User behavior recognition method, user equipment, and behavior recognition server

Patent metadata

Field	Value
Publication number	US-10409841-B2
Application number	US-201615358410-A
Country	US
Kind code	B2
Filing date	Nov 22, 2016
Priority date	May 22, 2014
Publication date	Sep 10, 2019
Grant date	Sep 10, 2019

Abstract

A user behavior recognition method, a user equipment, a behavior recognition server, and a behavior recognition system are presented, where the method includes acquiring, by a first user equipment, statistical distribution information of a target parameter corresponding to a target user behavior, where the target parameter includes at least one parameter in a behavior recognition model of the target user behavior, and the statistical distribution information of the target parameter is determined according to values of the target parameters in behavior recognition models of the target user behavior that are respectively corresponding to multiple other user equipment; and creating and saving, according to the statistical distribution information, a behavior recognition model of the target user behavior, to recognize the target user behavior.

First claim

What is claimed is:

1. A user behavior recognition method, comprising:
sending, by a first user equipment, a data download request to a behavior recognition server, wherein the data download request comprises an identifier of a target user behavior; and
receiving, by the first user equipment, the statistical distribution information of a target parameter according to the identifier of the target user behavior from the behavior recognition server, wherein the target parameter comprises at least one parameter in a behavior recognition model of the target user behavior, wherein the statistical distribution information of the target parameter is determined according to values of the target parameter in a plurality of behavior recognition models of the target user behavior that respectively correspond to a plurality of other user equipment; and
creating and saving, by the first user equipment according to the statistical distribution information of the target parameter, a behavior recognition model of the target user behavior that corresponds to the first user equipment to recognize the target user behavior.

2. The method according to claim 1 , wherein the statistical distribution information of the target parameter comprises at least one a probability distribution curve of the target parameter, an expected value of the target parameter, or a value of the target parameter that has a largest occurrence probability.

3. The method according to claim 1 , wherein the behavior recognition model is a naive Bayes classifier, wherein a first feature with continuous feature values in the naive Bayes classifier is a first target parameter, wherein a second feature with discrete feature values in the naive Bayes classifier is a second target parameter, and wherein the statistical distribution information of the target parameter comprises at least one of an expected value of a normal distribution parameter that the first feature has in each category of the target user behavior, or a statistical distribution curve of the second feature for each category of the target user behavior.

4. The method according to claim 1 , further comprising sending, by the first user equipment, the behavior recognition model of the target user behavior to a behavior recognition server, such that the behavior recognition server updates the statistical distribution information of the target parameter according to a value of the target parameter in the behavior recognition model.

5. A user behavior recognition method, comprising:
receiving, by a behavior recognition server, a plurality of behavior recognition models of a target user behavior from a plurality of other user equipment;
determining, by the behavior recognition server according to the behavior recognition models of the target user behavior, a plurality of values of a target parameter that respectively correspond to the other user equipment, wherein the target parameter comprises at least one parameter in a behavior recognition model of the target user behavior;
determining, by the behavior recognition server, statistical distribution information of the target parameter according to the values of the target parameter in the behavior recognition models that respectively correspond to the other user equipment;
receiving, by the behavior recognition server, a data download request from a first user equipment, wherein the data download request comprises an identifier of the target user behavior;
determining, by the behavior recognition server, the statistical distribution information of the target parameter corresponding to the target user behavior according to the identifier of the target user behavior; and
sending, by the behavior recognition server, the statistical distribution information of the target parameter to the first user equipment to create the behavior recognition model of the target user behavior according to the statistical distribution information of the target parameter.

6. The method according to claim 5 , wherein the statistical distribution information of the target parameter comprises at least one of a probability distribution curve of the target parameter, an expected value of the target parameter, or a value of the target parameter that has a largest occurrence probability.

7. The method according to claim 5 , wherein the behavior recognition model of the target user behavior is a naive Bayes classifier, wherein a first feature with continuous feature values in the naive Bayes classifier is a first target parameter, wherein a second feature with discrete feature values in the naive Bayes classifier is a second target parameter, and wherein the statistical distribution information of the target parameter comprises at least one of an expected value of a normal distribution parameter that the first feature has in each category of the target user behavior, or a statistical distribution curve of the second feature for each category of the target user behavior.

8. The method according to claim 5 , wherein the behavior recognition server has a basic knowledge base and a plurality of coordinated knowledge bases, wherein the basic knowledge base is used to store statistical distribution information of a target parameter corresponding to at least one user behavior, wherein the statistical distribution information is commonly used by a plurality of user groups, and wherein each coordinated knowledge base in the coordinated knowledge bases is used to store statistical distribution information of the target parameter corresponding to the at least one user behavior, wherein the statistical distribution information is exclusively used by one user group in the user groups, wherein the at least one user behavior comprises the target user behavior, and wherein determining, by the behavior recognition server according to the identifier of the target user behavior, the statistical distribution information of the target parameter corresponding to the target user behavior comprises:
determining, by the behavior recognition server, a user group of the first user equipment;
determining, by the behavior recognition server from the basic knowledge base, the statistical distribution information of the target parameter corresponding to the target user behavior in response to the first user equipment not belonging to any user group in the multiple user groups; and
determining, by the behavior recognition server from a coordinated knowledge base corresponding to a first user group, the statistical distribution information of the target parameter corresponding to the target user behavior in response to the first user equipment belonging to the first user group.

9. The method according to claim 5 , further comprising:
receiving, by the behavior recognition server, the behavior recognition model of the target user behavior according to the statistical distribution information of the target parameter from the first user equipment; and
updating, by the behavior recognition server, the statistical distribution information of the target parameter according to the behavior recognition model.

10. The method according to claim 9 , wherein updating the statistical distribution information of the target parameter according to the behavior recognition model comprises:
determining, by the behavior recognition server according to the behavior recognition model received from the first user equipment, a value of the target parameter that corresponds to the first user equipment;
collecting, by the behavior recognition server, mathematical statistics on the value of the target parameters corresponding to the first user equipment according to the values of the target parameter that respectively correspond to the other user equ

Assignees

• Huawei Tech Co Ltd

Inventors

• Yang Qiang
• Deng Ke
• Leung Wing Ki

Classifications

• G06F3/011 Primary

  Arrangements for interaction with the human body, e.g. for user immersion in virtual reality (blind teaching G09B21/00) · CPC title

• G06F18/24155

  Bayesian classification · CPC title

• G06F16/2462

  Approximate or statistical queries · CPC title

• G06F16/285 Primary

  Clustering or classification · CPC title

• G06K9/6278

  Physics · mapped topic