Granted Patent B2

US 12,423,368 · App. 18/441,389 · Granted Sep 23, 2025

Inventors: Jieming Zhu (Shenzhen, CN); Zhou Zhao (Hangzhou, CN); Shengyu Zhang (Hangzhou, CN); Xiuqiang He (Shenzhen, CN); Li Qian (Shenzhen, CN)

Assignee: Huawei Technologies Co., Ltd.

G06F16/9535G06V10/40

Abstract

Claims (75)

1. A method, comprising:

obtaining a plurality of images, wherein each image comprises one candidate interface and one type of candidate content presented by using the candidate interface, and each image comprises a plurality of regions;

obtaining image feature data of each image, wherein the image feature data of each image comprises a plurality of local eigenvectors that correspond to the plurality of regions respectively;

predicting, by using a prediction model, a degree of preference of a target user for each image based on user feature data of the target user and the image feature data, wherein input for the prediction model is determined based on the user feature data and the image feature data; and

selecting, based on the degree of preference, at least one of a candidate interface or candidate content from candidate interfaces and candidate content that are comprised in the plurality of images, to perform recommendation.

2. The method according to claim 1 , wherein;

the predicting, by using a prediction model, a degree of preference of a target user for each image based on user feature data of the target user and the image feature data comprises:

for each image, obtaining N word vectors based on candidate content in the image, wherein each word vector represents one word in the candidate content, and N is a positive integer;

for each word vector, calculating, by using a model of an attention mechanism, an attention weight of each of the plurality of local eigenvectors based on the word vector and the plurality of local eigenvectors, wherein the attention weight indicates a degree to which the target user pays attention to a region represented by a respective local eigenvector when the target user reads a word represented by each word vector;

performing fusion on each word vector and the plurality of local eigenvectors based on the attention weight of each of the plurality of local eigenvectors to obtain a first fusion eigenvector, wherein one first fusion eigenvector is correspondingly obtained for each word vector; and

predicting, by using the prediction model, the degree of preference of the target user for each image based on a user eigenvector and N first fusion eigenvectors corresponding to the N word vectors, wherein input for the prediction model is determined based on the user eigenvector and the N first fusion eigenvectors, and the user eigenvector represents the user feature data of the target user.

3. The method according to claim 2 , wherein the predicting, by using the prediction model, the degree of preference of the target user for each image based on the user eigenvector and N first fusion eigenvectors corresponding to the N word vectors comprises:

for each image, processing, by using a model of a self-attention mechanism, the N first fusion eigenvectors corresponding to the N word vectors to obtain N semantic-enhanced eigenvectors, wherein each first fusion eigenvector corresponds to one semantic-enhanced eigenvector; and

predicting, by using the prediction model, the degree of preference of the target user for each image based on the user eigenvector and the N semantic-enhanced eigenvectors, wherein input for the prediction model is determined based on the user eigenvector and the N semantic-enhanced eigenvectors.

4. The method according to claim 3 , wherein the predicting, by using the prediction model, the degree of preference of the target user for each image based on the user eigenvector and the N semantic-enhanced eigenvectors comprises:

for each image, performing fusion on the N semantic-enhanced eigenvectors by using a model of an additive attention mechanism to obtain a second fusion eigenvector; and

predicting, by using the prediction model, the degree of preference of the target user for each image based on the user eigenvector and the second fusion eigenvector, wherein input for the prediction model is determined based on the user eigenvector and the second fusion eigenvector.

5. The method according to claim 1 , wherein the image feature data of each image comprises a global eigenvector, and the global eigenvector represents the image.

6. The method according to claim 5 , wherein the predicting, by using a prediction model, a degree of preference of a target user for each image based on user feature data of the target user and the image feature data comprises:

for each image, obtaining a content eigenvector based on the candidate content in the image, wherein the content eigenvector represents the candidate content;

determining a weight of the content eigenvector and a weight of the global eigenvector based on the content eigenvector and the global eigenvector;

performing fusion on the content eigenvector and the global eigenvector based on the weight of the content eigenvector and the weight of the global eigenvector to obtain a third fusion eigenvector; and

predicting, by using the prediction model, the degree of preference of the target user for each image based on a user eigenvector and the third fusion eigenvector, wherein input for the prediction model is determined based on the user eigenvector and a third fusion vector, and the user eigenvector represents the user feature data of the target user.

7. The method according to claim 1 , wherein the selecting, based on the degree of preference, at least one of a candidate interface or candidate content from candidate interfaces and candidate content that are comprised in the plurality of images, to perform recommendation comprises:

selecting, based on the degree of preference, a type of candidate content from the candidate content comprised in the plurality of images as target candidate content; and

selecting, based on the degree of preference, a candidate interface from candidate interfaces of images that comprise the target candidate content as a target candidate interface, to recommend the target candidate content by using the target candidate interface.

8. The method according to claim 7 , wherein after the selecting, based on the degree of preference, a candidate interface from candidate interfaces of images that comprise the target candidate content as a target candidate interface, the method further comprises:

sending the target candidate content and metadata of the target candidate interface to a terminal device, wherein the terminal device displays the target candidate interface based on the metadata, and recommends the target candidate content to the target user by using the target candidate interface.

9. A method, comprising:

obtaining a plurality of sample images, wherein each sample image comprises one sample candidate interface and one type of sample candidate content presented by using the sample candidate interface, and each sample image comprises a plurality of regions;

obtaining image feature data of each sample image, wherein the image feature data of each sample image comprises a plurality of local eigenvectors that correspond to the plurality of regions respectively;

predicting, by using a prediction model, a degree of preference of a sample user for each sample image based on user feature data of the sample user and the image feature data, wherein input for the prediction model is determined based on the user feature data and the image feature data; and

adjusting the prediction model based on the degree of preference and historical click-through data of the sample user for the sample candidate content.

10. The method according to claim 9 , wherein;

the predicting, by using a prediction model, a degree of preference of a sample user for each sample image based on user feature data of the sample user and the image feature data comprises:

for each sample image, obtaining N word vectors based on the sample candidate content in each sample image, wherein each word vector represents one word in the sample candidate content, and N is a positive integer;

for each word vector, calculating, by using a model of an attention mechanism, an attention weight of each of the plurality of local eigenvectors based on the word vector and the plurality of local eigenvectors, wherein the attention weight indicates a degree to which the sample user pays attention to a region represented by a respective local eigenvector when the sample user reads a word represented by each word vector;

performing fusion on each word vector and the plurality of local eigenvectors based on the attention weight of each of the plurality of local eigenvectors to obtain a first fusion eigenvector, wherein one first fusion eigenvector is correspondingly obtained for each word vector; and

predicting, by using the prediction model, the degree of preference of the sample user for each sample image based on user eigenvector and N first fusion eigenvectors corresponding to the N word vectors, wherein input for the prediction model is determined based on the user eigenvector and the N first fusion eigenvectors, and the user eigenvector represents the user feature data of the sample user.

11. The method according to claim 10 , wherein the predicting, by using the prediction model, the degree of preference of the sample user for each sample image based on the user eigenvector and N first fusion eigenvectors corresponding to the N word vectors comprises:

for each sample image, processing, by using a model of a self-attention mechanism, the N first fusion eigenvectors corresponding to the N word vectors to obtain N semantic-enhanced eigenvectors, wherein each first fusion eigenvector corresponds to one semantic-enhanced eigenvector; and

predicting, by using the prediction model, the degree of preference of the sample user for each sample image based on the user eigenvector and the N semantic-enhanced eigenvectors, wherein input for the prediction model is determined based on the user eigenvector and the N semantic-enhanced eigenvectors.

12. The method according to claim 11 , wherein the predicting, by using the prediction model, the degree of preference of the sample user for each sample image based on the user eigenvector and the N semantic-enhanced eigenvectors comprises:

for each sample image, performing fusion on the N semantic-enhanced eigenvectors by using a model of an additive attention mechanism to obtain a second fusion eigenvector; and

predicting, by using the prediction model, the degree of preference of the sample user for each sample image based on the user eigenvector and the second fusion eigenvector, wherein input for the prediction model is determined based on the user eigenvector and the second fusion eigenvector.

13. The method according to claim 9 , wherein the image feature data of each sample image comprises a global eigenvector, and the global eigenvector represents the sample image.

14. The method according to claim 13 , wherein the predicting, by using a prediction model, a degree of preference of a sample user for each sample image based on user feature data of the sample user and the image feature data comprises:

for each sample image, obtaining a content eigenvector based on the sample candidate content in each sample image, wherein the content eigenvector represents the sample candidate content;

determining a weight of the content eigenvector and a weight of the global eigenvector based on the content eigenvector and the global eigenvector;

performing fusion on the content eigenvector and the global eigenvector based on the weight of the content eigenvector and the weight of the global eigenvector to obtain a third fusion eigenvector; and

predicting, by using the prediction model, the degree of preference of the sample user for each sample image based on a user eigenvector and the third fusion eigenvector, wherein input for the prediction model is determined based on the user eigenvector and a third fusion vector, and the user eigenvector represents the user feature data of the sample user.

15. A computer device, comprising one or more memories and one or more processors, wherein the one or more memories are coupled to the one or more processors and store computer-readable instructions for execution by the one or more processor to:

obtain a plurality of images, wherein each image comprises one candidate interface and one type of candidate content presented by using the candidate interface, and each image comprises a plurality of regions;

obtain image feature data of each image, wherein the image feature data of each image comprises a plurality of local eigenvectors that correspond to the plurality of regions respectively;

predict, by using a prediction model, a degree of preference of a target user for each image based on user feature data of the target user and the image feature data, wherein input for the prediction model is determined based on the user feature data and the image feature data; and

select, based on the degree of preference, at least one of a candidate interface or candidate content from candidate interfaces and candidate content that are comprised in the plurality of images, to perform recommendation.

16. A device, comprising one or more memories and one or more processors, wherein the one or more memories are coupled to the one or more processors and store computer-readable instructions for execution by the one or more processor to:

obtain a plurality of sample images, wherein each sample image comprises one sample candidate interface and one type of sample candidate content presented by using the sample candidate interface, and each sample image comprises a plurality of regions;

obtain image feature data of each sample image, wherein the image feature data of each sample image comprises a plurality of local eigenvectors that correspond to the plurality of regions respectively;

predict, by using a prediction model, a degree of preference of a sample user for each sample image based on user feature data of the sample user and the image feature data, wherein input for the prediction model is determined based on the user feature data and the image feature data; and

adjust the prediction model based on the degree of preference and historical click-through data of the sample user for the sample candidate content.

17. A non-transitory computer-readable storage medium, wherein the non-transitory computer-readable storage medium stores computer-readable instructions for execution by at least one processor to:

obtain a plurality of images, wherein each image comprises one candidate interface and one type of candidate content presented by using the candidate interface, and each image comprises a plurality of regions;

obtain image feature data of each image, wherein the image feature data of each image comprises a plurality of local eigenvectors that correspond to the plurality of regions respectively;

predict, by using a prediction model, a degree of preference of a target user for each image based on user feature data of the target user and the image feature data, wherein input for the prediction model is determined based on the user feature data and the image feature data; and

select, based on the degree of preference, at least one of a candidate interface or candidate content from candidate interfaces and candidate content that are comprised in the plurality of images, to perform recommendation.

18. A recommendation system, comprising a terminal device and a server, wherein the server comprises one or more first processors and one or more first memories coupled to the one or more first processors, and the terminal device comprises one or more second processors and one or more second memories coupled to the one or more second processors; and wherein:

the one or more first memories store programming instructions for execution by the one or more first processors to:

obtain a plurality of images, wherein each image comprises one candidate interface and one type of candidate content presented by using the candidate interface, and each image comprises a plurality of regions;

obtain image feature data of each image, wherein the image feature data of each image comprises a plurality of local eigenvectors that correspond to the plurality of regions respectively;

predict, by using a prediction model, a degree of preference of a target user for each image based on user feature data of the target user and the image feature data, wherein input for the prediction model is determined based on the user feature data and the image feature data; and

send, based on the degree of preference, target candidate content and metadata of a target candidate interface to a terminal device, wherein the terminal device displays the target candidate interface based on the metadata, and recommends the target candidate content to the target user by using the target candidate interface; and

the one or more second memories store programming instructions for execution by the one or more second processors to:

receive target candidate content and metadata of a target candidate interface from the server; and

display the target candidate interface based on the metadata, and recommend the target candidate content to a target user by using the target candidate interface.

Assignments (1)

Priority Claims (1)

Continuity (2)

References Cited (31)