game and game-specific criteria data for the video game, wherein the game-specific criteria data comprises criteria for detecting toxic gameplay incidents performed by users in an in-game environment of the video game;

detecting, using a machine learning model, an incident of the user based on the in-game user interaction data and the game-specific criteria data for the video game, wherein the machine learning model is trained to detect known incidents in training data comprising labeled native game data;

determining, using an additional machine learning model, that native game data from the video game corresponds to profile-stored behavior data from the cross-platform user profile, wherein the additional machine learning model is trained to detect the user's behavior in the video game in training data comprising labeled native game data;

updating a gaming-toxicity-related status of the cross-platform user profile based on the detected incident and the determination that the native game data corresponds to the profile-stored behavior data; and

generating for presentation, in a user interface for the account, the gaming-toxicity-related status of the cross-platform user profile.

2. A method for cross-platform user profiling based on disparate datasets using machine learning models, the method comprising:

retrieving, using control circuitry, a cross-platform profile comprising a profile linked to an account, for a user, that is used across multiple video games on different gaming platforms;

retrieving, using the control circuitry, in-game user interaction data for a video game and game-specific criteria data for the video game, wherein the game-specific criteria data comprises criteria for detecting gameplay incidents performed by users in an in-game environment of the video game;

detecting, using a machine learning model, an incident of the user based on the in-game user interaction data and the game-specific criteria data for the video game, wherein the machine learning model is trained to detect known incidents in training data comprising labeled native game data;

determining, using an additional machine learning model, that first native game data from the video game corresponds to profile-stored behavior data from the cross-platform profile, wherein the additional machine learning model is trained to detect the user's behavior in the video game in training data comprising labeled native game data;

updating, using the control circuitry, a reputation-related status of the cross- platform profile based on the detected incident and the determination that the first native game data corresponds to the profile-stored behavior data; and

generating for presentation, in a user interface for the account, the reputation- related status of the cross-platform profile.

3. The method of claim 2 , further comprising:

determining, by a cross-platform profile provider different from a game provider of the video game, based on the reputation-related status of the cross-platform profile, one or more in-game features to be granted to or revoked from the account; and

automatically granting or revoking the one or more in-game features to or from the account.

4. The method of claim 2 , further comprising:

receiving a request to access the video game using a game-specific account different from the account; and

in response to the request to access the video game using the game-specific account, determining that first user-specific data of the game-specific account corresponds to second user-specific data of the cross-platform profile, and requiring the user to use the cross-platform profile to access the video game based on the determination that the first user-specific data corresponds to the second user-specific data.

5. The method of claim 2 , further comprising:

retrieving the first native game data from the video game indicating the user's behavior in the video game and the profile-stored behavior data from the cross-platform profile, wherein the profile-stored behavior data represents the user's prior behavior in the video game that occurred prior to occurrence of the user's behavior indicated by the first native game data; and

determining to update the reputation-related status of the cross-platform profile based on the incident in response to determining that the first_native game data corresponds to the profile-stored behavior data.

6. The method of claim 2 , wherein retrieving the in-game user interaction data and the game-specific criteria data comprises:

retrieving native game data as at least part of the in-game user interaction data and a telemetry data pattern, that corresponds to a predetermined user behavior in the video game, as at least part of the game-specific criteria data,

wherein detecting the incident of the user comprises comparing the telemetry data pattern to telemetry data extracted from the native game data.

7. The method of claim 2 , further comprising:

receiving native game data for the video game; and

retrieving the in-game user interaction data from the native game data for the video game.

8. The method of claim 2 , further comprising:

determining a frequency at which to pull native game data for the video game from a data source based on a data storage period for the data source of the native game data;

pulling the native game data from the data source at the frequency derived from the data storage period; and

extracting, from the native game data, avatar telemetry data indicating a plurality of movements of an avatar of the user within the video game,

wherein detecting the incident is further based on the avatar telemetry data.

9. The method of claim 2 , further comprising:

determining an incident sampling frequency based on known incidents in user- specific data of the cross-platform profile; and

extracting telemetry data from native game data of the video game, and sampling the telemetry data at the incident sampling frequency to detect the incident of the user.

10. The method of claim 2 , wherein the reputation-related status indicates a reputation score associated with the cross-platform profile.

11. One or more non-transitory, computer-readable media comprising instructions that, when executed by one or more processors, cause operations comprising:

retrieving a cross-platform profile comprising a profile linked to an account, for a user, that is used across multiple games on different gaming platforms;

retrieving in-game interaction data for a game and game-specific criteria data for the game, wherein the game-specific criteria data comprises criteria for detecting gameplay incidents in the game;

detecting, using a machine learning model, an incident of the user based on the in-game interaction data and the game-specific criteria data, wherein the machine learning model is trained to detect known incidents in training data comprising labeled native game data;

determining, using an additional machine learning model, that first native game data from the game corresponds to profile-stored behavior data from the cross-platform profile, wherein the additional machine learning model is trained to detect the user's behavior in the game in training data comprising labeled native game data;

updating a reputation-related status of the cross-platform profile based on the incident and the determination that the first native game data corresponds to the profile-stored behavior data; and

generating for presentation, in a user interface for the account, the reputation- related status of the cross-platform profile.

12. The one or more non-transitory, computer-readable media of claim 11 , the operations further comprising:

determining, by a cross-platform profile provider different from a game provider of the game, based on the reputation-related status of the cross-platform profile, one or more in-game features to be granted to or revoked from the account; and

automatically granting or revoking the one or more in-game features to or from the account.

13. The one or more non-transitory, computer-readable media of claim 11 , the operations further comprising:

in response to a request to access the game using a game-specific account different from the account, requiring the user to use the cross-platform profile to access the game based on a determination that first user-specific data of the game-specific account corresponds to second user-specific data of the cross-platform profile.

14. The one or more non-transitory, computer-readable media of claim 11 , the operations further comprising:

retrieving the first native game data from the game indicating the user's behavior in the game and the profile-stored behavior data from the cross-platform profile, wherein the profile-stored behavior data represents the user's prior behavior in the game that occurred prior to occurrence of the user's behavior indicated by the first native game data; and

determining to update the reputation-related status of the cross-platform profile based on the incident in response to determining that the first_native game data corresponds to the profile-stored behavior data.

15. The one or more non-transitory, computer-readable media of claim 11 , the operations further comprising:

retrieving native game data as at least part of the in-game interaction data and a telemetry data pattern, that corresponds to a predetermined user behavior in the game, as at least part of the game-specific criteria data,

wherein detecting the incident of the user comprises comparing the telemetry data pattern to telemetry data extracted from the native game data.

16. The one or more non-transitory, computer-readable media of claim 11 , the operations further comprising:

receiving native game data for the game; and

retrieving the in-game interaction data from the native game data for the game.

17. The one or more non-transitory, computer-readable media of claim 11 , the operations further comprising:

determining a frequency at which to pull native game data for the game from a data source based on a data storage period for the data source of the native game data;

pulling the native game data from the data source at the frequency derived from the data storage period; and

extracting, from the native game data, avatar telemetry data indicating a plurality of movements of an avatar of the user within the game,

wherein detecting the incident is further based on the avatar telemetry data.

18. The one or more non-transitory, computer-readable media of claim 11 , the operations further comprising:

determining an incident sampling frequency based on known incidents in user-specific data of the cross-platform profile; and

extracting telemetry data from native game data of the game, and sampling the telemetry data at the incident sampling frequency to detect the incident of the user.

Assignments (1)

Continuity (1)

References Cited (16)