Data Processing Method and Computing Device

Abstract

An embodiment of the present application provides a data processing method and a computing device. The method includes: determining a target parameter instance corresponding to a proportional parameter in a target processing system; obtaining usage effect information generated by the target user using the target processing system corresponding to the target parameter instance; generating system feedback information generated by the target parameter instance in the target processing system based on the usage effect information; wherein the system feedback information is used to provide feedback to the target user. The embodiment of the present application improves the parameter update efficiency of the target processing system.

Description

Technical Field

[0001] This application relates to the technical field of electronic devices, and in particular, to a data processing method and a computing device. Background Art

[0002] A parameter is used to express the changes of several variables and the mutual relationship between variables in a mathematical calculation process. A proportional parameter can be composed of multiple parameters respectively according to a certain proportion, and the sum of the percentages of each parameter is 1. Assuming there are n parameters, {a1, a2, a3, …, a n}, where ai is a positive decimal within the range of [0, 1], and With the popularization of cloud computing, online systems based on cloud computing can provide a large amount of data computing to meet people's daily life needs. With the increase in system complexity, many proportional parameters are often involved in the system. The selection of the proportion of different parameters in the proportional parameters often has a relatively important impact on the calculation result. Therefore, it is necessary to set relatively optimal proportion values for each parameter in the proportional parameters.

[0003] In the prior art, when the parameter values of multiple parameters in a proportional parameter are determined respectively, a parameter instance can be formed. The setting of the parameter instance of the proportional parameter is usually carried out manually, relying on historical experience, and continuously performing parameter experiments to set different parameter instances, and performing application experiments on the specific application environment of each parameter instance to obtain the experimental results of each parameter instance, and selecting the target parameter instance from different parameter instances according to the experimental results of each parameter instance.

[0004] However, the method of manually setting parameter instances to obtain the target parameter instance requires the operation and maintenance personnel to set the proportion values of multiple parameters of the proportional parameter multiple times and perform usage tests to update the parameters, with high costs and low parameter update efficiency. Summary of the Invention

[0005] In view of this, embodiments of this application provide a data processing method and a computing device to solve the technical problem of low update efficiency of manually debugging proportional parameters in the prior art.

[0006] In a first aspect, an embodiment of this application provides a data processing method, including:

[0007] Determine a target parameter instance corresponding to a proportional parameter in a target processing system;

[0008] Obtain usage effect information generated by a target user using the target processing system corresponding to the target parameter instance;

[0009] Generate system feedback information generated by the target parameter instance in the target processing system based on the usage effect information; wherein, the system feedback information is used to provide feedback to the target user.

[0010] In a second aspect, an embodiment of the present application provides a data processing method, including:

[0011] Determine a system access request initiated by a target user for a network trading system;

[0012] In response to the system access request, determine a target parameter instance corresponding to a proportional parameter in the network trading system;

[0013] Obtain usage effect information generated by the target user using the network trading system corresponding to the target parameter instance;

[0014] Generate system feedback information generated by the target parameter instance in the network trading system based on the usage effect information; wherein, the system feedback information is used to provide feedback to the target user.

[0015] In a third aspect, an embodiment of the present application provides a data processing method, including:

[0016] In response to a usage request for invoking a data processing interface, determine processing resources corresponding to the data processing interface;

[0017] Use the processing resources corresponding to the data processing interface to perform the following steps:

[0018] Determine a target parameter instance corresponding to a proportional parameter in a target processing system;

[0019] Obtain usage effect information generated by a target user using the target processing system corresponding to the target parameter instance;

[0020] Generate system feedback information generated by the target parameter instance in the target processing system based on the usage effect information; wherein, the system feedback information is used to provide feedback to the target user.

[0021] In a fourth aspect, an embodiment of the present application provides a data processing device, including:

[0022] An instance determination module, configured to determine a target parameter instance corresponding to a proportional parameter in a target processing system;

[0023] An information acquisition module, configured to obtain usage effect information generated by a target user using the target processing system corresponding to the target parameter instance;

[0024] A feedback generation module, configured to generate system feedback information generated by the target parameter instance in the target processing system based on the usage effect information; wherein, the system feedback information is used to feedback to the target user.

[0025] In a fifth aspect, an embodiment of the present application provides a computing device, including:

[0026] A storage component and a processing component; the storage component is used to store one or more computer instructions, and the one or more computer instructions are called by the processing component;

[0027] The processing component is used to:

[0028] Determine a target parameter instance corresponding to a proportional parameter in the target processing system; obtain usage effect information generated by the target user using the target processing system corresponding to the target parameter instance; generate system feedback information generated by the target parameter instance in the target processing system based on the usage effect information; wherein, the system feedback information is used to feedback to the target user.

[0029] In the embodiment of the present application, a target parameter instance corresponding to a proportional parameter in the target processing system is determined. By obtaining the usage effect information generated by the target user using the target processing system corresponding to the target parameter instance, system feedback information generated by the target parameter instance in the target processing system can be generated based on the usage effect information, and the system feedback information is used to feedback to the target user. By automatically generating the target parameter instance and automatically evaluating the usage effect of the target parameter instance, the usage effect of the parameter instance can be quickly evaluated, and the test efficiency of the proportional parameter can be improved. Description of the Drawings

[0030] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the following will briefly introduce the drawings required for the description of the embodiments or the prior art. Obviously, the drawings in the following description are some embodiments of the present application. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts.

[0031] Figure 1 It is a flowchart of an embodiment of a data processing method provided by an embodiment of the present application;

[0032] Figure 2 It is a flowchart of another embodiment of a data processing method provided by an embodiment of the present application;

[0033] Figure 3 It is a flowchart of another embodiment of a data processing method provided by an embodiment of the present application;

[0034] Figure 4 An example diagram of a data processing method provided by an embodiment of the present application;

[0035] Figure 5 A flowchart of another embodiment of a data processing method provided by an embodiment of the present application;

[0036] Figure 6 A flowchart of another embodiment of a data processing method provided by an embodiment of the present application;

[0037] Figure 7 An example diagram of a data processing method provided by an embodiment of the present application;

[0038] Figure 8 A flowchart of another embodiment of a data processing method provided by an embodiment of the present application;

[0039] Figure 9 A schematic structural diagram of an embodiment of a computing device provided by an embodiment of the present application. Detailed implementation manners

[0040] To make the objectives, technical solutions and advantages of the embodiments of the present application clearer, the technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present application. Apparently, the described embodiments are some, but not all, of the embodiments of the present application. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present application without creative efforts shall fall within the protection scope of the present application.

[0041] The terms used in the embodiments of the present application are only for the purpose of describing specific embodiments, and are not intended to limit the present application. The singular forms "a", "said" and "the" used in the embodiments of the present application and the appended claims are also intended to include the plural forms unless the context clearly indicates otherwise. "Plural" generally includes at least two, but does not exclude the case of including at least one.

[0042] It should be understood that the term " / and" used herein is only a description of the association relationship of associated objects, indicating that there can be three relationships. For example, A and / or B can represent: A exists alone, A and B exist simultaneously, and B exists alone. In addition, the character " / " herein generally represents an "or" relationship between the associated objects before and after.

[0043] Depending on the context, as used herein, the words "if" and "when" can be interpreted as "when" or "while" or "in response to determining" or "in response to identifying". Similarly, depending on the context, the phrase "if determined" or "if (stated condition or event) is identified" can be interpreted as "when determined" or "in response to determining" or "when (stated condition or event) is identified" or "in response to identifying (stated condition or event)".

[0044] It should also be noted that the term "comprising", "including" or any other variation thereof is intended to cover non-exclusive inclusion, such that a commodity or system comprising a series of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such commodity or system. Without further limitation, an element defined by the statement "comprising an..." does not exclude the presence of additional identical elements in the commodity or system comprising said element.

[0045] The technical solution of the embodiment of the present application can be applied to the optimization scenario of proportional parameters. By predicting the effect of parameter instances of proportional parameters, automatic evaluation of parameter instances can be achieved, and the update efficiency of parameter instances can be improved.

[0046] In the prior art, in the parameter optimization scenario of an online system, there are many hyperparameter optimization problems. The selection of these hyperparameters will significantly affect the system operation ability. In order to obtain actual and accurate operation effects, generally, manual methods are adopted, relying on manual experience to set the parameter instances of hyperparameters in the online system. However, the cost of setting parameter instances by relying on manual methods is relatively high, the efficiency is relatively low, and the use effects of instances often cannot be guaranteed.

[0047] In the embodiment of the present application, a target parameter instance corresponding to a proportional parameter in a target processing system is determined. By obtaining the usage effect information generated by the target processing system when the target user uses the target parameter instance, based on this usage effect information, system feedback information generated by the target parameter instance in the target processing system can be generated, and this system feedback information is used to feedback to the target user. By automatically generating target parameter instances and automatically evaluating the usage effects of target parameter instances, rapid evaluation of the usage effects of parameter instances can be realized, and the update efficiency of proportional parameters can be improved.

[0048] The embodiments of the present application will be described in detail below with reference to the accompanying drawings.

[0049] As Figure 1 shown, it is a flowchart of an embodiment of a data processing method provided by an embodiment of the present application. The method may include the following steps:

[0050] 101: Determine the target parameter instance corresponding to the proportional parameter in the target processing system.

[0051] The data processing method provided by the embodiments of the present application can be applied to a computing device, which can include, for example: a computer, a server, a cloud server, a super personal computer, a laptop, a tablet computer, etc. The embodiments of the present application do not make excessive limitations on the specific type of the computing device.

[0052] Optionally, the target processing system may include a network platform capable of providing system application services. The system type of the target processing system may include an online system, which can provide online external services to target users. The target processing system may include, for example, an e-commerce platform, a financial cloud platform, a logistics cloud platform, a government affairs cloud platform, an online game platform, an online education platform, a social platform, an energy management platform, an intelligent manufacturing platform, a medical service platform, etc. The embodiments of the present application do not make excessive limitations on the specific type of the target processing system.

[0053] When a target user uses the target processing system, the user can first register in the target processing system to obtain a system account and password. In addition, in some embodiments, the user can also obtain the usage permission of the target processing system by initiating a usage request, paying usage fees, etc.

[0054] The proportional parameter may include a parameter associated with a system application request initiated by a user in the target processing system, which can be a normal parameter or a hyperparameter. During the process of the user using the target processing system, the proportional parameter will be used. The proportional parameter may include multiple system sub-parameters, and the sum of the parameter values corresponding to the multiple system sub-parameters is 1. Different system sub-parameters represent different system meanings. Taking the e-commerce scenario as an example, the proportional parameter may be related products recommended to the user. Assuming that the proportional parameter includes 3 system sub-parameters, among them, the first system sub-parameter may represent that the product type recommended to the user is the product browsed by the user, the second system sub-parameter may represent the latest product sorted based on the online time, and the third system sub-parameter may represent the product filtered based on the filtering technology.

[0055] The target parameter instance may include the parameter values of the proportional parameter. When the proportional parameter is composed of multiple system sub-parameters, the target parameter instance may include the parameter values corresponding to the multiple system sub-parameters respectively. The sum of the parameter values corresponding to the multiple system sub-parameters is 1. Assuming that there are n system sub-parameters, and the parameter values corresponding to the n sub-parameters are a i , where ai ∈ [0, 1],

[0056] 102: Obtain the usage effect information generated by the target user using the target processing system corresponding to the target parameter instance.

[0057] When determining the target parameter instance, the target processing system can be set according to the target parameter instance and presented to the target user for use. When the target user uses the target processing system, certain usage information will be generated. Through the usage information generated by the target user, the usage effect information of the target user using the target processing system can be obtained. The usage information may include the operation information performed by the target user on the target processing system. For example, click operations or non-click operations performed on page objects or controls of the target processing system, and the browsing time of the page of the target processing system, etc.

[0058] The usage effect information may include the evaluation results of the usage results or usage information generated by the target user using the target processing system. For example, the usage results generated by the target user using the target processing system can be scored or probability calculated to obtain the effect score. The usage effect information can be used to indicate the advantages and disadvantages of the target processing system corresponding to the target parameter instance.

[0059] 103: Generate system feedback information generated by the target parameter instance in the target processing system based on the usage effect information.

[0060] Among them, the system feedback information is used to feedback to the target user.

[0061] The usage effect information can be directly fed back to the target processing system. For example, the parameter instance of the proportional parameter can be tested through the usage effect information to update the target processing system, and the updated target processing system is presented to the target user for reuse. Using the usage effect information to update the target processing system in a timely manner can ensure the timeliness and effectiveness of the system feedback, continuously test the target processing system in the direction of improving the usage effect, and improve the usage effect of the target processing system.

[0062] The usage effect information can also be directly presented to the target user. Usually, the computing device can present the usage effect information to the target user, and the target user can adjust the usage method or process of the target processing system according to the usage effect information. In addition, the target user can also feedback usage suggestions to the computing device according to the usage effect information, and adjust the parameter instance of the proportional parameter through the usage suggestions feedback by the target user, so as to update the settings of the target processing system, improve the relevance between the target processing system and the user's usage habits, and provide more personalized system services.

[0063] In the embodiments of the present application, a target parameter instance corresponding to a proportional parameter in a target processing system is determined. By obtaining the usage effect information generated by the target user using the target processing system corresponding to the target parameter instance, based on this usage effect information, system feedback information generated by the target parameter instance in the target processing system can be generated. This system feedback information can be used to provide feedback to the target user. By automatically generating a target parameter instance and automatically evaluating the usage effect of the target parameter instance, the usage effect of the parameter instance can be quickly evaluated, the update efficiency of the proportional parameter can be improved, and effective update can be achieved.

[0064] In a possible design, the proportional parameter may include hyperparameters in the target processing system. A hyperparameter may be a parameter set for establishing the data processing system before the target processing system starts the calculation or learning process, rather than a model parameter used in the training process. Usually, a hyperparameter may include multiple sub-hyperparameters. And the types of hyperparameters can include multiple types. For example, the number of network layers of a deep network in a machine learning model, the learning rate of the model, etc. can both belong to two sub-hyperparameters among common hyperparameters. The proportional hyperparameter may include a hyperparameter in which the sum of the parameter ratios occupied by multiple sub-hyperparameters is 1. Assuming there are n sub-hyperparameters, the parameter values corresponding to the n sub-hyperparameters are a i , where a i ∈[0,1], In the embodiments of the present application, no excessive limitation is imposed on the specific type of the proportional parameter.

[0065] As Figure 2 shown, it is a flowchart of an embodiment of a data processing method provided by the embodiments of the present application. The method may include the following steps:

[0066] 201: Determine a system application request initiated by a target user for a target processing system.

[0067] Optionally, when the target processing system provides external services, the target user can use the external services provided by the target processing system. The user terminal can display the target processing system for the target user and detect the system application request initiated by the target user. The system application request may be initiated by the target user on the user terminal to the target processing system. After being detected by the user terminal, the system application request can be sent by the user terminal to a computing device configured with the data processing method provided by the embodiments of the present application.

[0068] 202: In response to the system application request, determine a target parameter instance corresponding to a proportional parameter in the target processing system.

[0069] 203: Obtain the usage effect information generated by the target user using the target processing system corresponding to the target parameter instance.

[0070] 204: Generate system feedback information generated by the target parameter instance in the target processing system based on the usage effect information.

[0071] Among them, the system feedback information is used to provide feedback to the target user.

[0072] Some steps in the embodiments of this application are the same as some steps in the foregoing embodiments. For the sake of simplicity of description, they will not be repeated here.

[0073] In the embodiments of this application, when determining the system application request initiated by the target user for the target processing system, the system application request can be responded to, and the target parameter instance corresponding to the proportional parameter in the target processing system can be determined. By obtaining the usage effect information generated by the target user using the target processing system corresponding to the target parameter instance, based on this usage effect information, the system feedback information generated by the target parameter instance in the target processing system can be generated. This system feedback information can be used to provide feedback to the target user. By automatically generating the target parameter instance and automatically evaluating the usage effect of the target parameter instance, the usage effect of the parameter instance can be quickly evaluated, the update efficiency of the proportional parameter can be improved, and effective update can be achieved.

[0074] As another embodiment, determining the target parameter instance corresponding to the proportional parameter in the target processing system may include:

[0075] When it is detected that the target processing system meets the update condition, determine the target parameter instance corresponding to the proportional parameter in the target processing system.

[0076] Optionally, the usage process of the target processing system can be detected. When the target processing system meets the update condition, the update of the target processing system can be started. At this time, a target parameter instance can be generated for the proportional parameter in the target processing system.

[0077] Optionally, when judging whether the target processing system meets the update condition, the historical usage information of the target user for the target processing system can be obtained, and the usage behavior of the target user can be analyzed according to the historical usage information. For example, the usage data such as the usage frequency, usage time, and click-through rate of the target user can be analyzed. When it is judged that the usage behavior of the target user does not meet the target behavior, it can be determined that the target processing system meets the update condition. In a possible design, the usage behavior of the target user not meeting the target behavior can specifically be that at least one of the multiple usage data of the target user does not meet the preset data threshold.

[0078] As an embodiment, after obtaining the usage effect information generated by the target user using the target processing system corresponding to the target parameter instance, the method may further include:

[0079] Based on the usage effect information of the target parameter instance, determine whether the target parameter instance meets the parameter usage conditions.

[0080] Optionally, determining whether the target parameter instance meets the parameter usage conditions may include: determining whether the usage effect information reaches a preset target effect. If so, determine that the target parameter instance meets the parameter usage conditions; if not, determine that the target parameter instance does not meet the parameter usage conditions. When the usage effect information is quantitative effect data such as click-through rate and number of clicks, whether the usage effect reaches the preset target effect may specifically include determining whether the effect data meets the preset effect threshold. Whether the effect data meets the effect threshold may, for example, include whether the effect data is greater than the effect threshold or whether the effect data is less than the effect threshold, which can be determined according to actual usage requirements.

[0081] Optionally, determining whether the target parameter instance meets the parameter usage conditions may include: determining whether the current iteration count reaches a preset iteration count threshold. If so, determine that the target parameter instance meets the parameter usage conditions; if not, determine that the target parameter instance does not meet the parameter usage conditions.

[0082] Optionally, the usage effect information may include the reward probabilities corresponding to at least one system sub-parameter. Whether the target parameter instance meets the parameter usage conditions may include: determining the usage effect information of the target parameter instance and the historical effect information of the historical parameter instance; calculating the expected value of the current cumulative regret using the usage effect information and the historical effect information. If the expected value of the cumulative regret meets the convergence condition, determine that the target parameter instance meets the parameter usage conditions. The expected value of the cumulative regret meeting the convergence condition may be that the expected value of the cumulative regret is minimized, for example, less than a certain expected threshold or less than all historical expectations.

[0083] At this time, based on the usage effect information, the system feedback information generated by the target parameter instance in the target processing system may include:

[0084] If the target parameter instance meets the parameter usage conditions, generate the system feedback information corresponding to the target parameter instance.

[0085] When the target parameter instance meets the usage conditions, the target processing system may be set based on the target parameter instance. At this time, the system feedback information may include the setting information of the target processing system.

[0086] In some embodiments, the method may further include:

[0087] If the target parameter instance does not meet the parameter usage conditions, based on the usage effect information corresponding to the target parameter instance, update the target parameter instance corresponding to the proportional parameter in the target processing system, and return to the step of obtaining the usage effect information generated by the target processing system when the target user uses the target parameter instance to continue execution.

[0088] When the target parameter instance does not meet the usage conditions, the target parameter instance can be adjusted continuously, and a parameter instance that meets the parameter usage conditions can be obtained, so that the obtained target parameter instance meets the parameter usage requirements.

[0089] Such as Figure 3 As shown, it is a flowchart of another embodiment of a data processing method provided by an embodiment of the present application. The method may include:

[0090] 301: Determine a system application request initiated by a target user for a target processing system.

[0091] Some steps in the embodiments of the present application are the same as some steps in the foregoing embodiments. For the sake of brevity of description, they will not be repeated here.

[0092] 302: In response to the system application request, determine a target parameter instance corresponding to a proportional parameter in the target processing system.

[0093] 303: Obtain usage effect information generated by the target processing system when the target user uses the target parameter instance.

[0094] 304: According to the usage effect information of the target parameter instance, determine whether the target parameter instance meets the parameter usage conditions.

[0095] 305: If the target parameter instance meets the parameter usage conditions, generate system feedback information corresponding to the target parameter instance. Among them, the system feedback information is used to feedback to the target user.

[0096] 306: If the target parameter instance does not meet the parameter usage conditions, update the target parameter instance corresponding to the proportional parameter based on the usage effect information; display the target processing system corresponding to the updated target parameter instance to the target user, and return to step 303.

[0097] Optionally, a black-box optimization algorithm can be used to update the target parameter instance to obtain an updated target parameter instance. Specifically, the usage effect information and the target parameter instance can be input into the black-box optimization algorithm to calculate and obtain the updated target parameter instance. The black-box optimization algorithm may include, for example, a Bayesian optimization algorithm.

[0098] Optionally, the adjustment information of the target parameter instance can be estimated based on the usage effect information corresponding to the target parameter instance and the historical effect information corresponding to the historical parameter instance, and the target parameter instance can be updated using the adjustment information. The target parameter instance includes parameter values corresponding to at least one system sub-parameter, and sub-adjustment information corresponding to at least one system sub-parameter can be obtained. The parameter values are adjusted using the sub-adjustment information corresponding to at least one system sub-parameter respectively to obtain the updated target parameter instance. The usage effect information may include reward probabilities corresponding to at least one system sub-parameter respectively, and the target parameter instance can be updated by the reward differences corresponding to at least one system sub-parameter between the historical parameter instance and the current target parameter instance. The reward difference of any one system sub-parameter may include the difference between the reward probability corresponding to the system sub-parameter in the target parameter instance and the historical reward probability corresponding to the historical parameter instance. In a possible design, the system sub-parameter with the largest positive reward difference between the target parameter instance and the historical parameter instance can be determined, and the adjustment value of this system sub-parameter is set higher than that of other system sub-parameters. The adjustment values of other system sub-parameters are also determined according to the magnitude of the reward difference. The parameter instance is adjusted in a timely manner according to the user usage situation, so that the adjustment direction has a higher correlation with the user, and efficient and accurate adjustment is achieved.

[0099] In the embodiments of the present application, when determining a system application request initiated by a target user for a target processing system, in response to the system application request, a target parameter instance corresponding to a proportional parameter in the target processing system can be determined, and thus the target processing system can be set according to the target parameter instance. When the target user uses the target processing system corresponding to the first parameter usage, a certain usage effect can be generated. The usage effect information generated by the target user using the target processing system corresponding to the target parameter instance can be obtained. Whether the target parameter instance meets the parameter usage condition can be judged through the usage effect information. If the target parameter instance meets the parameter usage condition, system feedback information of the target parameter instance can be generated. If the target parameter instance does not meet the parameter usage condition, the target parameter instance corresponding to the proportional parameter can be updated based on the usage effect information; the target processing system corresponding to the updated target parameter instance is displayed to the target user, so that the target user uses the new target processing system, and the step of obtaining the usage effect information generated by the target user using the target processing system corresponding to the target parameter instance is returned to continue execution. By continuously testing the target parameter instance of the proportional parameter, the target processing system is continuously corrected, so that the target processing system approaches the user's usage habit, the relevance between the target processing system and the user's usage is improved, automatic testing of the system is realized, and the testing efficiency and testing timeliness are improved.

[0100] The target parameter instance can be generated for the proportional parameter by a target generation algorithm.

[0101] As a possible implementation, in response to a system application request, determining a target parameter instance corresponding to a proportional parameter in a target processing system includes:

[0102] In response to a system application request, determining a target generation algorithm corresponding to a proportional parameter in a target processing system.

[0103] Generating a target parameter instance corresponding to a proportional parameter in a target processing system through the target generation algorithm.

[0104] The target generation algorithm may include a parameter generation algorithm. In a computing device, the target generation algorithm can be directly configured to generate a target parameter instance. Alternatively, algorithm generation services can also be provided to the computing device in the form of a program module or a third-party generation system. When the target generation algorithm is a third-party generation system or a program module, the computing device can initiate a parameter generation request to the third-party generation system or the program module, and the third-party generation system or the program module can, in response to the parameter generation request, use the target generation algorithm to generate a target parameter instance of the proportional parameter.

[0105] The usage effect information can be used to update the target generation algorithm for generating the target parameter instance, so that the target processing system corresponding to the target parameter instance is more in line with the user's usage habits, improving the accuracy and effectiveness of parameter setting.

[0106] In some embodiments, if the target parameter instance does not meet the parameter usage condition, based on the usage effect information corresponding to the target parameter instance, updating the target parameter instance corresponding to the proportional parameter in the target processing system and returning to the step of obtaining the usage effect information generated by the target user using the target processing system corresponding to the target parameter instance to continue execution includes:

[0107] If the target parameter instance does not meet the parameter usage condition, then update the target generation algorithm based on the usage effect information corresponding to the target parameter instance.

[0108] Regenerate the target parameter instance corresponding to the proportional parameter in the target processing system through the target generation algorithm.

[0109] Return to the step of obtaining the usage effect information generated by the target user using the target processing system corresponding to the target parameter instance to continue execution.

[0110] The target generation algorithm mainly estimates and ranks the future rewards that can be obtained for at least one system sub-parameter to obtain the probabilities of at least one system sub-parameter winning respectively, and uses the probabilities of at least one system sub-parameter winning respectively as the target parameter instance.

[0111] The usage effect information corresponding to the target parameter instance can be used as the estimation basis for the future rewards that can be obtained, improving the accuracy of the estimation result.

[0112] In the embodiments of the present application, the usage effect information corresponding to the target parameter instance is updated to the target generation algorithm, so as to generate a new target parameter instance through the target generation algorithm, realize the update of the parameter instance of the proportional parameter. By continuously iteratively updating the parameter instance of the proportional parameter, a more accurate parameter instance can be obtained, so as to set the target processing system according to the more accurate parameter instance, realize the online training of the target processing system, and obtain a target processing system with higher timeliness and usability.

[0113] The target generation algorithm can sequentially set parameter values for at least one system sub-parameter in the proportional parameter. In a possible design, the target generation algorithm can generate the target parameter instance corresponding to the proportional parameter in the target processing system in the following manner:

[0114] Determine at least one system sub-parameter corresponding to the proportional parameter in the target processing system;

[0115] Determine the target parameter instance composed of the parameter values respectively corresponding to at least one system sub-parameter.

[0116] The target parameter instance can be obtained by running the target generation algorithm.

[0117] In order to obtain accurate parameter instances, in some embodiments, the parameter values respectively corresponding to at least one system sub-parameter can be determined in the following manner:

[0118] Estimate the probabilities that the target user performs target operations on at least one system sub-parameter respectively, and obtain the trigger probabilities respectively corresponding to at least one system sub-parameter;

[0119] Determine that the trigger probabilities respectively corresponding to at least one system sub-parameter are the parameter values respectively corresponding to at least one system sub-parameter.

[0120] By estimating the probabilities that at least one system sub-parameter can respectively obtain the target user's execution of the target operation, the determination of the parameter values respectively corresponding to at least one system sub-parameter can be realized. The probabilities that at least one system sub-parameter can respectively obtain the target user's execution of the target operation are based on the probabilities that the target user performs target operations on the system information respectively corresponding to at least one system sub-parameter in a future period of time after setting the target processing system with the target parameter instance.

[0121] In order to make the correlation between the relevant information of the target user and the parameter instance of the proportional parameter closer and obtain more accurate parameter instances, as a possible implementation manner, estimating the probabilities that the target user performs target operations on at least one system sub-parameter respectively and obtaining the trigger probabilities respectively corresponding to at least one system sub-parameter includes:

[0122] Determine the user characteristics of the target user in the target processing system;

[0123] Estimate the probabilities that the target user performs target operations on at least one system sub-parameter respectively according to the user characteristics, and obtain the trigger probabilities respectively corresponding to at least one system sub-parameter.

[0124] The user characteristics may include the context characteristics corresponding to the system application request initiated by the target user, which are determined through the information of the target user. Adding the user characteristics to the probability estimation process can increase the accuracy of estimating the trigger probability.

[0125] The target operation is a related operation triggered by the target user for the target processing system. Taking the target processing system as an e-commerce platform as an example, the target user's execution of the target operation on the e-commerce platform may include the target user triggering the product page recommended by the e-commerce platform to the target user.

[0126] In the process of generating the target parameter instance of the proportional parameter, in addition to considering the relevant information of the target user, the parameter characteristics of each system sub-parameter in the proportional parameter can also be considered, so that the reference range of the parameter instance is more accurate. In some embodiments, the method may further include:

[0127] Obtain the parameter characteristics respectively corresponding to at least one system sub-parameter in the target processing system;

[0128] Estimating the probabilities that the target user performs target operations on at least one system sub-parameter respectively according to the user characteristics, and obtaining the trigger probabilities respectively corresponding to at least one system sub-parameter includes:

[0129] Estimate the probabilities that the target user performs target operations on at least one system sub-parameter respectively according to the user characteristics and the parameter characteristics respectively corresponding to at least one system sub-parameter, and obtain the trigger probabilities respectively corresponding to at least one system sub-parameter.

[0130] Using the user characteristics and the parameter characteristics respectively corresponding to at least one system sub-parameter to estimate the probabilities that the target user performs target operations on at least one system sub-parameter respectively increases the reference content of the probability estimation and improves the accuracy of the estimation result.

[0131] The usage effect information generated by the target user using the target processing system can actually be determined by the target operations performed by the user on the target processing system. Therefore, in a possible design, estimating the probabilities that the target user performs target operations on at least one system sub-parameter respectively according to the user characteristics and the parameter characteristics respectively corresponding to at least one system sub-parameter, and obtaining the trigger probabilities respectively corresponding to at least one system sub-parameter includes:

[0132] Estimate the triggering results of the target user performing a target operation on any one of at least one system sub-parameter according to the user characteristics and the parameter characteristics corresponding to at least one system sub-parameter respectively, and obtain N triggering results; wherein, any one of the triggering results is the target user performing a target operation on any one of at least one system sub-parameter.

[0133] Determine the triggering times corresponding to at least one system sub-parameter respectively according to the N triggering results.

[0134] Determine the triggering probabilities corresponding to at least one system sub-parameter respectively according to the ratios of the triggering times corresponding to at least one system sub-parameter respectively to N.

[0135] Wherein, N is a positive integer greater than 1. In order to obtain accurate triggering probabilities, N can be set to a relatively large value, such as 100, 200, or 500, etc. The N triggering results are respectively generated by the target user performing the target operation on any one of at least one system sub-parameter N times. The triggering times corresponding to at least one system sub-parameter respectively can be obtained by counting the number of times each system sub-parameter is triggered in the N triggering results.

[0136] In order to obtain the triggering results of the user performing a target operation on the target processing system, in a possible design, the triggering results are determined in the following manner:

[0137] Generate an online prediction model according to the user characteristics and the parameter characteristics corresponding to at least one system sub-parameter respectively.

[0138] Use the online prediction model to estimate the probabilities of the reward information obtained by at least one system sub-parameter respectively, and obtain the reward probabilities corresponding to at least one system sub-parameter respectively.

[0139] Determine the system sub-parameter with the highest reward probability as the system sub-parameter for the target user to perform the target operation according to the reward probabilities corresponding to at least one system sub-parameter respectively.

[0140] The reward information may include the target user performing a target operation on any one sub-parameter.

[0141] Optionally, the reward probabilities corresponding to at least one system sub-parameter can be sorted, and the system sub-parameter with the highest reward probability is obtained as the system sub-parameter for the target user to perform the target operation.

[0142] The online prediction model can include multiple types, and the model attributes of different online prediction models are different. The online prediction model can include an online processing model, such as UCB (The Upper Confidence Bound), TS (Tabu Search), LinUCB (Linear Upper Confidence Bound), or LinTS (Linear Thompson Sampling), etc., which can balance exploration and exploitation of parameter ratio values.

[0143] Therefore, in a possible design, when the online prediction model has random attribute information, using the online prediction model to estimate the probability of the reward information obtained by at least one system sub-parameter respectively, obtaining the reward probability corresponding to at least one system sub-parameter respectively includes:

[0144] Using the online prediction model, randomly perform multiple event samplings on the event that any one of at least one system sub-parameter is triggered, and obtain multiple sampling results; where any one sampling result is that any one of at least one system sub-parameter obtains the reward information of the target user;

[0145] According to the multiple sampling results, count the number of times that at least one system sub-parameter respectively obtains the reward information, so as to obtain the reward times corresponding to at least one system sub-parameter respectively;

[0146] According to the reward times corresponding to at least one system sub-parameter respectively, determine the reward probability corresponding to at least one system sub-parameter respectively.

[0147] The total reward times can be obtained by calculating the sum of the reward times corresponding to at least one system sub-parameter respectively, and the ratio of the reward times corresponding to at least one system sub-parameter respectively to the total reward times can be calculated to obtain the reward probability corresponding to at least one system sub-parameter respectively.

[0148] In addition, in another possible design, when the online prediction model does not have random attribute information, using the online prediction model to estimate the probability of the reward information obtained by at least one system sub-parameter respectively, obtaining the reward probability corresponding to at least one system sub-parameter respectively includes:

[0149] Using the online prediction model, estimate the reward scores corresponding to at least one system sub-parameter respectively;

[0150] According to the reward scores corresponding to at least one application sub-model respectively, generate a sampling distribution model corresponding to at least one system sub-parameter;

[0151] Using a sampling distribution model, sample to obtain the reward probabilities corresponding to at least one system sub-parameter respectively.

[0152] As a possible implementation, the sampling distribution model includes a uniform distribution model. Generating a sampling distribution model corresponding to at least one system sub-parameter respectively according to the reward scores corresponding to at least one application sub-model includes:

[0153] Normalize the reward scores corresponding to at least one application sub-model respectively to obtain the reward data corresponding to at least one application sub-model respectively;

[0154] Construct a uniform distribution model according to the reward data corresponding to at least one application sub-model respectively.

[0155] The uniform distribution model can perform effective sampling for simulating the user's execution of the target operation.

[0156] In some embodiments, generating an online prediction model according to the user characteristics and the parameter characteristics corresponding to at least one system sub-parameter respectively may include:

[0157] Obtain the historical reward information corresponding to at least one system sub-parameter respectively;

[0158] Generate an online prediction model according to the user characteristics and the parameter characteristics corresponding to at least one system sub-parameter respectively, in combination with the historical reward information corresponding to at least one system sub-parameter respectively.

[0159] Using the historical rewards, parameter characteristics, and user characteristics in the construction of the online prediction model can ensure that the online prediction model accurately models the operation results of the user's execution of the target operation, and use a more accurate model for the prediction of the use effect, improving the accuracy of the prediction.

[0160] In some embodiments, the target parameter instance is continuously updated. When the parameter instance is updated, the previous target parameter instance can become a historical parameter instance, and the historical parameter instance and the usage effect information of the historical parameter instance can also affect the determination of the usage effect information of the latest target parameter instance. As an embodiment, obtaining the usage effect information generated by the target user using the target processing system corresponding to the target parameter instance may include:

[0161] Obtain the historical parameter instance and the historical usage information corresponding to the historical parameter instance;

[0162] Generate the first sub-effect information corresponding to at least one system sub-parameter respectively according to the historical usage information corresponding to the historical parameter instance and the current usage information of the target parameter instance;

[0163] Determine the usage effect information constituted by the first sub-effect information corresponding to at least one system sub-parameter respectively.

[0164] Optionally, the usage effect information may include the analysis result of the usage information generated for the target user, and the analysis result may include the reward probability corresponding to at least one system sub-parameter respectively. The usage information may be that the user performs related operations on at least one system sub-parameter respectively, or at least one system sub-parameter obtains a reward. The analysis result in the usage effect information may be obtained by overall statistics of the current usage information and the historical usage information. For example, in a recommendation system, the system sub-parameters may include hyperparameters, and the effect information of different hyperparameters may represent the result ratios of different recall links. Suppose the overall feedback obtained currently is that the user clicks on a certain product, that is, the user obtains a reward, and the feedback is 1. Then, if the product comes from link 1, the effect feedback of hyperparameter 1 is 1, and the effect feedback of other hyperparameters is 0. By statistically analyzing the ratio of the number of times the user clicks on a certain hyperparameter to the total number of operations performed by the user, the reward probability corresponding to the system sub-parameter can be obtained. Generating the first sub-effect information corresponding to at least one system sub-parameter respectively according to the historical usage information corresponding to the historical parameter instance and the current usage information of the target parameter instance may specifically be to perform overall statistical analysis on the number of times of obtaining historical rewards and current rewards of at least one system sub-parameter respectively, so as to obtain the reward probability corresponding to at least one system sub-parameter respectively. The usage effect of the target parameter instance is marked by the reward probability, thereby obtaining the first sub-effect information corresponding to at least one system sub-parameter respectively.

[0165] In some embodiments, the system feedback information may be determined by the usage effect information of the parameter instance. Therefore, in some embodiments, if the target parameter instance meets the parameter usage condition, generating the system feedback information corresponding to the target parameter instance includes:

[0166] If the target parameter instance meets the parameter usage condition, generate the system setting information of the target processing system according to the target parameter instance;

[0167] The system feedback information is fed back to the target user in the following manner:

[0168] Display the target processing system corresponding to the system setting information to the target user for the target user to use the target processing system corresponding to the system setting information.

[0169] As a possible implementation manner, it further includes:

[0170] Determine at least one system sub-parameter corresponding to the ratio parameter in the target processing system;

[0171] Determine the system parameters corresponding to at least one system sub-parameter of the ratio parameter in the target processing system;

[0172] If the target parameter instance meets the parameter usage condition, generating the system setting information of the target processing system based on the target parameter instance includes:

[0173] If the target parameter instance meets the parameter usage condition, obtain the parameter values corresponding to the target parameter instance in at least one system sub-parameter respectively;

[0174] According to the parameter value corresponding to any one system sub-parameter, determine the system data of the system parameter corresponding to the system sub-parameter;

[0175] Determine the system setting information composed of the system data corresponding to at least one system parameter respectively.

[0176] In some embodiments, if the target parameter instance meets the parameter usage condition, the system feedback information generated corresponding to the target parameter instance may include:

[0177] If the target parameter instance meets the parameter usage condition, generate the parameter prompt information corresponding to the target parameter instance;

[0178] The system feedback information is fed back to the target user in the following manner:

[0179] Output the parameter prompt information for the target user for the target user to view the target parameter instance.

[0180] By generating and presenting the parameter prompt information corresponding to the target parameter instance for the target user, it can enable the target user to analyze the target parameter instance, so as to facilitate determining whether to use the target parameter instance or continue to perform the update of the parameter instance.

[0181] In the field of e-commerce, the target user can access the target processing system in the e-commerce field through a user device, such as a mobile phone, a tablet computer, a computer, a notebook, etc. terminal devices. At this time, the target processing system can include a network transaction system.

[0182] Reference Figure 4 , assuming that the target user uses the tablet computer M1 to use the network transaction system. At this time, the network transaction system can provide a network transaction page to the target user. The tablet computer M1 can include presenting the web page P1 provided by the network transaction system for the target user to the target user. The web page can include a recommendation page. When the target user browses the recommendation page of the target processing system, the target processing system can present the content recommended for the target user to the target user. The tablet computer M1 can detect the browsing operation triggered by the target user for the recommendation page, such as Figure 4 the sliding operation 401 shown in

[0183] In actual applications, since there are many types of transaction objects provided in the network transaction system, when providing the recommended objects of the recommendation page for the user, different transaction object sets can be used as a ratio parameter, and the sum of the recommended ratios corresponding to the multiple recommended objects can be 1, and the multiple recommended parameters can constitute a ratio parameter at this time. In the prior art, the ratio values ​​corresponding to the multiple recommended ratios are generally determined by historical experience and are fixed. Assuming that the technical solution of the embodiment of the present application is configured in the cloud server M2, when it is determined that the multiple recommended parameters constitute a ratio parameter, in the process of the user browsing the recommended page, it can be determined that the target user initiates a 402 system application request for the network transaction system, and in response to the system application request, the target parameter instance corresponding to the ratio parameter in the network transaction system can be determined 403. After that, the parameter values ​​corresponding to the multiple recommended parameters according to the target parameter instance can be re-determined 404 to display a new recommendation page to the target user, and sent 405 to the tablet computer M1. After the re-determined recommendation page is displayed to the target user, the use effect information can be generated, for example, the click operation performed by the target user on a certain recommended object S. The cloud server M2 can obtain 406 the use effect information generated by the network transaction system corresponding to this target parameter instance by the target user. Based on the use effect information, system feedback information generated by the target parameter instance in the target processing system is generated 407.

[0184] The system feedback information can be used to feed back to the target processing system for updating the target parameter instance of the proportional parameter. The system feedback information can be used to update the target parameter instance of the proportional parameter, and return to the step of determining the target parameter instance corresponding to the proportional parameter in the network transaction system to continue execution, thereby realizing online adjustment of the parameter instance of the proportional parameter, improving the timeliness of parameter updates, and matching the target parameter instance with the user's usage results, providing more personalized system applications, and improving the effectiveness of system use.

[0185] In the field of e-commerce, the use of online trading platforms is becoming more and more widespread. When any user accesses the online trading system, the online trading system displayed to the target user can be personalized and updated according to the user's usage information to improve the efficiency and effectiveness of the use of the online trading system.

[0186] like Figure 5 FIG. 1 is a flowchart of another embodiment of a data processing method provided in an embodiment of the present application. The method may include the following steps:

[0187] 501: Identify the system access request initiated by the target user to the network transaction system.

[0188] 502: In response to the system access request, determine a target parameter instance corresponding to the ratio parameter in the network transaction system.

[0189] 503: Obtain the usage effect information generated by the target user using the target parameter instance corresponding to the online trading system.

[0190] 504: Generate the system feedback information generated by the target parameter instance in the online trading system based on the usage effect information.

[0191] Among them, the system feedback information is used to feedback to the target user.

[0192] As an embodiment, determining the system access request initiated by the target user for the online trading system may include: determining the system access request initiated by the target user for the recommendation page of the online trading system.

[0193] Obtaining the usage effect information generated by the target user using the target parameter instance corresponding to the online trading system includes obtaining the browsing information and click information generated by the target user using the target parameter instance corresponding to the online trading system.

[0194] Generating the system feedback information generated by the target parameter instance in the online trading system based on the usage effect information may include: updating the target parameter instance corresponding to the proportional parameter based on the browsing information and click information of the target user; determining the parameter values corresponding to multiple recommendation parameters according to the updated target parameter instance; searching for the recommended objects corresponding to multiple recommendation parameters in the database according to the parameter values corresponding to multiple recommendation parameters; generating a new recommendation page based on the recommended objects corresponding to multiple recommendation parameters.

[0195] The system feedback information can be specifically fed back to the target user in the following way:

[0196] Output the new recommendation page for the target user through the online trading system for the target user to browse the new recommendation page. In practical applications, during the process of the target user browsing the new recommendation page, the usage effect information generated by the target user using the target parameter instance corresponding to the online trading system can also be re-obtained, continuously updating the target parameter instance of the proportional parameter, making the matching degree between the target parameter instance and the user's usage information higher, and obtaining a more accurate target parameter instance.

[0197] In an embodiment of the present application, during the process of a target user accessing a network trading system, a system access request initiated by the target user for the network trading system can be detected. At this time, in response to the system access request, a target parameter instance corresponding to a proportional parameter in the network trading system can be determined, so that usage effect information generated by the target user using the network trading system corresponding to the target parameter instance can be obtained. Based on the usage effect information, system feedback information generated by the target parameter instance in the network trading system can be generated. Through the usage information of the user, the network trading system presented to the target user can be personalized updated to improve the usage efficiency and effectiveness of the network trading system.

[0198] As an embodiment, determining a target parameter instance corresponding to a proportional parameter in a target processing system may include:

[0199] Determine multiple parameter generation algorithms corresponding to the proportional parameter in the target processing system;

[0200] Determine the target parameter instances generated by the multiple parameter generation algorithms for the proportional parameter respectively to obtain multiple target parameter instances;

[0201] Obtaining usage effect information generated by the target user using the target processing system corresponding to the target parameter instance may include:

[0202] Obtain the usage effect information generated by the target user using the target processing system corresponding to any one of the target parameter instances to obtain the usage effect information corresponding to the multiple target parameter instances respectively;

[0203] Based on the usage effect information, generating system feedback information generated by the target parameter instance in the target processing system may include:

[0204] Based on the usage effect information corresponding to the multiple target parameter instances respectively, determine the target parameter instance from the multiple target parameter instances; determine the parameter generation algorithm for generating the target parameter instance as the target generation algorithm.

[0205] As Figure 6 shown, it is a flowchart of another embodiment of a data processing method provided by an embodiment of the present application. The method may include the following steps:

[0206] 601: Determine multiple parameter generation algorithms corresponding to the proportional parameter in the target processing system.

[0207] Some steps in the embodiments of the present application are the same as some steps in the foregoing embodiments. For the sake of simplicity of description, they will not be repeated here.

[0208] 602: Determine the target parameter instances generated by the multiple parameter generation algorithms for the proportional parameter respectively to obtain multiple target parameter instances.

[0209] 603: Obtain the usage effect information generated by the target processing system using any target parameter instance for the target user, so as to obtain the usage effect information corresponding to multiple target parameter instances respectively.

[0210] 604: Based on the usage effect information corresponding to multiple target parameter instances respectively, determine the target parameter instance from multiple target parameter instances.

[0211] 605: Determine the parameter generation algorithm for generating the target parameter instance as the target generation algorithm.

[0212] 606: Generate algorithm prompt information in the target processing system according to the target generation algorithm.

[0213] 607: Display the algorithm prompt information to prompt the target generation algorithm that best matches the target user.

[0214] In the embodiments of the present application, for the target processing system, multiple parameter generation algorithms can be used to generate target parameter instances respectively. By evaluating the usage effects of multiple target parameter instances, the target generation algorithm is selected from multiple parameter generation algorithms to achieve the effective selection of the algorithm, obtain a generation algorithm with better generation effects, and then generate algorithm prompt information in the target processing system according to the target generation algorithm. Use the algorithm prompt information to prompt the target generation algorithm.

[0215] In addition to providing an optimization solution for the parameter instances of the online application system, a selection solution for the parameter generation algorithm in the target processing system can also be provided. In some embodiments, determining the system application request initiated by the target user for the target processing system may include:

[0216] Receive the algorithm selection request initiated by the optimization user;

[0217] In response to the algorithm selection request, generate the system application request initiated by the target user for the target processing system.

[0218] Optionally, determining the system application request initiated by the target user for the target processing system may further include: receiving the algorithm selection request initiated by the optimization user; in response to the algorithm selection request, simulate and generate the system application request initiated by the target user for the target processing system through the request simulation module. By providing multiple parameter generation algorithms for the optimization user and automatically testing multiple parameter generation algorithms, the target generation algorithm is obtained.

[0219] When optimizing the service for users to select a parameter generation algorithm, the target processing system may include an offline application system. The optimizing users may include those who use the algorithm selection service, and the optimizing users may initiate an algorithm selection request. At this time, the computing device may, in response to the algorithm selection request initiated by the optimizing user, simulate and generate a system application request initiated by the target user for the target processing system through a request simulation module. Displaying algorithm prompt information to prompt the target generation algorithm that best matches the target user may specifically include displaying algorithm prompt information to the optimizing user to prompt the target generation algorithm that best matches the target user.

[0220] Since the target processing system is an offline online application system at this time and the system application request corresponding to the target user is also simulated and generated, in order to accurately obtain the usage effect information of the target user for the target processing system, in some embodiments, a simulation system may be established for the target processing system to simulate the usage effect information generated by the target user in the target processing system.

[0221] Therefore, as an embodiment, the usage effect information generated by the target user using the target processing system corresponding to any target parameter instance is obtained through the following methods:

[0222] Generate a simulation system of the target processing system;

[0223] Based on the simulation system, simulate the simulation effect information generated by the target user using the target processing system;

[0224] Determine the usage effect information according to the target parameter instance and the simulation effect information.

[0225] When using the simulation system to simulate the usage effect of the user, in some embodiments, the method may further include:

[0226] Determine at least one system sub-parameter corresponding to the proportional parameter in the target processing system;

[0227] Based on the simulation system, simulating the simulation effect information generated by the target user using the target processing system includes:

[0228] Based on the simulation system, simulate the effect evaluation values generated by the target user using the target processing system for each of the at least one system sub-parameter.

[0229] Determining the usage effect information according to the target parameter instance and the simulation effect information includes:

[0230] According to the effect evaluation values corresponding to each of the at least one system sub-parameter and the parameter values corresponding to the target parameter instance for each of the at least one system sub-parameter, determine the second sub-effect information corresponding to each of the at least one system sub-parameter.

[0231] Determine the usage effect information composed of the second sub-effect information corresponding to at least one system sub-parameter respectively.

[0232] The simulation system can exist in the form of a probability distribution model. As a possible implementation, the simulation system for generating the target processing system includes:

[0233] Generate probability distribution models for at least one system sub-parameter respectively;

[0234] Based on the simulation system, when simulating the target user using the target processing system, the effect evaluation values generated by at least one system sub-parameter respectively include:

[0235] According to the probability distribution model corresponding to any one system sub-parameter, randomly perform multiple event samplings on the event of the system sub-parameter being triggered to obtain the effect evaluation value corresponding to the system sub-parameter, so as to obtain the effect evaluation values corresponding to at least one system sub-parameter respectively.

[0236] Optionally, the probability distribution models corresponding to at least one system sub-parameter respectively belong to a joint probability distribution model. Through the joint probability distribution model, the influence degree between at least one system sub-parameter can be quantified to achieve an accurate estimation of the influence effect.

[0237] In order to obtain accurate effect information, in a possible design, determining the second sub-effect information corresponding to at least one system sub-parameter respectively according to the effect evaluation values corresponding to at least one system sub-parameter respectively, and the parameter values corresponding to the target parameter instance in at least one system sub-parameter respectively may include:

[0238] According to the effect evaluation values corresponding to at least one system sub-parameter respectively, and the parameter values corresponding to the target parameter instance in at least one system sub-parameter respectively, multiply the effect evaluation value and the parameter value of any one system sub-parameter to obtain the second sub-effect information of the system sub-parameter, so as to obtain the second sub-effect information corresponding to at least one system sub-parameter respectively.

[0239] After presenting the system feedback information to the target user, if the feedback effect of the target user on the system feedback information is not satisfactory enough, a modification operation on the proportional parameter of the target processing system can be initiated. As an embodiment, after the system feedback information is fed back to the target user, the method may further include:

[0240] Detect a parameter adjustment request initiated by the target user for the proportional parameter;

[0241] Respond to the parameter adjustment request and obtain the parameter adjustment information provided by the target user;

[0242] Based on the parameter adjustment information, adjust the target parameter instance of the proportional parameter to obtain the adjusted target parameter instance.

[0243] After the target user adjusts the parameter instance of the proportional parameter, the target processing system corresponding to the adjusted parameter instance can be presented to the target user for the target user to use.

[0244] Therefore, in a possible design, after adjusting the parameter instance of the proportional parameter based on the parameter adjustment information and obtaining the expected parameter instance of the target user, the method may further include:

[0245] Generating expected setting information for the target processing system according to the second parameter instance;

[0246] Presenting the target processing system corresponding to the expected setting information to the target user who uses the target processing system for the target user to use the target processing system corresponding to the expected setting information.

[0247] Since different users have different usage habits, usage effects, user identities, and content of interest for the target processing system, when presenting system feedback information to the target user, the display type of the system feedback information can be determined according to the attribute information of the target user, so as to facilitate personalized display and improve the display effect. Therefore, as another embodiment, the method may further include:

[0248] Obtaining the user attribute information of the target user;

[0249] The system feedback information is fed back to the target user in the following manner:

[0250] Based on the user attribute information, selecting the target feedback type that best matches the target user from a preset plurality of feedback types;

[0251] Generating system display information corresponding to the system feedback information according to the target feedback type;

[0252] Presenting the system display information to the information target through the target processing system.

[0253] The user attribute information may include information associated with the target user itself, such as the historical usage effect information, user identity information, work type information, marked information, favorite information, etc. of the target user.

[0254] Determining the feedback type of the system feedback information through the user attribute information can make the feedback effect have a higher correlation with the user and higher feedback effectiveness.

[0255] For the sake of easy understanding, taking the target processing system as an e-commerce platform as an example, an application example of the embodiments of the present application is introduced in detail.

[0256] Reference Figure 7, in practical applications, user devices, such as mobile phone terminals, Internet of Things (IoT) terminals, etc., can interact with users, and user devices can communicate with an e-commerce platform. Taking the user device as the mobile phone terminal M3 and the e-commerce platform as the server M4 as an example. The mobile phone terminal M3 can detect a browsing request triggered by the target user for the e-commerce platform. This browsing request can be sent 701 by the mobile phone terminal M3 to the server M3 of the e-commerce platform as a system application request. The server M3 can determine 702 the browsing request initiated by the target user for the target processing system.

[0257] After that, the server M4 of the e-commerce platform can, in response to this browsing request, determine 703 the target parameter instance corresponding to the proportionality parameter of the products to be recommended to the user in the e-commerce platform. The proportionality parameter can include three system sub-parameters. These three system sub-parameters in the e-commerce platform can respectively include, for example: the first system sub-parameter corresponding to the ratio of the product type to the products browsed by the user historically, the second system sub-parameter corresponding to the ratio of the latest products sorted by the online time, and the third system sub-parameter corresponding to the ratio of the products filtered by the filtering technology. Assuming that the value of the first system sub-parameter is the first ratio, the value of the second system sub-parameter is the second ratio, and the value of the third system sub-parameter is the third ratio, the sum of the first ratio, the second ratio, and the third ratio is 1.

[0258] The target parameter instance can include the first ratio of the first system sub-parameter, the second ratio of the second system sub-parameter, and the third ratio of the third system sub-parameter. The e-commerce platform obtains 704 the products according to the target parameter instance, determines the first quantity of recommended products related to the products browsed by the user historically according to the first ratio, determines the second quantity of products sorted by the online time according to the second ratio, and determines the third quantity of products filtered by the filtering technology according to the third ratio. After that, the server M4 sends 705 the products to the mobile phone M3. The mobile phone M3 displays 706 the products for the target user.

[0259] The target user views the products recommended by the e-commerce platform and performs a target operation on the interested products, such as clicking, to obtain a trigger result. The trigger result 707 can include the usage effect information generated by the target user using the target processing system corresponding to the target parameter instance. The e-commerce platform can obtain this trigger result.

[0260] After that, the server M3 of the e-commerce platform can generate system feedback information generated by the 708 target parameter instance on the e-commerce platform based on the trigger result. This system feedback information is used to provide feedback to the target user. For example, this system feedback information is to update the 703 target parameter instance based on the trigger result, and re-obtain the products recommended by the e-commerce platform to the user through the target parameter instance.

[0261] The data processing method provided by the embodiments of the present application can be applied to different fields. To improve the application efficiency, the data processing method provided by the embodiments of the present application can be encapsulated and then configured in a cloud server so that multiple users can simultaneously initiate data processing requests to achieve the effective application of the technical solution.

[0262] As Figure 8 shown, it is a flowchart of another embodiment of a data processing method provided by the embodiments of the present application. The method may include:

[0263] 801: In response to a usage request for calling a data processing interface, determine the processing resources corresponding to the data processing interface.

[0264] Execute the following steps using the processing resources corresponding to the data processing interface:

[0265] 802: Determine the target parameter instance corresponding to the proportional parameter in the target processing system.

[0266] 803: Obtain the usage effect information generated by the target user using the target processing system corresponding to the target parameter instance.

[0267] 804: Generate system feedback information generated by the target parameter instance in the target processing system based on the usage effect information.

[0268] Among them, the system feedback information is used to provide feedback to the target user.

[0269] The specific steps executed by the processing resources corresponding to the data processing interface in the embodiments of the present application are the same as the processing steps executed by the Figure 1 shown data processing method. The specific implementation manners and technical effects of each step have been described in detail in the Figure 1 shown embodiments and will not be elaborated here.

[0270] In practical applications, the data processing method provided by the embodiments of the present application can be applied to various fields. For example, it can be applied to fields such as artificial intelligence interaction, data retrieval, content recommendation, click-through rate prediction, intelligent factory, industrial control, e-commerce field, video live broadcast field, social field, online education field, etc.

[0271] For ease of understanding, the following several actual usage fields will be used to introduce the embodiments of the present application in detail.

[0272] (1) E-commerce field. In the e-commerce field, application scenarios such as feature search, product recommendation in live broadcast scenarios, content recommendation, and calculation of advertising click-through rate are the most common. This embodiment takes the content recommendation scenario as an example and deploys an example. The general recommendation process in the recommendation scenario may include parameterizing the elements of the selected scenario, obtaining multiple system sub-parameters that affect the scenario, using multiple parameters to identify different features of the scenario, and obtaining a parameter instance by assigning features to multiple system sub-parameters.

[0273] Taking the click word recommendation scenario as an example, when the user clicks the search box in the APP (Application), the target processing system will recommend some search words (Query words) to the user. The purpose of recommending search words to users is to tap into the user's potential purchasing needs, increase the user's stickiness and increase the total number of commodity transactions. In the search system, using the technical solution of this case, the ratio corresponding to the search word types can be set as a ratio parameter. Different types of search words can correspond to different ratios. For example, assuming that 6 search words are recommended to the user, clothing search words can account for half of the 6 search words, beauty search words can account for a quarter, and maternal and child products can account for a quarter. When recommending search words to users, target parameter instances corresponding to multiple product types can be determined. The search words of multiple product types corresponding to the target parameter instance are displayed to the user. After obtaining the usage results generated by the target processing system corresponding to the user's use of the search words of the multiple product types, the search words can be retrieved based on the usage results and fed back to the target user.

[0274] (2) Social field: In the social field, it is also common to recommend content to social users and courses to students.

[0275] Recommendations in the social field usually involve social users browsing social applications, and the display interface of the application shows social content that the users are interested in. Generally, recommendations in the social field are usually based on feature parameters composed of options such as the user's historical browsing behavior, areas of interest, and user information. Different combinations of options can form different parameters. The sum of the proportions of different parameters in the recommended content is 1. When determining the parameters, the content related to the parameters can be determined, and by continuously optimizing the proportions of various types of parameters, accurate social content recommended to users can be obtained. The technical solution of the embodiment of this application can be configured in a social server. During the process of users browsing the social network, a system application request can be triggered. In response to this system application request, a target parameter instance composed of the proportions of different parameters in the social network can be determined. The social network can, based on the proportions of different parameters corresponding to the target parameter instance, search for relevant recommended content for the target user from the background, and detect the click results or browsing time of the target user on these recommended content to generate usage effect information of the target user. The usage effect information can be used to generate system feedback information generated by the target parameter instance in the target processing system. This system feedback information can, for example, include the content that the user is concerned about obtained by analyzing the usage effect information, and feedback the content that the user is concerned about to the social network so that the design network can update the proportions of different parameters based on the content that the user is concerned about. Through continuous iteration, the content recommended by the social network to the target user can have a higher matching degree with the user. (3) In the financial field, quantitative portfolio is a very important investment strategy issue. In the portfolio problem, the most common case is that when providing investment strategies for users in an investment processing system, it can include the user configuring multiple investment targets to make the portfolio return controllable. The portfolio return can be modeled as a process of adjusting a proportional value hyperparameter. Existing methods rely on the analysis of multi-factor investment variables and combine the personal experience of managers for judgment. The method of this patent can, on this basis, automatically dynamically adjust the proportional hyperparameters involved.

[0276] Using the technical solution provided by the embodiments of the present application, the multi-factor investment variable can be used as a proportional parameter. In the actual application process, when determining that the multi-factor investment variable is a proportional parameter, during the process of the target user consulting an investment strategy, the target user, that is, the investment strategy acquisition request initiated by the manager, can be determined, and the target parameter instance corresponding to the proportional parameter in the investment model can be determined. An investment model is constructed through the target parameter instance. During the process of the target user using the investment model to obtain an investment strategy, the usage effect information generated by the target user using the investment model corresponding to the target parameter instance can be obtained. The usage effect information can, for example, include receiving the investment strategy generated by the investment model or not receiving the investment strategy generated by the investment model. Through the usage effect information, the system feedback information generated by the target parameter instance in the investment model can be generated. This system feedback information can be used to update the target parameter instance. A new investment model is established through the updated target parameter instance. By continuously iterating the parameter instances and establishing the corresponding investment models to generate investment strategies, the parameter instances of the proportional parameter are updated online to achieve accurate adjustment of the proportional parameter.

[0277] (4) In the field of resource allocation, taking the allocation of electric power resources as an example. The allocation of electric power resources usually involves many regions, and each region can be represented by corresponding parameters. These parameters can respectively allocate a certain proportion of resources, and the allocation of resources will affect information such as regional economy, population, and environment.

[0278] The technical solution of the embodiments of the present application can be applied to the problems of dynamic pricing in the electricity market and economic load allocation in the electricity system. Below, a detailed description will be mainly made on the specific application fields of the power system.

[0279] In the problem of power economic load distribution, the power supply side can supply power resources to multiple regions simultaneously, and the power load capacity of each region accounts for a certain proportion of the power supply. The proportions of the power supply accounted for by multiple regions respectively can be used as a proportion parameter. By using the technical solution of the embodiment of the present application, the target user can include the management side of power resources, and the target processing system can include a calculation system for calculating the total power consumption of the power grid when supplying power to multiple regions. The management side of power resources can initiate a calculation request for the total power consumption of the power grid to the calculation system. Subsequently, in response to the calculation request, the ratios of the power load capacities of multiple regions to the power supply are set respectively to obtain a target parameter instance. After inputting the ratios of the power load capacities of each region in the target parameter instance to the power supply into the calculation system, the obtained total power consumption of the power grid can include the generated usage effect. The target parameter instance can be evaluated for its effect through the total power consumption of the power grid, thereby generating system feedback information generated by the target parameter instance in the calculation system. The system feedback information can, for example, include a prompt message of the total power consumption of the power grid, and this prompt message can be displayed to the management side of power resources so that the management side of power resources can set the power load capacities of multiple regions according to this prompt message. By adopting the technical solution of the embodiment of the present application, the power load capacities of each region can be automatically tested, the parameters can be set efficiently, and the setting effect can be improved.

[0280] (5) In the field of course recommendation, taking the target processing system as the course recommendation system as an example for detailed description.

[0281] The recommendation of online courses is usually that during the process of students or parents browsing the web page, the application program displays the target courses recommended for users on the displayed page. To improve the effectiveness of the recommendation, different factors such as the identity information, concerned fields, and historical courses of parents or students can be parameterized to obtain multiple recommendation factors, and the recommended ratios obtained by respectively taking the values of the multiple recommendation factors can be used as proportion parameters for optimization. During the optimization process, the target parameter instance corresponding to the proportion parameter in the course recommendation system can be determined first. The target parameter instance is also the proportion corresponding to each of the multiple recommendation factors. Through the target parameter instance, the target courses found by the course recommendation system are displayed to the target user. During the process of the target user browsing the target courses, it can be detected whether the target user views the target courses or performs a purchase operation on the target courses to generate usage effect information generated by the target user for the course recommendation information. This usage effect information can be used to generate system feedback information of the target processing system, and the system feedback information can be used to update the target parameter instance and return to continue executing the target parameter instance corresponding to the proportion parameter in the course recommendation system. By continuously adjusting the proportion parameters in the course recommendation system, the target parameter instance can be made more compatible with the user's usage behavior, the accuracy of the target parameter instance can be improved, and the effectiveness of the course recommendation can be promoted.

[0282] Such asFigure 9 As shown in the figure, it is a schematic structural diagram of an embodiment of a computing device provided by an embodiment of the present application. The device may include: a storage component 901 and a processing component 902; the storage component 901 is used to store one or more computer instructions, and the one or more computer instructions are called by the processing component;

[0283] The processing component 902 is used to:

[0284] Determine a target parameter instance corresponding to a proportional parameter in the target processing system; obtain usage effect information generated by the target user using the target processing system corresponding to the target parameter instance; generate system feedback information generated by the target parameter instance in the target processing system based on the usage effect information; wherein, the system feedback information is used to feedback to the target user.

[0285] As an embodiment, the processing component determining a target parameter instance corresponding to a proportional parameter in the target processing system may include:

[0286] Determine a system application request initiated by the target user for the target processing system;

[0287] In response to the system application request, determine a target parameter instance corresponding to a proportional parameter in the target processing system.

[0288] In some embodiments, the processing component determining a target parameter instance corresponding to a proportional parameter in the target processing system may include:

[0289] When detecting that the target processing system meets the update condition, determine a target parameter instance corresponding to a proportional parameter in the target processing system.

[0290] As an embodiment, the processing component may also be used to:

[0291] Judge whether the target parameter instance meets the parameter usage condition according to the usage effect information of the target parameter instance;

[0292] The processing component generating system feedback information generated by the target parameter instance in the target processing system based on the usage effect information may specifically include:

[0293] If the target parameter instance meets the parameter usage condition, generate system feedback information corresponding to the target parameter instance.

[0294] In some embodiments, the processing component may also be used to:

[0295] If the target parameter instance does not meet the parameter usage condition, update the target parameter instance corresponding to the proportional parameter based on the usage effect information; display the target processing system corresponding to the updated target parameter instance to the target user, and jump to continue executing to obtain the usage effect information generated by the target user using the target processing system corresponding to the target parameter instance.

[0296] In a possible design, for the processing component to determine the target parameter instance corresponding to the proportional parameter in the target processing system in response to a system application request may include:

[0297] In response to a system application request, determine the target generation algorithm corresponding to the proportional parameter in the target processing system;

[0298] Generate the target parameter instance corresponding to the proportional parameter in the target processing system through the target generation algorithm.

[0299] As a possible implementation, if the target parameter instance does not meet the parameter usage conditions, the processing component updates the target parameter instance corresponding to the proportional parameter in the target processing system based on the usage effect information corresponding to the target parameter instance, and returns to obtain the usage effect information generated by the target processing system when the target user uses the target parameter instance and continues to execute, which may specifically include:

[0300] If the target parameter instance does not meet the parameter usage conditions, then update the target generation algorithm based on the usage effect information corresponding to the target parameter instance;

[0301] Regenerate the target parameter instance corresponding to the proportional parameter in the target processing system through the target generation algorithm;

[0302] Return to the step of obtaining the usage effect information generated by the target processing system when the target user uses the target parameter instance and continue to execute.

[0303] In some embodiments, the processing component generates the target parameter instance corresponding to the proportional parameter in the target processing system by the following method:

[0304] Determine at least one system sub-parameter corresponding to the proportional parameter in the target processing system;

[0305] Determine the target parameter instance composed of the parameter values corresponding to at least one system sub-parameter respectively.

[0306] In a possible design, the processing component determines the parameter values corresponding to at least one system sub-parameter respectively by the following method:

[0307] Estimate the probabilities of the target user performing target operations on at least one system sub-parameter respectively, and obtain the trigger probabilities corresponding to at least one system sub-parameter respectively;

[0308] Determine that the trigger probabilities corresponding to at least one system sub-parameter respectively are the parameter values corresponding to at least one system sub-parameter respectively.

[0309] As an example, the processing component estimates the probabilities that the target user performs target operations on at least one system sub-parameter respectively, and obtaining the trigger probabilities corresponding to at least one system sub-parameter respectively may specifically include:

[0310] Determine the user characteristics of the target user in the target processing system;

[0311] Based on the user characteristics, estimate the probabilities that the target user performs target operations on at least one system sub-parameter respectively, and obtain the trigger probabilities corresponding to at least one system sub-parameter respectively.

[0312] In some embodiments, the processing component can also be used for:

[0313] Obtain the parameter characteristics corresponding to at least one system sub-parameter in the target processing system;

[0314] The processing component estimates the probabilities that the target user performs target operations on at least one system sub-parameter respectively, and obtaining the trigger probabilities corresponding to at least one system sub-parameter respectively may specifically include:

[0315] Based on the user characteristics and the parameter characteristics corresponding to at least one system sub-parameter respectively, estimate the probabilities that the target user performs target operations on at least one system sub-parameter respectively, and obtain the trigger probabilities corresponding to at least one system sub-parameter respectively.

[0316] In some embodiments, the processing component estimates the probabilities that the target user performs target operations on at least one system sub-parameter respectively, and obtaining the trigger probabilities corresponding to at least one system sub-parameter respectively may specifically include:

[0317] Based on the user characteristics and the parameter characteristics corresponding to at least one system sub-parameter respectively, estimate the trigger results of N times of the target user performing target operations on any one of at least one system sub-parameter, and obtain N trigger results; wherein, any one trigger result is the target user performing target operations on any one of at least one system sub-parameter;

[0318] Based on the N trigger results, determine the trigger times corresponding to at least one system sub-parameter respectively;

[0319] Based on the ratio of the trigger times corresponding to at least one system sub-parameter respectively to N, determine the trigger probabilities corresponding to at least one system sub-parameter respectively.

[0320] As a possible implementation, the processing component determines the trigger result through the following method:

[0321] Based on the user characteristics and the parameter characteristics corresponding to at least one system sub-parameter respectively, generate an online prediction model;

[0322] Using an online prediction model, predict the probabilities of the reward information obtained by at least one system sub-parameter respectively, and obtain the reward probabilities corresponding to at least one system sub-parameter respectively;

[0323] According to the reward probabilities corresponding to at least one system sub-parameter respectively, determine that the system sub-parameter with the highest reward probability is the system sub-parameter for the target user to perform the target operation.

[0324] In a possible design, the processing component's online prediction model has random attribute information; using the online prediction model, predicting the probabilities of the reward information obtained by at least one system sub-parameter respectively, and obtaining the reward probabilities corresponding to at least one system sub-parameter respectively may specifically include:

[0325] Using the online prediction model, randomly perform multiple event samplings on the event that any one of at least one system sub-parameter is triggered, and obtain multiple sampling results; wherein, any one sampling result is that any one of at least one system sub-parameter obtains the reward information of the target user;

[0326] According to the multiple sampling results, count the number of times that at least one system sub-parameter respectively obtains the reward information, so as to obtain the reward times corresponding to at least one system sub-parameter respectively;

[0327] According to the reward times corresponding to at least one system sub-parameter respectively, determine the reward probabilities corresponding to at least one system sub-parameter respectively.

[0328] In another possible design, the online prediction model does not have random attribute information; the processing component uses the online prediction model to predict the probabilities of the reward information obtained by at least one system sub-parameter respectively, and obtaining the reward probabilities corresponding to at least one system sub-parameter respectively may specifically include:

[0329] Using the online prediction model, predict the reward scores corresponding to at least one system sub-parameter respectively;

[0330] According to the reward scores corresponding to at least one application sub-model respectively, generate a sampling distribution model corresponding to at least one system sub-parameter;

[0331] Using the sampling distribution model, sample to obtain the reward probabilities corresponding to at least one system sub-parameter respectively.

[0332] In some embodiments, the sampling distribution model includes a uniform distribution model. The processing component generating a sampling distribution model corresponding to at least one system sub-parameter respectively according to the reward scores corresponding to at least one application sub-model respectively may specifically include:

[0333] Normalize the reward scores corresponding to at least one application sub-model to obtain the reward data corresponding to at least one application sub-model respectively;

[0334] Construct a uniform distribution model according to the reward data corresponding to at least one application sub-model respectively.

[0335] As an embodiment, the processing component generates an online prediction model according to the user characteristics and the parameter characteristics corresponding to at least one system sub-parameter, which may specifically include:

[0336] Obtain the historical reward information corresponding to at least one system sub-parameter respectively;

[0337] Generate an online prediction model according to the user characteristics, the parameter characteristics corresponding to at least one system sub-parameter, and the historical reward information corresponding to at least one system sub-parameter respectively.

[0338] In some embodiments, the processing component obtains the usage effect information generated by the target processing system when the target user uses the target parameter instance, which may specifically include:

[0339] Obtain the historical parameter instance and the historical usage information corresponding to the historical parameter instance;

[0340] Generate the first sub-effect information corresponding to at least one system sub-parameter respectively according to the historical usage information corresponding to the historical parameter instance and the current usage information of the target parameter instance;

[0341] Determine the usage effect information composed of the first sub-effect information corresponding to at least one system sub-parameter respectively.

[0342] As another embodiment, when the processing component processes that the target parameter instance meets the parameter usage condition, generating the system feedback information corresponding to the target parameter instance may specifically include:

[0343] If the target parameter instance meets the parameter usage condition, generate the system setting information of the target processing system according to the target parameter instance;

[0344] The processing component feeds back the system feedback information to the target user in the following manner:

[0345] Display the target processing system corresponding to the system setting information to the target user for the target user to use the target processing system corresponding to the system setting information.

[0346] In some embodiments, the processing component can also be used for:

[0347] Determine at least one system sub-parameter corresponding to the proportional parameter in the target processing system;

[0348] At least one system sub-parameter for determining a proportional parameter is a system parameter corresponding to a target processing system;

[0349] If a target parameter instance meets the parameter usage condition, the processing component processes to generate system setting information for the target processing system based on the target parameter instance. Specifically, it may include:

[0350] If a target parameter instance meets the parameter usage condition, obtain the parameter values corresponding to the target parameter instance in at least one system sub-parameter respectively;

[0351] Based on the parameter value corresponding to any one system sub-parameter, determine the system data of the system parameter corresponding to the system sub-parameter;

[0352] Determine the system setting information composed of the system data corresponding to at least one system parameter respectively.

[0353] In some embodiments, if a target parameter instance meets the parameter usage condition, the processing component processes to generate system feedback information corresponding to the target parameter instance. Specifically, it may include:

[0354] If a target parameter instance meets the parameter usage condition, generate parameter prompt information corresponding to the target parameter instance;

[0355] The system feedback information is fed back to the target user in the following ways:

[0356] Output parameter prompt information for the target user so that the target user can view the target parameter instance.

[0357] As an embodiment, the target parameter instance corresponding to the proportional parameter in the target processing system determined by the processing component may include:

[0358] Determine multiple parameter generation algorithms corresponding to the proportional parameter in the target processing system;

[0359] Determine the target parameter instances generated by the multiple parameter generation algorithms for the proportional parameter respectively to obtain multiple target parameter instances;

[0360] The processing component obtains the usage effect information generated by the target user using the target processing system corresponding to the target parameter instance. Specifically, it may include:

[0361] Obtain the usage effect information generated by the target user using the target processing system corresponding to any one target parameter instance to obtain the usage effect information corresponding to each of the multiple target parameter instances;

[0362] Based on the usage effect information, the processing component generates system feedback information generated by the target parameter instance in the target processing system. Specifically, it may include:

[0363] Determine a target parameter instance from multiple target parameter instances based on the usage effect information corresponding to each of the multiple target parameter instances;

[0364] Generate algorithm prompt information in the target processing system according to the target generation algorithm;

[0365] The processing component feeds back system feedback information to the target user in the following way:

[0366] Display the algorithm prompt information for the target user so that the target user can view the parameter generation algorithm corresponding to the target parameter instance.

[0367] In a possible design, the processing component obtains the usage effect information generated by the target processing system when the target user uses any target parameter instance in the following way:

[0368] Generate a simulation system of the target processing system;

[0369] Based on the simulation system, simulate the simulation effect information generated when the target user uses the target processing system;

[0370] Determine the usage effect information according to the target parameter instance and the simulation effect information.

[0371] In some embodiments, the processing component can also be used for:

[0372] Determine at least one system sub-parameter corresponding to the proportional parameter in the target processing system;

[0373] The processing component simulates the simulation effect information generated when the target user uses the target processing system based on the simulation system, which specifically may include:

[0374] Based on the simulation system, simulate the effect evaluation values generated by the target user when using the target processing system for each of at least one system sub-parameter;

[0375] The processing component determines the usage effect information according to the target parameter instance and the simulation effect information, which specifically may include:

[0376] According to the effect evaluation values corresponding to each of at least one system sub-parameter, and the parameter values corresponding to the target parameter instance for each of at least one system sub-parameter, determine the second sub-effect information corresponding to each of at least one system sub-parameter;

[0377] Determine the usage effect information composed of the second sub-effect information corresponding to each of at least one system sub-parameter.

[0378] In some embodiments, the processing component generates a simulation system of the target processing system, which specifically may include:

[0379] Generate probability distribution models for at least one system sub-parameter respectively;

[0380] When the processing component is based on the simulation system and simulates the target user using the target processing system, the effect evaluation values respectively generated at at least one system sub-parameter may specifically include:

[0381] According to the probability distribution model corresponding to any one system sub-parameter, randomly perform multiple event samplings on the event of the system sub-parameter being triggered to obtain the effect evaluation value corresponding to the system sub-parameter, so as to obtain the effect evaluation values respectively corresponding to at least one system sub-parameter.

[0382] In a possible design, when the processing component determines the second sub-effect information respectively corresponding to at least one system sub-parameter according to the effect evaluation values respectively corresponding to at least one system sub-parameter and the parameter values of the target parameter instance respectively corresponding to at least one system sub-parameter, it may specifically include:

[0383] According to the effect evaluation values respectively corresponding to at least one system sub-parameter and the parameter values of the target parameter instance respectively corresponding to at least one system sub-parameter, multiply the effect evaluation value and the parameter value of any one system sub-parameter to obtain the second sub-effect information of the system sub-parameter, so as to obtain the second sub-effect information respectively corresponding to at least one system sub-parameter.

[0384] As another embodiment, when the processing component determines the system application request initiated by the target user for the target processing system, it may specifically include:

[0385] Receive the algorithm selection request initiated by the optimization user;

[0386] In response to the algorithm selection request, generate the system application request initiated by the target user for the target processing system.

[0387] As an embodiment, the processing component can also be used for:

[0388] Detect the parameter adjustment request initiated by the target user for the proportional parameter;

[0389] In response to the parameter adjustment request, obtain the parameter adjustment information provided by the target user;

[0390] Based on the parameter adjustment information, adjust the target parameter instance of the proportional parameter to obtain the adjusted target parameter instance.

[0391] In some embodiments, the processing component can also be used for:

[0392] Generate the expected setting information of the target processing system according to the second parameter instance;

[0393] Display the target processing system corresponding to the expected setting information to the target user using the target processing system for the target user to use the target processing system corresponding to the expected setting information.

[0394] As a possible implementation, the processing component can also be used for:

[0395] Obtain the user attribute information of the target user;

[0396] The processing component feeds back system feedback information to the target user in the following manner:

[0397] Based on the user attribute information, select the target feedback type that best matches the target user from a preset multiple feedback types;

[0398] Generate the system display information corresponding to the system feedback information according to the target feedback type;

[0399] Display the system display information to the information target through the target processing system.

[0400] As another embodiment, the processing component is also used for:

[0401] Determine the system access request initiated by the target user for the network trading system; in response to the system access request, determine the target parameter instance corresponding to the proportional parameter in the network trading system; obtain the usage effect information generated by the target user using the network trading system corresponding to the target parameter instance; based on the usage effect information, generate the system feedback information generated by the target parameter instance in the network trading system; wherein, the system feedback information is used to feedback to the target user.

[0402] Figure 9 The implemented computing device can execute Figure 1 The data processing methods of the embodiments such as etc., and their implementation principles and technical effects will not be elaborated. The specific manners of each step executed by the processing component in the above embodiments have been described in detail in the embodiments related to the method, and will not be elaborated here.

[0403] In addition, the embodiment of the present application also provides a computer-readable storage medium, which is used to store a computer program, and when the computer program is executed, it can execute the data processing method in the foregoing embodiments.

[0404] The device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place, or may be distributed to multiple network units. Some or all of the modules can be selected according to actual needs to achieve the purpose of the solution of this embodiment. Those of ordinary skill in the art can understand and implement without creative labor.

[0405] Through the description of the above embodiments, those skilled in the art can clearly understand that each embodiment can be implemented by means of adding a necessary general hardware platform, and of course, it can also be implemented by a combination of hardware and software. Based on such an understanding, the above technical solution, in essence, or the part that contributes to the prior art can be embodied in the form of a computer product. The present application can adopt the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk memories, CD-ROMs, optical memories, etc.) that contain computer-usable program codes.

[0406] The present application is described with reference to the flowcharts and / or block diagrams of methods, apparatuses (systems), and computer program products according to embodiments of the present application. It should be understood that each flow and / or block in the flowchart and / or block diagram, as well as the combination of flows and / or blocks in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to the processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing devices to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing devices generate means for implementing the functions specified in Figure 1 one flow or multiple flows and / or blocks Figure 1 one block or multiple blocks.

[0407] These computer program instructions can also be stored in a computer-readable memory that can direct a computer or other programmable data processing device to work in a specific manner, so that the instructions stored in the computer-readable memory generate a manufactured article including an instruction device, and the instruction device implements the functions specified in Figure 1 one flow or multiple flows and / or blocks Figure 1 one block or multiple blocks.

[0408] These computer program instructions can also be loaded onto a computer or other programmable data processing device, so that a series of operation steps are executed on the computer or other programmable device to generate a computer-implemented process, and thus the instructions executed on the computer or other programmable device provide steps for implementing the functions specified in Figure 1 one flow or multiple flows and / or blocks Figure 1 one block or multiple blocks.

[0409] In a typical configuration, a computing device includes one or more processors (CPUs), an input / output interface, a network interface, and a memory.

[0410] The memory may include non-permanent memory in the form of computer-readable media, random access memory (RAM) and / or non-volatile memory such as read-only memory (ROM) or flash RAM. The memory is an example of computer-readable media.

[0411] Computer-readable media includes both permanent and non-permanent, removable and non-removable media implemented by any method or technology for information storage. The information can be computer-readable instructions, data structures, program modules, or other data. Examples of computer storage media include, but are not limited to, phase change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technologies, compact disc read-only memory (CD-ROM), digital versatile discs (DVD) or other optical storage, magnetic cassettes, magnetic tape magnetic disk storage or other magnetic storage devices, or any other non-transitory media that can be used to store information accessible by a computing device. As defined herein, computer-readable media does not include transitory computer-readable media such as modulated data signals and carrier waves.

[0412] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present application and are not intended to limit them; although the present application has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that they can still modify the technical solutions recorded in the foregoing embodiments, or perform equivalent replacements for some of the technical features; and these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the various embodiments of the present application.

Claims

1. A data processing method, characterized in that, Including: Determine a target parameter instance corresponding to a proportional parameter in a target processing system, including generating, through a target generation algorithm, a target parameter instance corresponding to the proportional parameter in the target processing system that includes at least one system sub-parameter. The target generation algorithm generates the target parameter instance in the following manner: Obtain the parameter characteristics respectively corresponding to the at least one system sub-parameter in the target processing system, and based on the parameter characteristics respectively corresponding to the at least one system sub-parameter, estimate the probabilities of a target user performing target operations on the at least one system sub-parameter respectively, obtain the trigger probabilities respectively corresponding to the at least one system sub-parameter. The trigger probabilities respectively corresponding to the at least one system sub-parameter are the parameter values respectively corresponding to the at least one system sub-parameter, and the parameter values respectively corresponding to the at least one system sub-parameter constitute the target parameter instance; Obtain usage effect information generated by the target user using the target processing system; Based on the usage effect information, generate system feedback information generated by the target parameter instance in the target processing system; Wherein, the system feedback information is used to provide feedback to the target user.

2. The method according to claim 1, wherein The determination of the target parameter instance corresponding to the proportional parameter in the target processing system includes: Determine a system application request initiated by the target user for the target processing system; In response to the system application request, determine the target parameter instance corresponding to the proportional parameter in the target processing system.

3. The method according to claim 1, wherein The determination of the target parameter instance corresponding to the proportional parameter in the target processing system includes: When it is detected that the target processing system meets the update condition, determine the target parameter instance corresponding to the proportional parameter in the target processing system.

4. The method according to claim 1, characterized in that After obtaining the usage effect information generated by the target user using the target processing system corresponding to the target parameter instance, it further includes: According to the usage effect information of the target parameter instance, determine whether the target parameter instance meets the parameter usage condition; The generation of the system feedback information generated by the target parameter instance in the target processing system based on the usage effect information includes: If the target parameter instance meets the parameter usage condition, generate the system feedback information corresponding to the target parameter instance.

5. The method according to claim 4, characterized in that, It further includes: If the target parameter instance does not meet the parameter usage condition, update the target parameter instance corresponding to the proportional parameter based on the usage effect information; Display to the target user the target processing system corresponding to the updated target parameter instance, and return to the step of obtaining the usage effect information generated by the target user using the target processing system corresponding to the target parameter instance and continue to execute.

6. The method according to claim 2, characterized in that The determination of the target parameter instance corresponding to the proportional parameter in the target processing system in response to the system application request includes: In response to the system application request, determine the target generation algorithm corresponding to the proportional parameter in the target processing system; Through the target generation algorithm, generate the target parameter instance corresponding to the proportional parameter in the target processing system.

7. The method according to claim 1, characterized in that, Estimating the probabilities that the target user performs target operations on the at least one system sub-parameter respectively according to the parameter characteristics corresponding to the at least one system sub-parameter, and obtaining the trigger probabilities corresponding to the at least one system sub-parameter respectively includes: Determining the user characteristics of the target user in the target processing system; Estimating the probabilities that the target user performs target operations on the at least one system sub-parameter respectively according to the user characteristics and the parameter characteristics corresponding to the at least one system sub-parameter, and obtaining the trigger probabilities corresponding to the at least one system sub-parameter respectively.

8. The method according to claim 7, wherein The estimating the probabilities that the target user performs target operations on the at least one system sub-parameter respectively according to the user characteristics and the parameter characteristics corresponding to the at least one system sub-parameter, and obtaining the trigger probabilities corresponding to the at least one system sub-parameter respectively includes: Estimating the trigger results of the target user performing target operations on any one of the at least one system sub-parameter N times according to the user characteristics and the parameter characteristics corresponding to the at least one system sub-parameter, and obtaining N trigger results; wherein, any one trigger result is the target user performing a target operation on any one of the at least one system sub-parameter; Determining the trigger times corresponding to the at least one system sub-parameter respectively according to the N trigger results; Determining the trigger probabilities corresponding to the at least one system sub-parameter respectively according to the ratios of the trigger times corresponding to the at least one system sub-parameter respectively to N.

9. The method according to claim 8, wherein The trigger result is determined by the following method: Generating an online prediction model according to the user characteristics and the parameter characteristics corresponding to the at least one system sub-parameter; Using the online prediction model to estimate the probabilities of the reward information obtained by the at least one system sub-parameter respectively, and obtaining the reward probabilities corresponding to the at least one system sub-parameter respectively; Determining the system sub-parameter with the maximum reward probability as the system sub-parameter for the target user to perform the target operation according to the reward probabilities corresponding to the at least one system sub-parameter respectively.

10. The method according to claim 9, wherein The generating an online prediction model according to the user characteristics and the parameter characteristics corresponding to the at least one system sub-parameter includes: Obtaining the historical reward information corresponding to the at least one system sub-parameter respectively; Generating the online prediction model according to the user characteristics and the parameter characteristics corresponding to the at least one system sub-parameter, in combination with the historical reward information corresponding to the at least one system sub-parameter respectively.

11. The method according to claim 1, characterized in that The obtaining the usage effect information generated by the target user using the target processing system corresponding to the target parameter instance includes: Obtaining historical parameter instances and the historical usage information corresponding to the historical parameter instances; Generating the first sub-effect information corresponding to the at least one system sub-parameter respectively according to the historical usage information corresponding to the historical parameter instances and the current usage information of the target parameter instance; Determining the usage effect information constituted by the first sub-effect information corresponding to the at least one system sub-parameter respectively.

12. The method according to claim 5, wherein If the target parameter instance meets the parameter usage condition, generating the system feedback information corresponding to the target parameter instance includes: If the target parameter instance meets the parameter usage condition, generating the system setting information of the target processing system according to the target parameter instance; The system feedback information is fed back to the target user in the following way: Showing the target processing system corresponding to the system setting information to the target user for the target user to use the target processing system corresponding to the system setting information.

13. The method according to claim 12, characterized in that, It also includes: Determining at least one system sub-parameter corresponding to the proportional parameter in the target processing system; Determining the system parameters corresponding to at least one system sub-parameter of the proportional parameter in the target processing system; If the target parameter instance meets the parameter usage condition, generating the system setting information of the target processing system according to the target parameter instance includes: If the target parameter instance meets the parameter usage condition, obtaining the parameter values corresponding to the target parameter instance in at least one system sub-parameter respectively; According to the parameter value corresponding to any one system sub-parameter, determining the system data of the system parameter corresponding to the system sub-parameter; Determining the system setting information composed of the system data corresponding to at least one system parameter respectively.

14. The method according to claim 4, wherein If the target parameter instance meets the parameter usage condition, generating the system feedback information corresponding to the target parameter instance includes: If the target parameter instance meets the parameter usage condition, generating the parameter prompt information corresponding to the target parameter instance; The system feedback information is fed back to the target user in the following way: Outputting the parameter prompt information for the target user for the target user to view the target parameter instance.

15. The method according to claim 1, wherein Determining the target parameter instance corresponding to the proportional parameter in the target processing system includes: Determining a plurality of parameter generation algorithms corresponding to the proportional parameter in the target processing system; Determining the target parameter instances generated by the plurality of parameter generation algorithms for the proportional parameter respectively to obtain a plurality of target parameter instances; Obtaining the usage effect information generated by the target user using the target processing system corresponding to the target parameter instance includes: Obtaining the usage effect information generated by the target user using the target processing system corresponding to any one target parameter instance to obtain the usage effect information corresponding to each of the plurality of target parameter instances; Based on the usage effect information, generating the system feedback information generated by the target parameter instance in the target processing system includes: Based on the usage effect information corresponding to each of the plurality of target parameter instances, determining a target parameter instance from the plurality of target parameter instances; Determining the parameter generation algorithm that generates the target parameter instance as the target generation algorithm; According to the target generation algorithm, generating the algorithm prompt information in the target processing system; The system feedback information is fed back to the target user in the following way: Showing the algorithm prompt information to prompt the target generation algorithm that best matches the target user.

16. The method according to claim 15, wherein, The usage effect information generated by the target user using the target processing system corresponding to any one target parameter instance is obtained by the following method: A simulation system for generating the target processing system; Based on the simulation system, simulating the simulated effect information generated by the target user using the target processing system; Determining the usage effect information according to the target parameter instance and the simulated effect information.

17. The method according to claim 16, wherein The determining the second sub-effect information corresponding to each of the at least one system sub-parameters according to the effect evaluation values corresponding to the at least one system sub-parameters respectively, and the parameter values corresponding to the target parameter instance in the at least one system sub-parameters respectively includes: According to the effect evaluation values corresponding to the at least one system sub-parameters respectively, and the parameter values corresponding to the target parameter instance in the at least one system sub-parameters respectively, multiplying the effect evaluation value and the parameter value of any one system sub-parameter to obtain the second sub-effect information of the system sub-parameter, so as to obtain the second sub-effect information corresponding to each of the at least one system sub-parameters.

18. The method according to claim 15, wherein The determining the system application request initiated by the target user for the target processing system includes: Receiving an algorithm selection request initiated by an optimization user; In response to the algorithm selection request, generating a system application request initiated by the target user for the target processing system.

19. The method according to claim 1, characterized in that After the system feedback information is fed back to the target user, it further includes: Detecting a parameter adjustment request initiated by the target user for the ratio parameter; Responding to the parameter adjustment request and obtaining parameter adjustment information provided by the target user; Based on the parameter adjustment information, adjusting the target parameter instance of the ratio parameter to obtain the adjusted target parameter instance.

20. The method according to claim 19, wherein After adjusting the target parameter instance of the ratio parameter based on the parameter adjustment information to obtain the adjusted target parameter instance, it further includes: Generating expected setting information of the target processing system according to the target parameter instance; Showing the target processing system corresponding to the expected setting information to the target user using the target processing system for the target user to use the target processing system corresponding to the expected setting information.

21. The method according to claim 1, characterized in that, It further includes: Obtaining the user attribute information of the target user; The system feedback information is fed back to the target user in the following manner: Based on the user attribute information, selecting the target feedback type that best matches the target user from a preset plurality of feedback types; Generating system display information corresponding to the system feedback information according to the target feedback type; Showing the system display information to the information target through the target processing system.

22. A data processing method, characterized in that, It includes: Determining a system access request initiated by the target user for the online trading system; In response to the system access request, determine a target parameter instance corresponding to a proportional parameter in the network transaction system, including generating, through a target generation algorithm, a target parameter instance corresponding to the proportional parameter in the network transaction system that includes at least one system sub-parameter. The target generation algorithm generates the target parameter instance in the following manner: obtain the parameter characteristics respectively corresponding to the at least one system sub-parameter in the network transaction system, and based on the parameter characteristics respectively corresponding to the at least one system sub-parameter, estimate the probabilities of the target user performing target operations on the at least one system sub-parameter respectively, to obtain the trigger probabilities respectively corresponding to the at least one system sub-parameter. The trigger probabilities respectively corresponding to the at least one system sub-parameter are the parameter values respectively corresponding to the at least one system sub-parameter, and the parameter values respectively corresponding to the at least one system sub-parameter constitute the target parameter instance; Obtain the usage effect information generated by the target user using the network transaction system corresponding to the target parameter instance; Based on the usage effect information, generate system feedback information generated by the target parameter instance in the network transaction system; Wherein, the system feedback information is used to provide feedback to the target user.

23. A data processing method, characterized in that, Including: In response to a usage request for invoking a data processing interface, determine the processing resources corresponding to the data processing interface; Use the processing resources corresponding to the data processing interface to perform the following steps: Determine a target parameter instance corresponding to a proportional parameter in a target processing system, including generating, through a target generation algorithm, a target parameter instance corresponding to the proportional parameter in the target processing system that includes at least one system sub-parameter. The target generation algorithm generates the target parameter instance in the following manner: obtain the parameter characteristics respectively corresponding to the at least one system sub-parameter in the target processing system, and based on the parameter characteristics respectively corresponding to the at least one system sub-parameter, estimate the probabilities of the target user performing target operations on the at least one system sub-parameter respectively, to obtain the trigger probabilities respectively corresponding to the at least one system sub-parameter. The trigger probabilities respectively corresponding to the at least one system sub-parameter are the parameter values respectively corresponding to the at least one system sub-parameter, and the parameter values respectively corresponding to the at least one system sub-parameter constitute the target parameter instance; Obtain the usage effect information generated by the target user using the target processing system corresponding to the target parameter instance; Based on the usage effect information, generate system feedback information generated by the target parameter instance in the target processing system; Wherein, the system feedback information is used to provide feedback to the target user.

24. A computing device, characterized in that, Including: a storage component and a processing component; the storage component is used to store one or more computer instructions, and the one or more computer instructions are invoked by the processing component; The processing component is used for: Determine the target parameter instance corresponding to the proportional parameter in the target processing system, including generating the target parameter instance corresponding to the proportional parameter including at least one system sub-parameter in the target processing system through a target generation algorithm. The target generation algorithm generates the target parameter instance in the following manner: obtain the parameter characteristics corresponding to the at least one system sub-parameter in the target processing system respectively, estimate the probabilities of the target user performing target operations on the at least one system sub-parameter respectively according to the parameter characteristics corresponding to the at least one system sub-parameter respectively, obtain the trigger probabilities corresponding to the at least one system sub-parameter respectively, the trigger probabilities corresponding to the at least one system sub-parameter respectively are the parameter values corresponding to the at least one system sub-parameter respectively, and the parameter values corresponding to the at least one system sub-parameter respectively constitute the target parameter instance; obtain the usage effect information generated by the target user using the target processing system corresponding to the target parameter instance; based on the usage effect information, generate the system feedback information generated by the target parameter instance in the target processing system; wherein, the system feedback information is used to feedback to the target user.

Images