CN-121996842-A - Optimization method and device for recommended object, electronic equipment and program product

CN121996842ACN 121996842 ACN121996842 ACN 121996842ACN-121996842-A

Abstract

The application discloses a recommendation object optimization method, a recommendation object optimization device, electronic equipment and a recommendation object optimization program product, and relates to the technical field of content recommendation, wherein the method comprises the steps of obtaining theme characteristics of a recommendation set, wherein the recommendation set is used for representing multi-mode recommendation information; the method comprises the steps of obtaining a matching relation between a theme feature and an interest portrait of a first recommended object, wherein the interest portrait is determined based on first interaction data of the first recommended object, screening the first recommended object based on the matching relation to obtain a second recommended object, recommending a recommended collection to the second recommended object, obtaining second interaction data of the second recommended object aiming at the recommended collection, and optimizing the second recommended object by utilizing the second interaction data. By implementing the technical scheme of the application, the target users can be screened and recommended based on the accurate matching of the recommendation set theme and the user interests, and then the target user group is continuously and dynamically optimized by utilizing the feedback data, so that the recommendation accuracy and the interactive conversion rate are effectively improved.

Inventors

XUE JIN
LI JUNNING
WANG JIAJIE
LI YONG

Assignees

杭州网易云音乐科技有限公司

Dates

Publication Date: 20260508
Application Date: 20251231

Claims (10)

1. A method of optimizing a recommended object, the method comprising: obtaining theme characteristics of a recommendation set, wherein the recommendation set is used for representing multi-mode recommendation information; Acquiring a matching relationship between the subject feature and an interest portrait of a first recommended object, wherein the interest portrait is determined based on first interaction data of the first recommended object; Screening the first recommended object based on the matching relation to obtain a second recommended object; recommending the recommendation set to the second recommendation object, and acquiring second interaction data of the second recommendation object aiming at the recommendation set; and optimizing the second recommended object by utilizing the second interaction data.
2. The method of claim 1, wherein optimizing the second recommended object using the second interaction data comprises: Processing the second interaction data to determine an evaluation index of the recommended set; And optimizing the second recommended object based on the evaluation index.
3. The method of claim 2, wherein the set of recommendations comprises at least one recommendation, wherein the processing the second interaction data to determine an evaluation index for the set of recommendations after being recommended comprises: determining target consumption completion degrees of the recommended items based on the content consumption records in the second interaction data; Determining a target weighting coefficient corresponding to each target consumption completion degree based on a mapping relation between the consumption completion degrees and the weighting coefficients; Determining a consumption time length weighting score corresponding to the recommendation set based on the superposition result of each target weighting coefficient; Determining a preference expression rate corresponding to the recommendation set based on the preference expression operation quantity in the second interaction data, and determining a focus conversion rate corresponding to the recommendation set based on the focus operation quantity in the second interaction data; the assessment index is determined based on the consumption time weighted score, the preferred expression rate, and the conversion rate of interest.
4. The method according to claim 1, wherein the method further comprises: Analyzing the second interaction data and determining an operation behavior sequence corresponding to each second recommended object; Counting the occurrence frequency of each operation behavior sequence; And determining the operation behavior sequence with the occurrence frequency exceeding a preset frequency threshold and the evaluation index exceeding a preset index threshold as a target behavior sequence.
5. The method according to claim 4, wherein the method further comprises: Analyzing the evaluation index and the target behavior sequence to generate an optimization suggestion aiming at the recommendation set; and responding to the adjustment operation executed on the recommendation set according to the optimization suggestion, and obtaining a modified recommendation set.
6. The method of claim 1, wherein the obtaining the matching relationship between the subject feature and the interest image of the first recommended object comprises: Extracting behavior characteristic data under a content consumption dimension, an interaction operation dimension and a content theme preference dimension from the first interaction data; Processing the behavior characteristic data of the content consumption dimension to obtain content preference values of the first recommended object for each content type; processing the behavior characteristic data of the interaction operation dimension to obtain interaction behavior intensity values of the first recommended object on various interaction types; Processing the behavior characteristic data of the content theme preference dimension to obtain theme preference degrees of the first recommended object on each content theme; and constructing an interest portrait of the first recommended object based on the content preference value, the interaction behavior intensity value and the theme preference, and establishing a matching relation between the theme characteristics and the interest portrait.
7. The method of claim 6, wherein the screening the first recommended object based on the matching relationship to obtain a second recommended object comprises: Matching the target content theme characterized by the theme characteristics with the theme preference degree in the interest portrait to obtain a matching result; If the matching result represents that the first recommended object is successfully matched, judging whether the interaction behavior intensity value corresponding to the first recommended object exceeds a preset intensity threshold value or not, and judging whether the content preference value of any content type exceeds the preset preference threshold value or not, so as to obtain a judging result; And if the judging result indicates that the interactive behavior intensity value exceeds the preset intensity threshold and the content preference value of any one content type exceeds the preset preference threshold, determining the first recommended object as a second recommended object.
8. An optimization apparatus for a recommended object, the apparatus comprising: The system comprises a first acquisition module, a second acquisition module and a third acquisition module, wherein the first acquisition module is used for acquiring theme characteristics of a recommendation set, and the recommendation set is used for representing multi-modal recommendation information; the second acquisition module is used for acquiring the matching relation between the theme characteristics and the interest portraits of the first recommended objects, wherein the interest portraits are determined based on the first interaction data of the first recommended objects; The screening module is used for screening the first recommended object based on the matching relation to obtain a second recommended object; The recommendation module is used for recommending the recommendation set to the second recommendation object and acquiring second interaction data of the second recommendation object aiming at the recommendation set; And the optimizing module is used for optimizing the second recommended object by utilizing the second interaction data.
9. An electronic device, comprising: A memory and a processor, the memory and the processor being communicatively connected to each other, the memory having stored therein computer instructions, the processor executing the computer instructions to perform the method of optimizing the recommended object according to any of claims 1 to 7.
10. A computer program product comprising computer instructions for causing a computer to perform the method of optimizing a recommended object according to any one of claims 1 to 7.

Description

Optimization method and device for recommended object, electronic equipment and program product Technical Field The present application relates to the field of content recommendation technologies, and in particular, to a method and apparatus for optimizing a recommendation object, an electronic device, and a program product. Background With the rapid development of digital content consumption, various platforms push integrated content to users in the form of recommendation aggregation and the like, and become an important means for improving user viscosity and activity. However, when pushing a recommendation set to a large number of users, the platform generally recommends the same recommendation set to all users without any personalized distinction, so that the recommendation effect is poor. Disclosure of Invention In view of the above, the present application provides a method, apparatus, electronic device and program product for optimizing a recommendation object, so as to solve the problem of poor recommendation effect of a recommendation set. The application provides a recommendation object optimization method, which comprises the steps of obtaining theme characteristics of a recommendation set, wherein the recommendation set is used for representing multi-mode recommendation information, obtaining a matching relation between the theme characteristics and an interest portrait of a first recommendation object, the interest portrait is determined based on first interaction data of the first recommendation object, screening the first recommendation object based on the matching relation to obtain a second recommendation object, recommending the recommendation set to the second recommendation object, obtaining second interaction data of the second recommendation object aiming at the recommendation set, and optimizing the second recommendation object by utilizing the second interaction data. In an alternative embodiment, optimizing the second recommended object by using the second interaction data includes processing the second interaction data, determining an evaluation index after the recommendation set is recommended, and optimizing the second recommended object based on the evaluation index. In an alternative implementation mode, the recommendation set comprises at least one recommendation item, the second interaction data is processed to determine an evaluation index after the recommendation set is recommended, the evaluation index comprises the steps of determining target consumption completion degrees of all recommendation items based on content consumption records in the second interaction data, determining target weighting coefficients corresponding to all target consumption completion degrees based on mapping relations between the consumption completion degrees and the weighting coefficients, determining consumption duration weighting scores corresponding to the recommendation set based on superposition results of all target weighting coefficients, determining preference expression rates corresponding to the recommendation set based on the number of preference expression operations in the second interaction data, determining attention conversion rates corresponding to the recommendation set based on the number of attention operations in the second interaction data, and determining the evaluation index based on the consumption duration weighting scores, the preference expression rates and the attention conversion rates. In an alternative implementation mode, the second interaction data are analyzed, operation behavior sequences corresponding to the second recommended objects are determined, the occurrence frequency of the operation behavior sequences is counted, the operation behavior sequences with the occurrence frequency exceeding a preset frequency threshold and the evaluation index exceeding the preset index threshold are determined to be target behavior sequences. In an alternative embodiment, the evaluation index and the target behavior sequence are analyzed to generate an optimization suggestion for the recommendation set, and the modified recommendation set is obtained in response to an adjustment operation performed on the recommendation set according to the optimization suggestion. In an alternative embodiment, the method comprises the steps of extracting content consumption dimension, interaction operation dimension and behavior feature data under the content theme preference dimension from first interaction data, processing the behavior feature data of the content consumption dimension to obtain content preference values of the first recommendation object for each content type, processing the behavior feature data of the interaction operation dimension to obtain interaction behavior intensity values of the first recommendation object for a plurality of interaction types, processing the behavior feature data of the content theme preference dimension to obtain theme preference degrees of