CN-116204730-B - Content recommendation method and device

CN116204730BCN 116204730 BCN116204730 BCN 116204730BCN-116204730-B

Abstract

The embodiment of the application provides a content recommendation method and device, wherein the method comprises the steps of obtaining social relation data of a first user and a first content list issued by a second user with interaction behavior with the first user, obtaining a second content list of second content of which the second user has performed the first behavior, pulling all first content corresponding to the set number of second users in the user list from the first content list under the condition that the number of the second users in the user list exceeds a threshold value, recommending the content to the first user according to the pulled first content, pulling the first content corresponding to the second users in the user list from the first content list under the condition that the number of the second users in the user list does not exceed the threshold value, and pulling the second content corresponding to the second users in the user list from the second content list, and recommending the content to the first user according to the pulled first content and second content.

Inventors

ZHANG SHU
ZHANG JUNQI
XU DABO

Assignees

微梦创科网络科技（中国）有限公司

Dates

Publication Date: 20260512
Application Date: 20230221

Claims (9)

1. A content recommendation method, comprising: receiving a content request initiated by a first user; Responding to the content request, acquiring social relation data of a first user and a first content list issued by a second user with interactive behaviors with the first user, wherein the social relation data comprises a user list of the second user with interactive behaviors with the first user; Acquiring a second content list of second content of which the second user performs the first action, wherein the interaction action comprises the first action; Pulling all first contents corresponding to the set number of second users in the user list from the first content list under the condition that the number of the second users in the user list exceeds a threshold value; Pulling first contents of which the number corresponds to all second users in the user list from the first content list and pulling second contents of which the number corresponds to all second users in the user list from the second content list under the condition that the number of the second users in the user list does not exceed the threshold value; The recommending the content to the first user according to the pulled first content comprises the following steps: sorting the pulled first content; determining the first content with the forefront ordering as a target first content; determining first content corresponding to a second user who publishes the target first content; And recommending the target first content and target second content to the first user, wherein the target second content is at least one piece of content except the target first content in the first content corresponding to the second user who publishes the target first content.
2. The content recommendation method of claim 1 wherein said ordering said pulled first content comprises: Inputting the first content corresponding to each second user, the reading time of the first content and the type of the interaction behavior of the first content into a CTR model for prediction to obtain the ranking score of the first content corresponding to each second user; in a refresh period, determining the exposure times of a second user corresponding to the first content in the refresh period; Determining a final sorting score of the first content corresponding to each second user according to the sorting score and the exposure times; and sorting the first content according to the final sorting score.
3. The content recommendation method according to claim 1, wherein before the content recommendation is made to the first user based on the pulled first content, the method further comprises: acquiring theme tag information of the first content; Aggregating the first content of the same subject tag information; The recommending the content to the first user according to the pulled first content comprises the following steps: determining first content with the forefront ordering as target first content, and determining theme tag information of the target first content; And recommending the target first content and the first content with the same theme label information as the target first content to the first user.
4. The content recommendation method according to claim 1, wherein the pulling from the first content list a number of first content corresponding to each of all second users in the user list comprises: Determining a relationship score of each second user having interaction behavior with the first user according to social relationship data of the first user, wherein the relationship score is used for describing the intimacy between the second user and the first user, and the relationship score is in direct proportion to the intimacy; and pulling a first content with the quantity corresponding to the relation score of the second user from the first content list for any second user, wherein the quantity of the pulled first content is proportional to the relation score of the second user.
5. The content recommendation method of claim 4 wherein determining a relationship score for each second user that has interactive activity with the first user based on the social relationship data of the first user comprises: Determining the type of the interaction behavior between the second user and the first user and the weight of the interaction behavior corresponding to the type according to the social relation data; Determining the number of times the second user has occurred the type of interaction behavior and a time decay factor for the type of interaction behavior within a predetermined period of time; And determining the relation score according to the weight, the times and the time attenuation coefficient.
6. The content recommendation method according to claim 1, wherein the recommending content to the first user based on the pulled first content and second content comprises: sorting the pulled first content and second content; determining the content with the top ranking from the first content and the second content which are pulled as target content; Determining at least one first content and at least one second content corresponding to a second user who publishes the target content, wherein the at least one first content or the at least one second content comprises the target content; the at least one first piece of content and the at least one second piece of content are recommended to the first user.
7. A content recommendation device, comprising: The receiving module is used for receiving a content request initiated by a first user; The acquisition module is used for responding to the content request, acquiring social relation data of a first user and a first content list issued by a second user with interactive behaviors with the first user, wherein the social relation data comprises a user list of the second user with interactive behaviors with the first user; The acquisition module is further configured to acquire a second content list of second content that is subjected to the first behavior by the second user, where the interactive behavior includes the first behavior; The pulling module is used for pulling all first contents corresponding to the set number of second users in the user list from the first content list under the condition that the number of the second users in the user list exceeds a threshold value; The recommendation module is used for recommending the content to the first user according to the pulled first content; The pulling module is further configured to, when the number of second users in the user list does not exceed the threshold value, pull, from the first content list, first content corresponding to each of all second users in the user list, and pull, from the second content list, second content corresponding to each of all second users in the user list; The recommendation module is also used for recommending the content to the first user according to the pulled first content and the pulled second content; the recommendation module is specifically configured to sort the pulled first content, determine a first content with the highest sorting as a target first content, determine a first content corresponding to a second user who publishes the target first content, and recommend the target first content and the target second content to the first user, where the target second content is at least one piece of content except the target first content in the first content corresponding to the second user who publishes the target first content.
8. An electronic device, comprising: A processor; a memory for storing the processor-executable instructions; wherein the processor is configured to execute the instructions to implement the content recommendation method of any one of claims 1 to 6.
9. A computer readable storage medium, characterized in that instructions in the storage medium, when executed by a processor of an electronic device, enable the electronic device to perform the content recommendation method according to any one of claims 1 to 6.

Description

Content recommendation method and device Technical Field The invention relates to the technical field of internet, in particular to a content recommendation method and device. Background With the continuous development of internet technology, information flow is an important way for users to passively acquire information, and occupies most of fragmentation time of users. The information stream products distribute the content products in a waterfall stream mode, common information stream products such as shopping applications are commodity and news applications corresponding to the distributed content products, and news information corresponding to the distributed content products. In some scenarios, information flow is a form of data that continuously enhances content to users, which is a resource aggregator made up of multiple content sources that can be actively subscribed to and push data content to users. The content distribution based on social relations is commonly adopted at present by constructing a virtual social relation network of users and carrying out split propagation and diffusion on content generated by common people along a relation link of one stage and the other stage. However, in this way, when the social relationship links of the users are more, the content material pulling stage is filtered by adopting a TOPN cut-off mode due to the limitation of machine resources, and the cut-off mode is usually mainly based on time or random modes, so that the content released by part of the users is filtered, the content loss of the users is caused, the diversity of the content is reduced, and the accuracy and reliability of content recommendation are lower. Disclosure of Invention The embodiment of the application aims to provide a content recommendation method and device, which are used for solving the problems of content loss of a user and lower accuracy and reliability of content recommendation. In order to solve the technical problems, the embodiment of the application is realized as follows: In a first aspect, an embodiment of the present application provides a content recommendation method, which includes receiving a content request initiated by a first user, obtaining social relationship data of the first user and a first content list published by a second user having an interaction behavior with the first user in response to the content request, where the social relationship data includes a user list of the second user having the interaction behavior with the first user, obtaining a second content list of second content of the second user subjected to the first behavior, where the interaction behavior includes the first behavior, pulling all first contents corresponding to a set number of second users in the first content list from the first content list when the number of second users in the user list exceeds a threshold, recommending content to the first user according to the pulled first contents, pulling all first contents corresponding to the first users in the first content list from the first content list when the number of second users in the user list does not exceed the threshold, and pulling all first contents corresponding to the first users in the first content list from the first content list according to the pulled first contents. In a second aspect, an embodiment of the present application provides a content recommendation device, where the content recommendation device is configured to receive a content request initiated by a first user, obtain, in response to the content request, social relationship data of the first user and a first content list published by a second user having an interaction behavior with the first user, where the social relationship data includes a user list of the second user having the interaction behavior with the first user, obtain, in addition, a second content list of second content in which the second user has performed the first behavior, where the interaction behavior includes the first behavior, pull, in addition, obtain, in addition, all first content corresponding to a set number of second users in the user list from the first content list when a number of second users in the user list exceeds a threshold, recommend, in addition, pull, in addition, all first content corresponding to the set number of second users in the user list from the first content list, and pull, in addition, all content corresponding to the first user list from the first content list, and all content corresponding to the first user list from the first user list. In a third aspect, an embodiment of the present application provides an electronic device, including a processor, a communication interface, a memory, and a communication bus, where the processor, the communication interface, and the memory complete communication with each other through the bus, the memory is configured to store a computer program, and the processor is configured to execute the program stored