CN116186398A - Template recommendation method, device, equipment and medium - Google Patents

Template recommendation method, device, equipment and medium Download PDF

Info

Publication number
CN116186398A
CN116186398A CN202310108211.6A CN202310108211A CN116186398A CN 116186398 A CN116186398 A CN 116186398A CN 202310108211 A CN202310108211 A CN 202310108211A CN 116186398 A CN116186398 A CN 116186398A
Authority
CN
China
Prior art keywords
template
sample
target object
target
information
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Pending
Application number
CN202310108211.6A
Other languages
Chinese (zh)
Inventor
昝晓飞
冯俊
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Beijing Dajia Internet Information Technology Co Ltd
Original Assignee
Beijing Dajia Internet Information Technology Co Ltd
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Beijing Dajia Internet Information Technology Co Ltd filed Critical Beijing Dajia Internet Information Technology Co Ltd
Priority to CN202310108211.6A priority Critical patent/CN116186398A/en
Publication of CN116186398A publication Critical patent/CN116186398A/en
Pending legal-status Critical Current

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
    • G06F16/90—Details of database functions independent of the retrieved data types
    • G06F16/95—Retrieval from the web
    • G06F16/953—Querying, e.g. by the use of web search engines
    • G06F16/9535—Search customisation based on user profiles and personalisation
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N20/00—Machine learning

Landscapes

  • Engineering & Computer Science (AREA)
  • Theoretical Computer Science (AREA)
  • Databases & Information Systems (AREA)
  • Data Mining & Analysis (AREA)
  • Physics & Mathematics (AREA)
  • General Engineering & Computer Science (AREA)
  • General Physics & Mathematics (AREA)
  • Software Systems (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Evolutionary Computation (AREA)
  • Medical Informatics (AREA)
  • Artificial Intelligence (AREA)
  • Computing Systems (AREA)
  • Mathematical Physics (AREA)
  • Management, Administration, Business Operations System, And Electronic Commerce (AREA)
  • Information Retrieval, Db Structures And Fs Structures Therefor (AREA)

Abstract

The disclosure provides a template recommendation method, a device, equipment and a medium, and relates to the field of computer surgery, wherein the template recommendation method comprises the following steps: object feature information of a target object is obtained, and terminal performance information corresponding to the target object is obtained; determining a recommended value of each candidate template with respect to the target object according to the object characteristic information, the terminal performance information and the template characteristic data of the plurality of candidate templates; determining a target template associated with the target object from a plurality of candidate templates according to the recommended value of each candidate template on the target object and a predetermined screening condition, and recommending the target template to the terminal equipment of the target object; the matching degree of the template recommended for the user, the user preference and the equipment performance is improved.

Description

Template recommendation method, device, equipment and medium
Technical Field
The disclosure relates to the technical field of computers, and in particular relates to a template recommendation method, device, equipment and medium.
Background
With the continuous perfection of the social application, in the process of releasing social content by the user, the social application can provide a social content making template for the user to select a favorite template to release social content.
Typically, to facilitate the user's posting of social content, a social application may make templates for the user to recommend social content for the user to select; however, in the current template recommendation logic, the factors considered in the template recommendation process are single, so that the template recommendation result is generally not in line with the actual performance of the terminal equipment, and social content release experience of the user is affected.
Disclosure of Invention
The disclosure provides a method, a device, equipment and a medium for recommending templates, which are used for at least solving the problem that the template recommending result in the related technology does not accord with the actual state of terminal equipment. The technical scheme of the present disclosure is as follows:
according to a first aspect of an embodiment of the present disclosure, there is provided a template recommendation method, including:
object feature information of a target object is obtained, and terminal performance information corresponding to the target object is obtained;
determining a recommended value of each candidate template with respect to the target object according to the object characteristic information, the terminal performance information and the template characteristic data of the candidate templates, wherein the recommended value is used for representing the priority of recommending the candidate templates for the target object;
and determining a target template associated with the target object from the plurality of candidate templates according to the recommended value of each candidate template about the target object and a predetermined screening condition, and recommending the target template to the terminal equipment of the target object.
Optionally, the determining, according to the object feature information, the terminal capability information and the template feature data of the plurality of candidate templates, a recommended value of each candidate template with respect to the target object includes:
determining a feature data set to be processed associated with each alternative template, wherein the feature data set to be processed comprises template feature data of the alternative templates, object feature information and terminal performance information;
and respectively inputting the feature data set to be processed associated with each alternative template into a pre-trained template recommendation model to obtain a recommendation value of each alternative template on the target object.
Optionally, the method further includes a training method of the template recommendation model, and the training method of the template recommendation model includes:
acquiring sample object characteristic information, sample terminal performance information of a sample object and sample template characteristic data of a sample alternative template used by the sample object;
and carrying out iterative training on the template recommendation model to be trained based on the sample object characteristic information, the sample terminal performance information and the sample template characteristic data of the sample alternative template to obtain the trained template recommendation model.
Optionally, before performing iterative training on the template recommendation model to be trained based on the sample object feature information, the sample terminal performance information and the sample template feature data of the sample alternative template to obtain the trained template recommendation model, the method further includes:
determining a sample recommendation value for each sample candidate template with respect to the sample object;
the performing iterative training on the template recommendation model to be trained based on the sample object feature information, the sample terminal performance information and the sample template feature data of the sample candidate template to obtain a trained template recommendation model comprises the following steps:
inputting the sample object characteristic information, the sample terminal performance information and sample template characteristic data of a target sample alternative template into the template recommendation model to be trained to obtain a prediction recommendation value of the sample object about the target sample alternative template;
determining a loss function value according to the sample recommended value and the predicted recommended value of the sample object relative to the target sample candidate template and a loss function;
and if the loss function value is smaller than or equal to a preset loss function threshold, determining that the template recommendation model converges, and obtaining the trained template recommendation model.
Optionally, the acquiring object feature information of the target object includes:
acquiring target historical behavior information corresponding to the target object according to the target object, wherein the target historical behavior information comprises data generated by the target object on the using behavior of the alternative template in a historical period;
determining first data and second data corresponding to the target object according to the target historical behavior information, wherein the first data are used for representing the sensitivity degree of the object to the template characteristics, and the second data are used for representing the sensitivity degree of the object to the template display effect;
and determining the first data and/or the second data corresponding to the target object as object characteristic information of the target object.
Optionally, before determining the first data and/or the second data corresponding to the target object as object feature information of the target object, the method further includes:
acquiring object attribute information corresponding to the target object;
the determining the first data and/or the second data corresponding to the target object as object feature information of the target object includes:
And determining at least one of the first data and the second data corresponding to the target object and the object attribute information as object feature information of the target object.
Optionally, the determining, according to the recommended value of each candidate template with respect to the target object and the predetermined screening condition, a target template associated with the target object from the plurality of candidate templates includes:
and selecting a preset number of alternative templates from the plurality of alternative templates according to the sequence of the recommended value from large to small, and obtaining a plurality of target templates associated with the target object.
According to a second aspect of the embodiments of the present disclosure, there is provided a template recommendation apparatus, including:
the terminal performance information acquisition module is configured to acquire object characteristic information of a target object and terminal performance information corresponding to the target object;
a first determining module configured to determine a recommendation value of each candidate template with respect to the target object according to the object feature information, the terminal performance information, and template feature data of a plurality of candidate templates, the recommendation value being used to characterize a priority of recommending the candidate templates for the target object;
And a second determining module configured to determine a target template associated with the target object among the plurality of candidate templates according to a recommended value of each candidate template with respect to the target object and a predetermined screening condition, and recommend the target template to a terminal device of the target object.
Optionally, the first determining module is configured to:
determining a feature data set to be processed associated with each alternative template, wherein the feature data set to be processed comprises template feature data of the alternative templates, object feature information and terminal performance information;
and respectively inputting the feature data set to be processed associated with each alternative template into a pre-trained template recommendation model to obtain a recommendation value of each alternative template on the target object.
Optionally, the apparatus further comprises a model training module configured to:
acquiring sample object characteristic information, sample terminal performance information of a sample object and sample template characteristic data of a sample alternative template used by the sample object;
and carrying out iterative training on the template recommendation model to be trained based on the sample object characteristic information, the sample terminal performance information and the sample template characteristic data of the sample alternative template to obtain the trained template recommendation model.
Optionally, the apparatus further includes a third determining module configured to:
determining a sample recommendation value for each sample candidate template with respect to the sample object;
the performing iterative training on the template recommendation model to be trained based on the sample object feature information, the sample terminal performance information and the sample template feature data of the sample candidate template to obtain a trained template recommendation model comprises the following steps:
inputting the sample object characteristic information, the sample terminal performance information and sample template characteristic data of a target sample alternative template into the template recommendation model to be trained to obtain a prediction recommendation value of the sample object about the target sample alternative template;
determining a loss function value according to the sample recommended value and the predicted recommended value of the sample object relative to the target sample candidate template and a loss function;
and if the loss function value is smaller than or equal to a preset loss function threshold, determining that the template recommendation model converges, and obtaining the trained template recommendation model.
Optionally, the first obtaining module is configured to:
acquiring target historical behavior information corresponding to the target object according to the target object, wherein the target historical behavior information comprises data generated by the target object on the using behavior of the alternative template in a historical period;
Determining first data and second data corresponding to the target object according to the target historical behavior information, wherein the first data are used for representing the sensitivity degree of the object to the template characteristics, and the second data are used for representing the sensitivity degree of the object to the template display effect;
and determining the first data and/or the second data corresponding to the target object as object characteristic information of the target object.
Optionally, the apparatus further includes a second acquisition module configured to:
acquiring object attribute information corresponding to the target object;
the determining the first data and/or the second data corresponding to the target object as object feature information of the target object includes:
and determining at least one of the first data and the second data corresponding to the target object and the object attribute information as object feature information of the target object.
Optionally, the second determining module is configured to:
and selecting a preset number of alternative templates from the plurality of alternative templates according to the sequence of the recommended value from large to small, and obtaining a plurality of target templates associated with the target object.
According to a third aspect of embodiments of the present disclosure, there is provided 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 template recommendation method according to the first aspect.
According to a fourth aspect of embodiments of the present disclosure, there is provided a computer readable storage medium, which when executed by a processor of an electronic device, causes the electronic device to perform the template recommendation method as described in the first aspect.
The technical scheme provided by the embodiment of the disclosure at least brings the following beneficial effects:
according to the template recommendation method, device, equipment and medium, on one hand, the template is recommended without using the association relation between the template and the terminal equipment model, which is established manually, so that interference of human factors in the template recommendation process based on the terminal equipment performance is reduced, and the accuracy of the determined template recommendation result is improved; on the other hand, in the template recommendation process, the template recommendation is performed by combining the object characteristic information and the characteristics of the template while considering the performance of the terminal equipment, so that the template use preference of the object can be further met and the template use experience of the object is improved on the premise of ensuring that the template recommendation result is matched with the performance of the terminal equipment.
It is to be understood that both the foregoing general description and the following detailed description are exemplary and explanatory only and are not restrictive of the disclosure.
Drawings
The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments consistent with the disclosure and together with the description, serve to explain the principles of the disclosure and do not constitute an undue limitation on the disclosure.
FIG. 1 is a schematic architecture diagram of a template recommendation system, shown in accordance with an exemplary embodiment;
FIG. 2 is a flowchart illustrating a template recommendation method according to an exemplary embodiment;
FIG. 3 is a flowchart illustrating a template recommendation model training method, according to an exemplary embodiment;
FIG. 4 is a block diagram of a template recommender, according to an exemplary embodiment;
fig. 5 is a block diagram of an electronic device, according to an example embodiment.
Detailed Description
In order to enable those skilled in the art to better understand the technical solutions of the present disclosure, the technical solutions of the embodiments of the present disclosure will be clearly and completely described below with reference to the accompanying drawings.
It should be noted that the terms "first," "second," and the like in the description and claims of the present disclosure and in the foregoing figures are used for distinguishing between similar objects and not necessarily for describing a particular sequential or chronological order. It is to be understood that the data so used may be interchanged where appropriate such that the embodiments of the disclosure described herein may be capable of operation in sequences other than those illustrated or described herein. The implementations described in the following exemplary examples are not representative of all implementations consistent with the present disclosure. Rather, they are merely examples of apparatus and methods consistent with some aspects of the present disclosure as detailed in the accompanying claims.
It should be noted that, information related to the user (including, but not limited to, terminal device information of the user, personal information of the user, etc.) and data (including, but not limited to, data for presentation, analyzed data, etc.) related to the present disclosure are information and data authorized by the user or sufficiently authorized by each party.
Templates which can be recommended to the user in the social application are various, wherein the data amount of different templates is different, the template with more abundant display effect is displayed, and the more data amount is, the more data amount is displayed by the terminal equipment, the more data processing resources are needed.
In the related art, in order to recommend a suitable template to an object, an association relationship between a terminal device and a recommendable template may be pre-established, and after model information of the terminal device used by a target object to be subjected to template recommendation is obtained, the association relationship between the terminal device and the recommendable template may be searched to determine a template that may be recommended for the target object.
However, on the one hand, because the association relationship between the terminal device and the recommendable template is manually established, errors may exist, that is, the recommendable template of a certain type of terminal device established in the association relationship may not be suitable for the type of terminal device, for example, a template with a larger data volume is recommended to a terminal device with low performance, which may cause a situation of stuck or poor definition in the process of displaying the template, and affect the use experience of an object; on the other hand, since the template recommendation process only considers the equipment condition of the terminal equipment, the recommended template may not conform to the preference of the object and also influence the use experience of the object.
In view of the foregoing, exemplary embodiments of the present disclosure provide a template recommendation method, an application scenario of which includes, but is not limited to: when determining that the object needs to release social content, acquiring object characteristic information of a target object and acquiring terminal performance information corresponding to the target object; determining a recommended value of each candidate template with respect to the target object according to the object characteristic information, the terminal performance information and the template characteristic data of the plurality of candidate templates; and determining a target template associated with the target object from a plurality of candidate templates according to the recommended value of each candidate template about the target object and a predetermined screening condition, and recommending the target template to the terminal equipment of the target object. The social content can comprise short videos, image-text information and the like, and the target object is an object which needs to be subjected to template recommendation. The method can combine object characteristic information, terminal performance information of the terminal equipment and characteristics of the templates, consider various factors to determine a target template which can be recommended for the object, improve matching degree of the recommended template and actual conditions of the terminal equipment, and can also meet preference of the object.
In order to implement the above-described template recommendation method, an exemplary embodiment of the present disclosure provides a template recommendation system. Fig. 1 shows a schematic architecture diagram of the template recommendation system. As shown in fig. 1, the template recommendation system 100 may include a server 110 and a terminal device 120. The server 110 is a background server deployed by a social application service, and the social application may be a short video application, a shopping application, a music application, or the like, which is not limited by the embodiment of the present disclosure. The terminal device 120 may be a terminal device used by an object to which a social application is installed, and more particularly, the terminal device may be, for example, a smart phone, a personal computer, a tablet computer, or the like. The server 110 and the terminal device 120 can establish a connection through a network to realize template recommendation.
It should be understood that the server 110 may be one server or may be a cluster formed by a plurality of servers, and the specific architecture of the server 110 is not limited in this disclosure.
In an alternative embodiment, the terminal device 120 may obtain object feature information of the target object after detecting the template loading operation, and obtain terminal performance information corresponding to the target object; determining a recommended value of each candidate template with respect to the target object according to the object characteristic information, the terminal performance information and the template characteristic data of the plurality of candidate templates; finally, according to the recommended value of each candidate template about the target object and the predetermined screening condition, determining the target template associated with the target object from the plurality of candidate templates, and displaying the target template in the terminal device 120, wherein the target object can be the object logged in the terminal device.
In an alternative embodiment, the terminal device 120 may collect terminal performance information after detecting a template loading operation, generate a template recommendation request according to the target object and the terminal performance information, and send the template recommendation request to the server 110, after receiving the template recommendation request, the server 110 may parse the template recommendation request to obtain object feature information and terminal performance information of the target object, and determine a recommendation value of each candidate template with respect to the target object according to the object feature information, the terminal performance information, and template feature data of a plurality of candidate templates; finally, according to the recommended value of each candidate template with respect to the target object and the predetermined screening condition, determining a target template associated with the target object from the plurality of candidate templates, generating template recommendation response information according to the target template associated with the target object, transmitting the template recommendation response information to the terminal device 120, and the terminal device 120 may display the target template.
Fig. 2 is a flowchart illustrating a template recommendation method according to an exemplary embodiment, and the embodiment of the present disclosure uses the template recommendation method applied to a server as an example, and the template recommendation method is illustrated in fig. 2 and includes the following steps.
Step S201, obtaining object characteristic information of a target object and obtaining terminal performance information corresponding to the target object;
step S202, determining a recommended value of each candidate template relative to a target object according to object feature information, terminal performance information and template feature data of a plurality of candidate templates;
the recommendation value is used for representing the priority of recommending the alternative templates for the target object;
step S203, determining a target template associated with the target object from a plurality of candidate templates according to the recommended value of each candidate template about the target object and the predetermined screening condition, and recommending the target template to the terminal equipment of the target object.
In summary, according to the template recommendation method provided by the embodiment of the present disclosure, on one hand, the template is recommended without using the association relationship between the template and the terminal device model, which is manually established, so that interference of human factors in the process of recommending the template based on the performance of the terminal device is reduced, and the accuracy of the determined template recommendation result is improved; on the other hand, in the template recommendation process, the template recommendation is performed by combining the object characteristic information and the characteristics of the template while considering the performance of the terminal equipment, so that the template use preference of the object can be further met and the template use experience of the object is improved on the premise of ensuring that the template recommendation result is matched with the performance of the terminal equipment.
The following describes in detail the specific implementation of each step in the embodiment shown in fig. 2:
in step S201, the server may acquire object feature information of the target object, and acquire terminal capability information corresponding to the target object.
In an embodiment of the disclosure, the object feature information may include first data and second data, the first data is used for representing a sensitivity degree of the object to a template feature, and the template feature may include a template theme, a template style, a template tone, and the like; the second data is used for representing the sensitivity degree of the object to the template display effect; the template display effect may include definition, smoothness, etc. in the template display process.
It should be noted that, in the embodiment of the present disclosure, the larger the first data, the more easily the template usage behavior of the object is affected by the template feature, where the template of the template feature is recommended for the object, and the frequency of using the template by the object may increase; the larger the second data is, the more easily the template use behavior of the object is affected by the template display effect, wherein the template display effect liked by the object is recommended for the object, and the frequency of using the template by the object is increased; the terminal performance information is used to characterize the hardware resource status of the target terminal, and may include central processing unit (Central Processing Unit, CPU) usage, memory usage, graphics processor (graphics processing unit, GPU) usage, etc. of the terminal device.
It will be appreciated that the object characteristic information of the target object may include first data associated with the target object for characterizing the sensitivity of the target object to the template characteristics and second data associated with the target object for characterizing the sensitivity of the target object to the template presentation effects.
In an alternative embodiment, when the object issues social content, after detecting a template loading operation, the terminal device collects terminal performance information, generates a template recommendation request according to the target object and the terminal performance information, and sends the template recommendation request to the server, and the server obtains object feature information of the target object, and the process of obtaining the terminal performance information corresponding to the target object may include: after receiving the template recommendation request, the template recommendation request can be analyzed to obtain a target object and terminal performance information corresponding to the target object, and object feature information of the target object is obtained.
In an alternative embodiment, the process of obtaining object feature information of the target object by the server may include: acquiring target historical behavior information corresponding to a target object according to the target object; determining first data and second data corresponding to the target object according to the target historical behavior information; determining the first data and/or the second data corresponding to the target object as object feature information of the target object; the target historical behavior information comprises data generated by using behaviors of the target object on the alternative templates in a historical period; the historical period may be determined based on actual needs, which is not limited by the embodiments of the present disclosure. For example, the historical period may be one week before the current time. The sensitivity degree of the object to the template characteristics and/or the sensitivity degree of the object to the template display effect can be determined as the object characteristic information, so that the template liked by the object is recommended according to the object characteristic information, the satisfaction degree of the object to the template recommendation result is improved, and the use rate of the template can be improved.
The process of determining, by the server, the first data and the second data corresponding to the target object according to the target historical behavior information may include: acquiring a first social content posting quantity and a second social content posting quantity associated with a target object in a historical period, wherein the first social content posting quantity is the quantity of the object posting social contents when template characteristics of a historical recommendation template are unchanged; when the second social content release amount is the template characteristic of the change history recommendation template, the object releases the quantity of the social content; further, a first difference value between the second social content distribution amount and the first social content distribution amount can be determined, and in a corresponding relation between the pre-established social content distribution amount difference value and the first data, the first data corresponding to the first difference value is searched for, so that the first data corresponding to the target object is obtained. It should be noted that, in the process of determining, by the server, the second data corresponding to the target object according to the target historical behavior information, reference may be made to the process of determining, by the server, the first data corresponding to the target object according to the target historical behavior information, which is not described in detail in the embodiments of the present disclosure.
In an alternative embodiment, the process of obtaining object feature information of the target object by the server may include: and acquiring object characteristic information of the target object according to the target object in the object characteristic information database. In the determining process of the object feature information of the target object stored in the object feature information database, reference may be made to the above embodiment, which is not described in detail in this disclosure. The efficiency of obtaining object feature information of a target object may be provided by looking up object feature information of the target object in an object feature information database.
In an optional implementation manner, in order to further improve the matching degree of the determined template recommendation result and the personal preference of the object, before determining the first data and/or the second data corresponding to the target object as the object feature information of the target object, the server may further acquire object attribute information corresponding to the target object; the process of determining the first data and/or the second data corresponding to the target object as the object feature information of the target object by the server may include: at least one of the first data and the second data corresponding to the target object, and the object attribute information are determined as object feature information of the target object. Wherein the object attribute information is used to characterize the object's own attribute information, such as the object's age, sex, city, occupation, etc. The own attribute information of the object may be determined as object feature information so as to recommend a template for the object to better conform to the object preference according to the object feature information.
In step S202, the server may determine a recommended value of each of the candidate templates with respect to the target object according to the object feature information, the terminal capability information, and the template feature data of the plurality of candidate templates.
In the embodiment of the disclosure, the alternative templates are available recommended templates provided by the server; the template characteristic data of the alternative template can comprise resource characteristics and template characteristics of the alternative template, wherein the resource characteristics of the alternative template can comprise the number of layers in the alternative template, the resolution of the alternative template, the number of background videos in the alternative template and the like; template features of the alternative template may include style of the alternative template, subject matter of the alternative template, hue of the alternative template, and the like; the recommendation value is used to characterize the priority of recommending the alternative template to the target object, and the higher the recommendation value of the alternative template with respect to the target object, the higher the priority of recommending the alternative template to the target object.
In an alternative embodiment, the process of determining, by the server, a recommended value of each candidate template with respect to the target object according to the object feature information, the terminal capability information, and the template feature data of the plurality of candidate templates may include: determining a feature data set to be processed associated with each alternative template, wherein the feature data set to be processed comprises template feature data, object feature information and terminal performance information of the alternative templates; and inputting the feature data set to be processed associated with each alternative template into a pre-trained template recommendation model respectively to obtain a recommendation value of each alternative template on the target object. The efficiency and the accuracy of determining the recommended value of the candidate template relative to the target object can be improved through a pre-trained template recommendation model.
In an alternative embodiment, the process of inputting the feature data set to be processed associated with each candidate template into the pre-trained template recommendation model, respectively, to obtain the recommendation value of each candidate template with respect to the target object may include: and sequentially inputting the feature data set to be processed associated with each alternative template into a pre-trained template recommendation model to obtain a recommendation value of each alternative template on the target object. The recommended value of each alternative template relative to the target object can be determined through a template recommended model, so that the occupation of memory resources of the server is reduced.
In an alternative embodiment, the process of inputting the feature data set to be processed associated with each candidate template into the pre-trained template recommendation model, respectively, to obtain the recommendation value of each candidate template with respect to the target object may include: and inputting the feature data set to be processed associated with each alternative template into a plurality of pre-trained template recommendation models simultaneously to obtain a recommendation value of each alternative template on the target object. The recommendation value of each alternative template relative to the target object can be determined through a plurality of template recommendation models, so that the efficiency of determining the recommendation value of each alternative template relative to the target object is improved, and the template recommendation efficiency is further improved. Wherein the model structure and model parameters of the plurality of template recommendation models are the same.
In an alternative embodiment, the template recommendation method further comprises a training method of the template recommendation model. As shown in fig. 3, the training method of the template recommendation model includes:
step S301, sample object feature information, sample terminal performance information of a sample object, and sample template feature data of a sample candidate template used by the sample object are acquired.
In embodiments of the present disclosure, the sample object may include any object that publishes social content with a social application prior to training of the template recommendation model.
In an alternative embodiment, the process of obtaining, by the server, sample object feature information of the sample object, sample terminal performance information, and sample template feature data of the sample object used sample candidate template may include: sample object feature information, sample terminal performance information, and sample template feature data of a sample object used sample candidate template are obtained in a sample history period, wherein the sample history period can be determined based on actual needs, and the embodiment of the disclosure is not limited to this. For example, the sample history period may be one week before the model training time.
It should be noted that, in the embodiment of the present disclosure, for an object that uses a social application to publish social content, sample object feature information of the object, terminal performance information, and information of an alternative template used by the object may be stored in a database in a server, so as to directly obtain sample data in a training stage of a template recommendation model.
Step S302, performing iterative training on a template recommendation model to be trained based on sample object feature information, sample terminal performance information and sample template feature data of a sample alternative template to obtain a trained template recommendation model.
In the embodiment of the present disclosure, the template recommendation model to be trained may be a machine learning model, and by way of example, the template recommendation model to be trained may be a neural network model, or a clustering model, etc., which is not limited in the embodiment of the present disclosure. The template recommendation model of the multi-element feature recommendation template can be trained by utilizing sample object portrait data of sample objects, sample terminal performance information and sample template feature data of sample alternative templates, and the accuracy of a template recommendation result determined by the trained template recommendation model in practical application is improved.
In an alternative embodiment, the server training the template recommendation model to be trained may be a supervised training process, and the server may further perform iterative training on the template recommendation model to be trained based on the sample object feature information, the sample terminal performance information, and the sample template feature data of the sample candidate template before obtaining the trained template recommendation model: a sample recommendation value for each sample candidate template is determined for the sample object, it being understood that the sample recommendation score is used to characterize the priority of recommending sample templates for the sample object, the higher the sample candidate template recommendation value for the sample object, the higher the priority of recommending sample templates to the sample object.
Wherein the process of the server determining a sample recommendation value for each sample candidate template with respect to the sample object may include: determining the number of uses of the sample object for each sample alternative template over a sample history period; and determining a sample recommended value corresponding to the use times of each sample alternative template according to the pre-established association relation between the use times of the sample alternative templates and the sample recommended value, and obtaining the sample recommended value of each sample alternative template about the sample object.
In an alternative embodiment, in the supervised training process, the server performs iterative training on the template recommendation model to be trained based on the sample object feature information, the sample terminal performance information and the sample template feature data of the sample candidate template, and the process of obtaining the trained template recommendation model may include: inputting sample object characteristic information, sample terminal performance information and sample template characteristic data of a target sample alternative template into a template recommendation model to be trained to obtain a prediction recommendation value of a sample object about the target sample alternative template; determining a loss function value according to the sample recommended value and the predicted recommended value of the sample object relative to the target sample candidate template and the loss function; if the loss function value is smaller than or equal to a preset loss function threshold value, determining that the template recommendation model converges to obtain a trained template recommendation model; if the loss function value is larger than the preset loss function threshold, determining that the template recommendation model is not converged, continuously adjusting parameters of the template recommendation model, and repeating the process until the loss function value is smaller than or equal to the preset loss function threshold, and obtaining the trained template recommendation model. The template recommendation model can be subjected to supervised training by combining with sample recommendation value training, the sample recommendation value is determined according to the sample candidate template use behavior of the sample object, and the recommendation result determined by the template recommendation model in the actual application process can be more in line with the actual template use preference of the object.
In an optional implementation manner, the server training the template recommendation model to be trained may be an unsupervised training process, and the server performs iterative training on the template recommendation model to be trained based on sample object feature information, sample terminal performance information and sample template feature data of the sample alternative template, so that a process of obtaining a trained template recommendation model may be implemented based on a clustering algorithm, which is not described in detail in the embodiments of the present disclosure.
Step S203, determining a target template associated with the target object from a plurality of candidate templates according to the recommended value of each candidate template about the target object and the predetermined screening condition, and recommending the target template to the terminal equipment of the target object.
In an alternative embodiment, the process of determining the target template associated with the target object among the plurality of candidate templates according to the recommended value of each candidate template with respect to the target object and the predetermined screening condition by the server may include: and selecting a preset number of alternative templates from the plurality of alternative templates according to the sequence of the recommended value from large to small, so as to obtain a plurality of target templates associated with the target object. The preset number may be determined based on actual needs, which is not limited in the embodiment of the present disclosure, and the preset number may be 10, for example, and may provide a plurality of target templates for the object to select, so as to further improve the diversity of object template selection and improve the use experience of the object.
The process of recommending the target template to the terminal device of the target object by the server can comprise the following steps: generating template recommendation response information according to the plurality of target templates, and sending the template recommendation response information to terminal equipment of the target object; the terminal device may display a plurality of target templates on the template display page in response to receiving the template recommendation response information.
In an alternative embodiment, after the server recommends the target template to the terminal device of the target object, a recommended template update request sent by the terminal device may be received, and a preset number of candidate templates are selected from the updated plurality of candidate templates according to the order of the recommended values from large to small, so as to obtain updated plurality of target templates associated with the target object, and the updated plurality of target templates are sent to the terminal device. Wherein the updated plurality of alternative templates does not include the alternative templates that have been sent to the terminal device. A new template may be recommended for the object when the object is not full of the first recommended template.
In an alternative embodiment, the process of determining the target template associated with the target object among the plurality of candidate templates according to the recommended value of each candidate template with respect to the target object and the predetermined screening condition by the server may include: and determining the candidate template with the highest recommended value as a target template, and sending the target template to the terminal equipment of the target object for display by the terminal equipment.
In an alternative embodiment, the process of determining, by the server, a recommended value of each candidate template with respect to the target object according to the object feature information, the terminal capability information, and the template feature data of the plurality of candidate templates may include: and for each alternative template, determining a feature data set to be processed associated with each alternative template, and processing the feature data set to be processed associated with each alternative template by utilizing a pre-established fitting function to obtain a recommended value of each alternative template relative to the target object.
Fig. 4 is a template recommending apparatus according to an exemplary embodiment, and as shown in fig. 4, a template recommending apparatus 400 includes:
a first obtaining module 401 configured to obtain object feature information of a target object, and obtain terminal performance information corresponding to the target object;
a first determining module 402 configured to determine a recommendation value of each candidate template with respect to the target object according to the object feature information, the terminal capability information, and the template feature data of the plurality of candidate templates, the recommendation value being used for characterizing a priority of recommending the candidate templates for the target object;
the second determining module 403 is configured to determine a target template associated with the target object among the plurality of candidate templates according to the recommended value of each candidate template with respect to the target object and a predetermined screening condition, and recommend the target template to the terminal device of the target object.
Optionally, the first determining module 402 is configured to:
determining a feature data set to be processed associated with each alternative template, wherein the feature data set to be processed comprises template feature data, object feature information and terminal performance information of the alternative templates;
and respectively inputting the feature data sets to be processed associated with each alternative template into a pre-trained template recommendation model to obtain a recommendation value of each alternative template on the target object.
Optionally, as shown in fig. 4, the template recommending apparatus 400 further includes a model training module 404 configured to:
acquiring sample object feature information, sample terminal performance information of a sample object and sample template feature data of a sample alternative template used by the sample object;
and carrying out iterative training on the template recommendation model to be trained based on the sample object characteristic information, the sample terminal performance information and the sample template characteristic data of the sample alternative template to obtain a trained template recommendation model.
Optionally, as shown in fig. 4, the template recommending apparatus 400 further includes a third determining module 405 configured to:
determining a sample recommendation value for each sample candidate template with respect to the sample object;
Based on sample object feature information, sample terminal performance information and sample template feature data of a sample alternative template, performing iterative training on a template recommendation model to be trained to obtain a trained template recommendation model, including:
inputting sample object characteristic information, sample terminal performance information and sample template characteristic data of a target sample alternative template into a template recommendation model to be trained to obtain a prediction recommendation value of a sample object about the target sample alternative template;
determining a loss function value according to the sample recommended value and the predicted recommended value of the sample object relative to the target sample candidate template and the loss function;
and if the loss function value is smaller than or equal to the preset loss function threshold value, determining that the template recommendation model converges, and obtaining a trained template recommendation model.
Optionally, the first obtaining module 401 is configured to:
acquiring target historical behavior information corresponding to a target object according to the target object, wherein the target historical behavior information comprises data generated by the target object on the use behavior of the alternative template in a historical period;
determining first data and second data corresponding to the target object according to the target historical behavior information, wherein the first data is used for representing the sensitivity degree of the object to the template characteristics, and the second data is used for representing the sensitivity degree of the object to the template display effect;
And determining the first data and/or the second data corresponding to the target object as object characteristic information of the target object.
Optionally, as shown in fig. 4, the template recommending apparatus 400 further includes a second obtaining module 406 configured to:
acquiring object attribute information corresponding to a target object;
determining the first data and/or the second data corresponding to the target object as object feature information of the target object, including:
at least one of the first data and the second data corresponding to the target object, and the object attribute information are determined as object feature information of the target object.
Optionally, the second determining module 403 is configured to:
and selecting a preset number of alternative templates from the plurality of alternative templates according to the sequence of the recommended value from large to small, so as to obtain a plurality of target templates associated with the target object.
The specific manner in which the various modules perform the operations in the apparatus of the above embodiments have been described in detail in connection with the embodiments of the method, and will not be described in detail herein.
The exemplary embodiments of the present disclosure also provide an electronic device, which may be a terminal device or a server. The electronic device is described below with reference to fig. 5. It should be understood that the electronic device is described below with reference to fig. 5. It should be understood that the electronic device 500 shown in fig. 5 is merely an example and should not be construed to limit the functionality and scope of use of embodiments of the present disclosure in any way.
As shown in fig. 5, the electronic device 500 is embodied in the form of a general purpose computing device. The components of electronic device 500 may include, but are not limited to: at least one processing unit 510, at least one memory unit 520, a bus 530 connecting the different system components, including the memory unit 520 and the processing unit 510.
Wherein the storage unit stores program code that is executable by the processing unit 510 such that the processing unit 510 performs steps according to various exemplary embodiments of the present invention described in the above section of the "exemplary method" of the present specification. For example, the processing unit 510 may perform the method steps shown in fig. 2-3, etc.
The memory unit 520 may include volatile memory units, such as Random Access Memory (RAM) 521 and/or cache memory 522, and may further include Read Only Memory (ROM) 523.
The storage unit 520 may also include a program/utility 524 having a set (at least one) of program modules 525, such program modules 525 including, but not limited to: an operating system, one or more application programs, other program modules, and program data, each or some combination of which may include an implementation of a network environment.
Bus 530 may include a data bus, an address bus, and a control bus.
The electronic device 500 may also communicate with one or more external devices 600 (e.g., keyboard, pointing device, bluetooth device, etc.) via an input/output (I/O) interface 540. Electronic device 500 may also communicate with one or more networks such as a Local Area Network (LAN), a Wide Area Network (WAN), and/or a public network such as the internet through network adapter 580. As shown, network adapter 580 communicates with other modules of electronic device 500 over bus 530. It should be appreciated that although not shown, other hardware and/or software modules may be used in connection with electronic device 500, including, but not limited to: microcode, device drivers, redundant processing units, external disk drive arrays, RAID systems, tape drives, data backup storage systems, and the like.
It should be noted that although in the above detailed description several modules or units of a device for action execution are mentioned, such a division is not mandatory. Indeed, the features and functionality of two or more modules or units described above may be embodied in one module or unit in accordance with exemplary embodiments of the present disclosure. Conversely, the features and functions of one module or unit described above may be further divided into a plurality of modules or units to be embodied.
Those skilled in the art will appreciate that the various aspects of the present disclosure may be implemented as a system, method, or program product. Accordingly, various aspects of the disclosure may be embodied in the following forms, namely: an entirely hardware embodiment, an entirely software embodiment (including firmware, micro-code, etc.) or an embodiment combining hardware and software aspects may be referred to herein as a "circuit," module "or" system. Other embodiments of the disclosure will be apparent to those skilled in the art from consideration of the specification and practice of the disclosure disclosed herein. This disclosure is intended to cover any adaptations, uses, or adaptations of the disclosure following the general principles of the disclosure and including such departures from the present disclosure as come within known or customary practice within the art to which the disclosure pertains. It is intended that the specification and examples be considered as exemplary only, with a true scope and spirit of the disclosure being indicated by the following claims.
In addition, the present disclosure also provides a computer-readable storage medium, which when executed by a processor of an electronic device, enables the electronic device to perform the template recommendation method as provided in the above embodiments.
In addition, the present disclosure also provides a computer program product comprising computer instructions which, when run on an electronic device, cause the electronic device to perform the template recommendation method as provided in the above embodiments.
Other embodiments of the disclosure will be apparent to those skilled in the art from consideration of the specification and practice of the disclosure disclosed herein. This application is intended to cover any adaptations, uses, or adaptations of the disclosure following, in general, the principles of the disclosure and including such departures from the present disclosure as come within known or customary practice within the art to which the disclosure pertains. It is intended that the specification and examples be considered as exemplary only, with a true scope and spirit of the disclosure being indicated by the following claims.

Claims (10)

1. A template recommendation method, comprising:
object feature information of a target object is obtained, and terminal performance information corresponding to the target object is obtained;
determining a recommended value of each candidate template with respect to the target object according to the object characteristic information, the terminal performance information and template characteristic data of a plurality of candidate templates, wherein the recommended value is used for representing priority of recommending the candidate templates to the target object;
And determining a target template associated with the target object from the plurality of candidate templates according to the recommended value of each candidate template about the target object and a predetermined screening condition, and recommending the target template to the terminal equipment of the target object.
2. The template recommendation method according to claim 1, wherein the determining a recommendation value of each candidate template with respect to the target object based on the object feature information, terminal capability information, and template feature data of a plurality of candidate templates comprises:
determining a feature data set to be processed associated with each alternative template, wherein the feature data set to be processed comprises template feature data of the alternative templates, object feature information and terminal performance information;
and respectively inputting the feature data set to be processed associated with each alternative template into a pre-trained template recommendation model to obtain a recommendation value of each alternative template on the target object.
3. The template recommendation method according to claim 2, further comprising a training method of the template recommendation model, the training method of the template recommendation model comprising:
Acquiring sample object characteristic information, sample terminal performance information of a sample object and sample template characteristic data of a sample alternative template used by the sample object;
and carrying out iterative training on the template recommendation model to be trained based on the sample object characteristic information, the sample terminal performance information and the sample template characteristic data of the sample alternative template to obtain the trained template recommendation model.
4. The template recommendation method according to claim 3, wherein before performing iterative training on a template recommendation model to be trained based on the sample object feature information, the sample terminal performance information, and the sample template feature data of the sample candidate template, the method further comprises:
determining a sample recommendation value for each sample candidate template with respect to the sample object;
the performing iterative training on the template recommendation model to be trained based on the sample object feature information, the sample terminal performance information and the sample template feature data of the sample candidate template to obtain a trained template recommendation model comprises the following steps:
Inputting the sample object characteristic information, the sample terminal performance information and sample template characteristic data of a target sample alternative template into the template recommendation model to be trained to obtain a prediction recommendation value of the sample object about the target sample alternative template;
determining a loss function value according to the sample recommended value and the predicted recommended value of the sample object relative to the target sample candidate template and a loss function;
and if the loss function value is smaller than or equal to a preset loss function threshold, determining that the template recommendation model converges, and obtaining the trained template recommendation model.
5. The template recommendation method according to claim 1, wherein the obtaining object feature information of the target object includes:
acquiring target historical behavior information corresponding to the target object according to the target object, wherein the target historical behavior information comprises data generated by the target object on the using behavior of the alternative template in a historical period;
determining first data and second data corresponding to the target object according to the target historical behavior information, wherein the first data are used for representing the sensitivity degree of the object to the template characteristics, and the second data are used for representing the sensitivity degree of the object to the template display effect;
And determining the first data and/or the second data corresponding to the target object as object characteristic information of the target object.
6. The template recommendation method according to claim 5, wherein before determining the first data and/or the second data corresponding to the target object as object feature information of the target object, the method further comprises:
acquiring object attribute information of the target object;
the determining the first data and/or the second data corresponding to the target object as object feature information of the target object includes:
and determining at least one of the first data and the second data corresponding to the target object and the object attribute information as object feature information of the target object.
7. The template recommendation method according to any one of claims 1 to 6, wherein the determining a target template associated with the target object among the plurality of candidate templates according to a recommended value of each candidate template with respect to the target object and a predetermined screening condition, comprises:
and selecting a preset number of alternative templates from the plurality of alternative templates according to the sequence of the recommended value from large to small, and obtaining a plurality of target templates associated with the target object.
8. A template recommending apparatus, comprising:
the terminal performance information acquisition module is configured to acquire object characteristic information of a target object and terminal performance information corresponding to the target object;
a first determining module configured to determine a recommendation value of each candidate template with respect to the target object according to the object feature information, the terminal performance information, and template features of a plurality of candidate templates, the recommendation value being used to characterize a priority of recommending the candidate templates for the target object;
and a second determining module configured to determine a target template associated with the target object among the plurality of candidate templates according to a recommended value of each candidate template with respect to the target object and a predetermined screening condition, and recommend the target template to a terminal device of the target object.
9. 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 template recommendation method of any one of claims 1 to 7.
10. A computer readable storage medium, characterized in that instructions in the computer readable storage medium, when executed by a processor of an electronic device, enable the electronic device to perform the template recommendation method of any one of claims 1 to 7.
CN202310108211.6A 2023-02-01 2023-02-01 Template recommendation method, device, equipment and medium Pending CN116186398A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202310108211.6A CN116186398A (en) 2023-02-01 2023-02-01 Template recommendation method, device, equipment and medium

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202310108211.6A CN116186398A (en) 2023-02-01 2023-02-01 Template recommendation method, device, equipment and medium

Publications (1)

Publication Number Publication Date
CN116186398A true CN116186398A (en) 2023-05-30

Family

ID=86436031

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202310108211.6A Pending CN116186398A (en) 2023-02-01 2023-02-01 Template recommendation method, device, equipment and medium

Country Status (1)

Country Link
CN (1) CN116186398A (en)

Similar Documents

Publication Publication Date Title
CN109902849B (en) User behavior prediction method and device, and behavior prediction model training method and device
WO2020135535A1 (en) Recommendation model training method and related apparatus
US20200286154A1 (en) Utilizing item-level importance sampling models for digital content selection policies
JP2024503774A (en) Fusion parameter identification method and device, information recommendation method and device, parameter measurement model training method and device, electronic device, storage medium, and computer program
CN111142968A (en) Electronic equipment configuration recommendation processing method and device and storage medium
US11663509B2 (en) System and method for a personalized machine learning pipeline selection and result interpretation
CN111881358B (en) Object recommendation system, method and device, electronic equipment and storage medium
CN113254679B (en) Multimedia resource recommendation method and device, electronic equipment and storage medium
CN112883265A (en) Information recommendation method and device, server and computer readable storage medium
CN113919923B (en) Live broadcast recommendation model training method, live broadcast recommendation method and related equipment
CN113641835B (en) Multimedia resource recommendation method and device, electronic equipment and medium
CN112269943B (en) Information recommendation system and method
CN111680218B (en) User interest identification method and device, electronic equipment and storage medium
CN117235371A (en) Video recommendation method, model training method and device
CN113836388A (en) Information recommendation method and device, server and storage medium
CN112269942B (en) Method, device and system for recommending object and electronic equipment
CN113420165B (en) Training of classification model and classification method and device of multimedia data
CN116186398A (en) Template recommendation method, device, equipment and medium
CN114510627A (en) Object pushing method and device, electronic equipment and storage medium
CN114580652A (en) Method and system for item recommendation applied to automatic artificial intelligence
CN111382874B (en) Method and device for realizing update iteration of online machine learning model
US11611483B2 (en) Learning-based dynamic determination of synchronous/asynchronous behavior of computing services
CN114297417A (en) Multimedia resource recommendation method and related device
CN114528472A (en) Resource recommendation model training method, resource information recommendation method and device
CN116662652A (en) Model training method, resource recommendation method, sample generation method and device

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination