CN-121980086-A - Content recommendation method, device, computer equipment and storage medium

Abstract

The application relates to a content recommendation method, a content recommendation device, computer equipment and a storage medium. The method comprises the steps of clustering user browsing data of different user accounts in an acquisition period, dynamically dividing interest classification thresholds corresponding to different content types, pushing the interest content of different content types for each user account according to the interest classification thresholds determined dynamically, and not relying on fixed system score points or normal distribution assumptions, so that interference of hot content or extreme users on classification results can be reduced, objectivity of the interest classification thresholds is improved, and the problem that interest classification generalization is insufficient in some interest classification schemes is solved.

Inventors

• DU HONGGUANG
• XU XIAOLE
• ZHANG QUNFANG
• HE KAI
• ZHANG SHIJUN

Assignees

• 北京奇艺世纪科技有限公司

Dates

Publication Date

20260505

Application Date

20260205

Claims (10)

1. 1. A content recommendation method, the method comprising: Acquiring user browsing data of different user accounts in an acquisition period; Determining interest grading thresholds corresponding to different content types according to the clustering result of the user browsing data; determining interest grades of the user accounts on different content types according to the interest grading threshold; And pushing corresponding interested contents for each user account according to the interest level of each user account for different content types.
2. 2. The method according to claim 1, wherein determining the interest ranking threshold corresponding to different content types according to the clustering result of the user browsing data comprises: Determining interest confidence of the user account to the target content type according to the user browsing data; And determining an interest grading threshold corresponding to the target content type according to clustering results of interest confidence degrees of different user accounts on the target content type.
3. 3. The method of claim 2, wherein determining the confidence of interest of the user account in the target content type based on the user browsing data comprises: and determining the interest confidence of the user account to the target content type according to the weighted sum of different behavioral characteristics of the target content type in the user browsing data.
4. 4. The method according to claim 2, wherein the determining the interest ranking threshold corresponding to the target content type according to the clustering result of interest confidence of different user accounts for the target content type includes: Determining a plurality of clustering centers according to clustering results of interest confidence degrees of different user accounts on the target content types; Determining interest grading point positions according to the critical point positions among the clustering centers; And determining a plurality of interest grading thresholds corresponding to the target content type based on the interval ranges among different interest grading points.
5. 5. The method of claim 4, wherein determining the interest classification point based on the critical point locations between the cluster centers comprises: and determining the interest grading point positions according to the central point positions between any two adjacent clustering centers.
6. 6. The method of claim 5, wherein determining the interest classification point based on the center point between any adjacent two of the cluster centers comprises: determining the central point position between any two adjacent clustering centers as a grading point position to be selected; And carrying out noise reduction treatment on each candidate grading point to obtain an interest grading point corresponding to each candidate grading point.
7. 7. The method of claim 6, wherein the performing noise reduction on each candidate classification point location to obtain an interest classification point location corresponding to each candidate classification point location includes: acquiring a last grading point position generated in a last threshold calculation period; and carrying out exponential weighted moving average processing on the grading point positions to be selected according to the last grading point position and the smoothing coefficient to obtain the interest grading point positions corresponding to the grading point positions to be selected.
8. 8. A content recommendation device, the device comprising: The acquisition module is used for acquiring user browsing data of different user accounts in the acquisition period; the grading module is used for determining interest grading thresholds corresponding to different content types according to the clustering result of the user browsing data; The determining module is used for determining the interest level of each user account to different content types according to the interest level threshold; and the pushing module is used for pushing the corresponding interested contents for each user account according to the interest level of each user account for different content types.
9. 9. A computer device comprising a memory, a processor and a computer program stored on the memory and executable on the processor, characterized in that the processor implements the method of any of claims 1 to 7 when executing the computer program.
10. 10. A computer readable storage medium, on which a computer program is stored, characterized in that the computer program, when being executed by a processor, implements the method of any of claims 1 to 7.

Description

Content recommendation method, device, computer equipment and storage medium Technical Field The present application relates to the field of content recommendation, and in particular, to a content recommendation method, apparatus, computer device, and storage medium. Background In the field of content personalized recommendation, some personalized distribution strategies are usually required to be realized according to the interest intensity of a user, such as grading the interests of the user according to weight scores, and then performing personalized recommendation strategy adjustment according to the grading data of the interests. The general interest grading scheme has the problem of insufficient interest grading generalization, namely, the interest grading scheme depends on fixed statistical score points or normal distribution assumptions and cannot adapt to multi-channel data differences and dynamic fluctuation of popular contents. Disclosure of Invention The application provides a content recommendation method, a content recommendation device, computer equipment and a storage medium, which are used for solving the problem that the existing interest classification scheme has insufficient interest classification generalization. In a first aspect, the present application provides a content recommendation method, the method comprising: Acquiring user browsing data of different user accounts in an acquisition period; Determining interest grading thresholds corresponding to different content types according to the clustering result of the user browsing data; determining interest grades of the user accounts on different content types according to the interest grading threshold; And pushing corresponding interested contents for each user account according to the interest level of each user account for different content types. In a second aspect, the present application provides a content recommendation apparatus, the apparatus comprising: The acquisition module is used for acquiring user browsing data of different user accounts in the acquisition period; the grading module is used for determining interest grading thresholds corresponding to different content types according to the clustering result of the user browsing data; The determining module is used for determining the interest level of each user account to different content types according to the interest level threshold; and the pushing module is used for pushing the corresponding interested contents for each user account according to the interest level of each user account for different content types. In a third aspect, the present application provides a computer device comprising a memory, a processor and a computer program stored on the memory and executable on the processor, the processor implementing the content recommendation method described above when executing the computer program. In a fourth aspect, the present application also provides a computer storage medium storing computer-executable instructions for performing the content recommendation method described above. Compared with the prior art, the technical scheme provided by the embodiment of the application has the advantages that the method provided by the embodiment of the application obtains the user browsing data of different user accounts in the acquisition period, determines the interest classification threshold corresponding to different content types according to the clustering result of the user browsing data, determines the interest level of each user account to different content types according to the interest classification threshold, and pushes corresponding interest content for each user account according to the interest level of each user account to different content types. Based on the method, the user browsing data of different user accounts in the acquisition period are clustered, so that the interest classification threshold corresponding to different content types is dynamically divided, and further, the interest content of different content types is pushed to each user account according to the interest classification threshold determined dynamically, and the interference of hot content or extreme users on classification results can be reduced without depending on fixed system score points or normal distribution assumptions, the objectivity of the interest classification threshold is improved, and the problem that interest classification generalization is insufficient in some interest classification schemes is solved. Drawings The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments consistent with the invention and together with the description, serve to explain the principles of the invention. In order to more clearly illustrate the embodiments of the invention or the technical solutions of the prior art, the drawings which are used in the description of the embodiments or the prior art will be brief