Virtual product standard image storage method, device, equipment and medium

Abstract

Embodiments of the present disclosure disclose a virtual product standard image storage method, device, equipment and medium. A specific embodiment of the method comprises: obtaining a product use data set; for each product use object, generating an object image; performing attribute and attribute information clipping on the object image set to obtain at least one product use attribute and a post-removal object image set; for each product use attribute, performing a generation step: determining a product attribute information set; obtaining an attribute anomaly determination rule set and an attribute anomaly generation model set; respectively screening out a target attribute anomaly determination rule and a target attribute anomaly generation model; generating attribute anomaly standard information; generating a first standard information image; and storing the first standard information image and the post-removal object image set. The embodiment can efficiently and accurately generate a standard information image for abnormality judgment of a target virtual product while adapting to current computing residual resources.

Description

TECHNICAL FIELD

[0001] Embodiments of the present disclosure relate to the technical field of computer technology, and particularly to a virtual product standard image storage method, device, equipment and medium. BACKGROUND

[0002] At present, in the use scene of virtual products, the abnormality determination of virtual products is crucial, which can effectively avoid the value loss. For the generation of abnormality determination standards corresponding to virtual products, the commonly used way is: directly input the product use data set corresponding to the target virtual product into the multi-attribute abnormality standard information output model to obtain the abnormality determination standards under each attribute.

[0003] However, when the above method is used, the following technical problems often exist:

[0004] The product use data set often has a large amount of data, which leads to a large waste of computing time when directly inputting into the multi-attribute abnormality standard information output model, and makes the multi-attribute abnormality standard information output model learn more redundant features. And in the case that the multi-attribute abnormality standard information output model needs a large amount of computing resources, if the terminal cannot effectively supply the computing resources, the output will be interrupted and the generation of abnormality determination standards will be affected.

[0005] The above information disclosed in the background section is only intended to enhance the understanding of the background of the present inventive concept, and therefore, it can include information that does not form the prior art known to those of ordinary skill in the art in the country. SUMMARY

[0006] The summary section is provided to introduce the concepts briefly in a simplified form, which will be described in detail in the specific embodiments section later. The summary section is not intended to identify key or essential features of the claimed technical solutions, nor is it intended to be used to limit the scope of the claimed technical solutions.

[0007] Some embodiments of the present disclosure propose a virtual product standard image storage method, device, equipment and medium to solve the technical problems mentioned in the background section.

[0008] In a first aspect, some embodiments of the present disclosure provide a virtual product standard image storage method, comprising: obtaining a product use data set of a target virtual product within a predetermined time period; for each product use object in the product use object set, generating an object image corresponding to the product use object according to the product use data corresponding to the product use object; performing attribute and attribute information pruning on the product use attribute set corresponding to the obtained object image set to obtain at least one product use attribute and a post-removal object image set; for each product use attribute in the at least one product use attribute, using a distributed graphics processor cluster to perform a generation step: determining a product attribute information set corresponding to the product use attribute in the object image set; obtaining a pre-stored attribute anomaly determination rule set and an attribute anomaly generation model set corresponding to the product use attribute; according to the remaining computing resources corresponding to the graphics processor cluster, screening out a target attribute anomaly determination rule and a target attribute anomaly generation model from the attribute anomaly determination rule set and the attribute anomaly generation model set, respectively; according to the product attribute information set, using the target attribute anomaly determination rule and the target attribute anomaly generation model to generate attribute anomaly standard information for the product use attribute; generating a first standard information image corresponding to the obtained at least one attribute anomaly standard information, wherein the first standard information image and the object image are of the same tree structure; storing the first standard information image and the post-removal object image set in a target storage terminal.

[0009] In a second aspect, some embodiments of the present disclosure provide a virtual product standard image storage device, comprising: an acquisition unit configured to acquire a product use data set of a target virtual product within a predetermined time period; a first generation unit configured to, for each product use object in a product use object set, generate an object image corresponding to the product use object according to product use data corresponding to the product use object; a clipping unit configured to clip a product use attribute set corresponding to the obtained object image set in terms of attributes and attribute information, to obtain at least one product use attribute and a post-removal object image set; an execution unit configured to, for each product use attribute in the at least one product use attribute, execute the following generation step using a distributed graphics processor cluster: determining a product attribute information set in the object image set corresponding to the product use attribute; acquiring a pre-stored attribute anomaly determination rule set and an attribute anomaly generation model set corresponding to the product use attribute; according to the corresponding calculation remaining resources of the graphics processor cluster, screening a target attribute anomaly determination rule and a target attribute anomaly generation model from the attribute anomaly determination rule set and the attribute anomaly generation model set, respectively; generating attribute anomaly standard information for the product use attribute according to the product attribute information set, the target attribute anomaly determination rule, and the target attribute anomaly generation model; a second generation unit configured to generate a first standard information image corresponding to the obtained at least one attribute anomaly standard information, wherein the first standard information image and the object image are of the same tree structure; and a storage unit configured to store the first standard information image and the post-removal object image set in a target storage terminal.

[0010] In a third aspect, some embodiments of the present disclosure provide an electronic device, comprising: one or more processors; and a storage device having one or more programs stored thereon, when the one or more programs are executed by the one or more processors, the one or more processors implement the method described in any implementation manner of the first aspect.

[0011] In a fourth aspect, some embodiments of the present disclosure provide a computer readable medium having a computer program stored thereon, wherein the program is executed by a processor to implement the method described in any implementation manner of the first aspect.

[0012] The above various embodiments of the present disclosure have the following beneficial effects: through the virtual product standard portrait storage method of some embodiments of the present disclosure, the standard information portrait for abnormal judgment of the target virtual product can be efficiently and accurately generated in the case of adapting to the current remaining computing resources. Specifically, the reason for the insufficient accuracy and timeliness of the related abnormal judgment is that the product usage dataset is often large in data volume, resulting in a waste of a large amount of computing time when directly input into the multi-attribute abnormal standard information output model, and causing the multi-attribute abnormal standard information output model to learn more redundant features. And in the case that the multi-attribute abnormal standard information output model needs a large amount of computing resources, if the terminal cannot effectively supply computing resources, it will cause the output to be interrupted and affect the generation of abnormal judgment standards. Based on this, the virtual product standard portrait storage method of some embodiments of the present disclosure first acquires the product usage dataset of the target virtual product within a predetermined time period, which is used as the basic dataset for abnormal judgment of each product usage attribute. Based on this, the accurate generation of the virtual product standard portrait can be realized. Then, for each product usage object in the product usage object set, the object portrait corresponding to the product usage object can be accurately generated according to the product usage data corresponding to the product usage object. Through the object portrait, the object portrait of each usage object corresponding to the target virtual product is determined, which facilitates the subsequent query of the product usage object related information. Next, the product usage attribute set corresponding to the obtained object portrait set is subjected to attribute and attribute information pruning to obtain at least one product usage attribute and a removed object portrait set, so as to determine at least one product usage attribute for key monitoring and a removed object portrait set including key product usage attribute information. Then, for each product usage attribute in the at least one product usage attribute, a distributed graphics processor cluster is used to perform the following generation steps: first, determine the product attribute information set in the object portrait set corresponding to the product usage attribute, so as to facilitate the generation of attribute abnormal standard information corresponding to the product usage attribute. In addition, the distributed graphics processor cluster can effectively provide computing resources available for computing. Second, acquire the pre-stored attribute abnormal determination rule set and attribute abnormal generation model set corresponding to the product usage attribute, so as to jointly determine the abnormal judgment standard from the rule and neural network model angles, which can greatly improve the accuracy of the judgment. Third, according to the remaining computing resources corresponding to the graphics processor cluster, the target attribute abnormal determination rule and the target attribute abnormal generation model are selected from the attribute abnormal determination rule set and the attribute abnormal generation model set, respectively, so as to effectively select the rules and models for subsequent real-time generation of attribute abnormal standard information in the case of current computing resources. Fourth, according to the product attribute information set, the target attribute abnormal determination rule and the target attribute abnormal generation model can be used to accurately generate attribute abnormal standard information for the product usage attribute.Further, a first standard information portrait corresponding to the generated at least one attribute exception standard information is generated, wherein the first standard information portrait and the object portrait are of the same tree structure. Here, through the form of the first standard information portrait, the efficiency of the exception determination can be improved and the determination time can be saved when the subsequent exception determination is performed. Finally, the first standard information portrait and the object portrait set after the removal are stored in a target storage terminal, so as to facilitate the execution of the subsequent exception determination and the periodic update of the first standard information portrait. BRIEF DESCRIPTION OF DRAWINGS

[0013] The above and other features, aspects and advantages of the embodiments of the present disclosure will become more apparent from the following detailed description when taken in conjunction with the accompanying drawings. Throughout the drawings, the same or similar reference numerals are used to represent the same or similar elements. It should be understood that the drawings are schematic and elements and features are not necessarily to scale.

[0014] Figure 1 is a flowchart of some embodiments of a virtual product standard portrait storage method according to the present disclosure;

[0015] Figure 2 is a structural schematic diagram of some embodiments of a virtual product standard portrait storage device according to the present disclosure;

[0016] Figure 3 is a structural schematic diagram of an electronic device suitable for use to implement some embodiments of the present disclosure. DETAILED DESCRIPTION

[0017] Embodiments of the present disclosure will be described in more detail below with reference to the accompanying drawings. Although some embodiments of the present disclosure are shown in the drawings, it should be understood that the present disclosure can be implemented in various forms, and should not be interpreted as being limited to the embodiments set forth herein. On the contrary, these embodiments are provided to make the present disclosure more thorough and complete. It should be understood that the drawings and embodiments of the present disclosure are only for exemplary purposes, and are not intended to limit the scope of protection of the present disclosure.

[0018] It should also be noted that, for the sake of description, only the parts related to the present invention are shown in the drawings. The embodiments in the present disclosure and the features in the embodiments can be combined with each other without conflict.

[0019] It should be noted that the terms "first", "second", etc. mentioned in the present disclosure are only used to distinguish different devices, modules or units, and are not intended to limit the order or interdependence of the functions performed by these devices, modules or units.

[0020] It should be noted that the modification of "one", "multiple" mentioned in the present disclosure is illustrative but not restrictive, and those skilled in the art should understand that unless otherwise explicitly indicated in the context, it should be understood as "one or more".

[0021] The names of the messages or information exchanged between the plurality of devices in the embodiments of the present disclosure are only for illustrative purposes, and are not used to limit the scope of the messages or information.

[0022] The present disclosure will be described in detail below with reference to the accompanying drawings and in conjunction with embodiments.

[0023] Reference Figure 1 , shows the flow 100 of some embodiments of the virtual product standard portrait storage method according to the present disclosure. The virtual product standard portrait storage method includes the following steps:

[0024] Step 101, obtaining a product use data set of a target virtual product in a predetermined time period. In some embodiments, the execution subject (for example, an electronic device) of the virtual product standard portrait storage method described above can obtain the product use data set of the target virtual product in the predetermined time period through wired connection or wireless connection. Wherein, the target virtual product can be a virtual product to be determined for abnormal standard. In practice, the virtual product can be a virtual product related to value. For example, in the field of credit investigation, the virtual product can be a credit product. For another example, for the express delivery scene, the corresponding virtual product can be an express delivery App. The predetermined time period can be a pre-set time period. The predetermined time period can be a historical time period before the current time. The product use data can be the use data of the product use object using the target virtual product in the predetermined time period. Each product use data corresponds to a product use object. The product use object can be an object using the target virtual product.

[0025] Step 102, for each product use object in the product use object set, generating an object portrait corresponding to the product use object according to the product use data corresponding to the product use object.

[0026] In some embodiments, the execution subject can generate an object portrait corresponding to each product use object in the product use object set according to the product use data corresponding to the product use object. Wherein, the object portrait can be a description portrait corresponding to the product use object. The object portrait can represent the use feature situation of each aspect of the product use object using the target virtual product. In practice, the object portrait can include but not limited to at least one of the following: product use object name, product use time, input amount, profit amount, product use duration, product natural person.

[0027] As an example, first, the execution subject can extract each product usage attribute information corresponding to each product usage attribute from the product usage data. Then, according to each product usage attribute information, an object portrait of a tree structure is constructed.

[0028] In step 103, the product usage attribute set corresponding to the obtained object portrait set is subjected to attribute and attribute information pruning, and at least one product usage attribute and the object portrait set after removal are obtained.

[0029] In some embodiments, the execution subject can perform attribute and attribute information pruning on the product usage attribute set corresponding to the obtained object portrait set, and obtain at least one product usage attribute and the object portrait set after removal. The at least one product usage attribute can represent each product usage attribute that is crucial in the current abnormality determination process of the target virtual product. That is, the abnormality of the product usage object using the target virtual product is determined by the attribute content corresponding to the at least one product usage attribute. The object portrait after removal can be a portrait in which the object portrait removes each product usage attribute information that is crucial in the abnormality determination process of the target virtual product.

[0030] As an example, first, the product usage object selects each product usage attribute that is crucial in the abnormality determination process of the target virtual product on the target interface as at least one product usage attribute. Then, for each object portrait in the object portrait set, the attribute information corresponding to the at least one product usage attribute is removed from the object portrait to obtain the object portrait after removal.

[0031] In some optional implementations of some embodiments, the execution subject can perform attribute and attribute information pruning on the product usage attribute set corresponding to the obtained object portrait set to obtain at least one product usage attribute and the object portrait set after removal, including the following steps:

[0032] First, generate attribute prediction effective information corresponding to each product usage attribute in the product usage attribute set to obtain an attribute prediction effective information set. The attribute prediction effective information is an effective value representing the importance of the attribute and the strength of the attribute prediction ability. The strength of the attribute prediction ability can represent the ease of obtaining the attribute content corresponding to the attribute by prediction. The higher the value corresponding to the strength of the attribute prediction ability, the easier it is to obtain the attribute content by prediction. The attribute prediction effective information can be a value set by a relevant technical personnel for each attribute based on historical experience.

[0033] Second, remove the product usage attribute corresponding to the attribute prediction effective information less than the target value from the product usage attribute set to obtain at least one product usage attribute. The target value can be a value set in advance by a relevant technical personnel according to experience.

[0034] In the third step, for each object image in the object image set, the attribute information corresponding to the at least one product use attribute is removed from the object image to obtain a removed object image.

[0035] In step 104, for each product use attribute in the at least one product use attribute, a distributed graphics processor cluster is used to perform the generating step.

[0036] In step 1041, a product attribute information set corresponding to the product use attribute in the object image set is determined.

[0037] In some embodiments, the execution subject can determine the product attribute information set corresponding to the product use attribute in the object image set. The graphics processor cluster can be a graphics processor (GPU) based cluster. The object image in the object image set and the product attribute information in the product attribute information set have a use object correspondence.

[0038] In step 1042, the pre-stored attribute abnormality determination rule set and the attribute abnormality generation model set corresponding to the product use attribute are obtained.

[0039] In some embodiments, the execution subject can obtain the pre-stored attribute abnormality determination rule set and the attribute abnormality generation model set corresponding to the product use attribute. The attribute abnormality determination rule can be a determination rule for determining attribute abnormality pre-set by a relevant expert technician. The attribute abnormality generation model can be a pre-trained neural network model for predicting attribute abnormality information. In practice, the model structure or model training method of each attribute abnormality generation model in the attribute abnormality generation model set can be different. The rule content corresponding to each attribute abnormality determination rule in the attribute abnormality determination rule set can be different. The model structure corresponding to the attribute abnormality generation model is determined according to the content type of the attribute content corresponding to the product use attribute. For example, for the content type of label type, each attribute abnormality generation model in the attribute abnormality generation model set can be a classification model. For the content type of numerical continuous type, each attribute abnormality model in the attribute abnormality generation model set can be a regression model.

[0040] In step 1043, according to the calculation remaining resources of the graphics processor cluster, a target attribute abnormality determination rule and a target attribute abnormality generation model are selected from the attribute abnormality determination rule set and the attribute abnormality generation model set, respectively.

[0041] In some embodiments, the execution subject can filter out target attribute anomaly determination rules and target attribute anomaly generation models from the attribute anomaly determination rule set and the attribute anomaly generation model set respectively according to the remaining computing resources of the graphics processor cluster corresponding to the graphics processor cluster.

[0042] In some optional implementations of some embodiments, the execution subject can filter out target attribute anomaly determination rules and target attribute anomaly generation models from the attribute anomaly determination rule set and the attribute anomaly generation model set respectively according to the remaining computing resources of the graphics processor cluster corresponding to the graphics processor cluster, including the following steps:

[0043] First, determine the sequence of remaining computing resources of the graphics processor cluster in a target future time period. The target future time period can be a time period in a future predetermined duration. The future predetermined duration can be pre-set. The remaining computing resources in the sequence of remaining computing resources have a one-to-one correspondence with the future time points in the target future time period. The sequence of remaining computing resources can represent the change of resources of the graphics processor cluster in the target future time period.

[0044] Second, determine the attribute division resource sequence corresponding to the product use attribute from the sequence of remaining computing resources. Each product use attribute has a corresponding attribute division resource sequence. The attribute division resource can be the remaining resource that the product use attribute can use at the target future time point. The attribute division resources in the attribute division resource sequence have a one-to-one correspondence with the future time points in the target future time period. That is, the attribute division resource sequence can represent the change sequence of resources that the graphics processor can provide for the attribute calculation process corresponding to the product use attribute.

[0045] As an example, for each remaining computing resource in the sequence of remaining computing resources, determine the number of attributes corresponding to the at least one product use attribute. Then, divide the remaining computing resource by the number of attributes to obtain the attribute division resource corresponding to the product use attribute.

[0046] Third, determine the set of collocation combinations between the attribute anomaly determination rule set and the attribute anomaly generation model set. Each collocation combination includes an attribute anomaly determination rule and an attribute anomaly generation model. Each collocation combination in the set of collocation combinations is flexibly collocated according to the total amount of computation consumed by the attribute anomaly determination rule and the attribute anomaly generation model. The total amount of computation can represent the amount of resource consumption.

[0047] It should be noted that each attribute exception determination rule has a corresponding rule accuracy and computational amount, and the attribute exception generation model has a corresponding model accuracy and computational amount. Related researchers can also perform corresponding combinations of each attribute exception determination rule and each attribute exception generation model in the form of interface dragging to obtain a collocation combination set.

[0048] In the fourth step, for each collocation combination in the collocation combination set, the following generation steps are performed:

[0049] Substep 1, determine the corresponding estimated consumption time and estimated consumption resources of the attribute exception determination rule and the attribute exception generation model in the collocation combination. The estimated consumption time can be the total time predicted to be consumed by the attribute exception determination rule and the attribute exception generation model to complete attribute exception determination. The estimated consumption resources can be the total amount of resources predicted to be consumed by the attribute exception determination rule and the attribute exception generation model to complete attribute exception determination.

[0050] As an example, first, the first consumption time corresponding to the attribute exception determination rule and the second consumption time corresponding to the attribute exception generation model are determined. Then, the highest consumption time in the first consumption time and the second consumption time is determined as the estimated consumption time. Next, the first consumption resource corresponding to the attribute exception determination rule and the second consumption resource corresponding to the attribute exception generation model are determined. Then, the resource and between the first consumption resource and the second consumption resource is determined as the estimated consumption resource.

[0051] Substep 2, determine whether the attribute division resource sequence satisfies the consumption requirement corresponding to the estimated consumption time and the estimated consumption resource, and obtain satisfaction information. The satisfaction information can be one of the following: satisfaction information representing that the consumption requirement corresponding to the estimated consumption time and the estimated consumption resource is satisfied, and satisfaction information representing that the consumption requirement corresponding to the estimated consumption time and the estimated consumption resource is not satisfied. The consumption requirement can be that the attribute division resource subsequence corresponding to the estimated consumption time in the attribute division resource sequence has more attribute division resources than the estimated consumption resource.

[0052] In the fifth step, the satisfaction information representing that the consumption requirement corresponding to the estimated consumption time and the estimated consumption resource is satisfied is filtered out from the obtained satisfaction information set, and at least one satisfaction information is obtained.

[0053] In the sixth step, the collocation combination with the highest comprehensive accuracy and the shortest estimated consumption time is selected from the at least one collocation combination as the target collocation combination. The target collocation combination includes a target attribute exception determination rule and a target attribute exception generation model. The at least one satisfaction information and the at least one collocation combination have a one-to-one correspondence.

[0054] Optionally, the execution subject can determine, from the calculated residual resource sequence, an attribute partition resource sequence corresponding to the product use attribute, including the following steps:

[0055] First, determine an attribute weight corresponding to each product use attribute in the at least one product use attribute, to obtain at least one attribute weight. The attribute weight can represent the importance of the attribute corresponding to the product use attribute. The attribute weight can be a value between 0 and 1, and the higher the value, the more important the attribute content corresponding to the product use attribute.

[0056] Second, according to the at least one attribute weight, each calculation residual resource in the calculated residual resource sequence is correspondingly partitioned to generate the available calculation resource of the product use attribute at the corresponding time, to obtain at least one attribute partition resource sequence.

[0057] As an example, for each calculation residual resource in the calculation residual resource sequence, according to the at least one attribute weight, the calculation residual resource is divided into at least one, to obtain at least one attribute partition resource.

[0058] Third, from the at least one attribute partition resource sequence, the attribute partition resource sequence corresponding to the product use attribute is selected.

[0059] In some optional implementations of some embodiments, the execution subject can select, from the attribute anomaly determination rule set and the attribute anomaly generation model set, a target attribute anomaly determination rule and a target attribute anomaly generation model according to the calculation residual resource corresponding to the graphics processor cluster, including the following steps:

[0060] First, the calculation residual resource corresponding to the graphics processor cluster is displayed on a filtering interface. The filtering interface displays the calculation residual resource and the initial attribute use calculation resource corresponding to each attribute.

[0061] Second, obtain adjustment information of the target user adjusting the attribute use calculation resource of each attribute on the filtering interface.

[0062] Third, according to the adjustment information, the initial attribute use calculation resource corresponding to each attribute is adjusted, and the adjusted attribute use calculation resource corresponding to each attribute is displayed on the filtering interface.

[0063] In the fourth step, in response to determining that the target attribute corresponding to the target attribute control in the click filtering interface is determined, the rule and model configuration pop-up window corresponding to the target attribute is popped up. The left half of the rule and model configuration pop-up window is a first configuration area corresponding to the target attribute abnormality determination rule. The right half of the rule and model configuration pop-up window is a second configuration area corresponding to the target attribute abnormality generation model. The first configuration area and the second configuration area support custom setting of rules and models. The first configuration area and the second configuration area also support importing rules and models from external files.

[0064] In the fifth step, in response to determining that the attribute abnormality determination rule and the target attribute abnormality generation model for the target attribute are configured in the rule and model configuration pop-up window and the configuration save control is clicked, the target attribute control corresponding to the target attribute is set to a target color, and an identification number corresponding to the attribute abnormality determination rule and an identification number corresponding to the target attribute abnormality generation model are displayed in the upper right part of the target attribute control corresponding to the target attribute, so as to realize configuration of the target attribute abnormality determination rule and the target attribute abnormality generation model corresponding to the target attribute and resource configuration of the target attribute.

[0065] In step 1044, according to the product attribute information set, the target attribute abnormality determination rule and the target attribute abnormality generation model are used to generate attribute abnormality standard information for the product use attribute.

[0066] In some embodiments, the execution subject can generate attribute abnormality standard information for the product use attribute according to the product attribute information set, the target attribute abnormality determination rule and the target attribute abnormality generation model. The attribute abnormality standard information can represent the abnormality judgment standard of the product use attribute corresponding to the abnormality. That is, the attribute abnormality standard information represents when the product use attribute is abnormal and when it is not abnormal.

[0067] As an example, first, the execution subject can input the product attribute information set into the target attribute abnormality determination rule and the target attribute abnormality generation model to obtain first abnormality information and second abnormality information. Then, the average value of the first abnormality information and the second abnormality information is determined as the attribute abnormality standard value. Finally, the attribute abnormality standard judgment information corresponding to the attribute abnormality standard value is generated as the attribute abnormality standard information.

[0068] In some optional implementations of some embodiments, the execution subject can generate attribute abnormality standard information for the product use attribute according to the product attribute information set, the target attribute abnormality determination rule and the target attribute abnormality generation model, including the following steps:

[0069] In the first step, in response to determining that the comprehensive accuracy corresponding to the target combination of the target attribute anomaly determination rule and the target attribute anomaly generation model corresponding to the target combination is not the highest in the combination set, the remaining computing resources of the graphics processor cluster are waited to meet the resource requirements of the combination with the highest comprehensive accuracy. The comprehensive accuracy can be the accuracy between the target attribute anomaly determination rule and the target attribute anomaly generation model included in the target combination. That is, the comprehensive accuracy is the sum of the accuracy corresponding to the target attribute anomaly determination rule and the accuracy corresponding to the target attribute anomaly generation model.

[0070] In the second step, in response to determining that the remaining computing resources of the graphics processor cluster meet the resource requirements of the combination with the highest comprehensive accuracy, the candidate attribute anomaly standard information for the product use attribute is generated by using the product attribute information set, the remaining computing resources of the graphics processor cluster, and the combination with the highest comprehensive accuracy.

[0071] As an example, the execution subject can use the remaining computing resources to control the input of the product attribute information set into the attribute anomaly determination rule and the attribute anomaly generation model included in the combination with the highest comprehensive accuracy to generate the candidate attribute anomaly standard information for the product use attribute.

[0072] The method further includes:

[0073] In the first step, the second standard information portrait corresponding to the at least one generated candidate attribute anomaly standard information is generated. The second standard information portrait and the object portrait are in the same tree structure.

[0074] In the second step, the second standard information portrait and the object portrait set after removal are stored in the target storage terminal to replace the first standard information portrait and the object portrait set after removal.

[0075] In some optional implementations of some embodiments, the execution subject can generate the attribute anomaly standard information for the product use attribute according to the product attribute information set, the target attribute anomaly determination rule, and the target attribute anomaly generation model, including the following steps:

[0076] In the first step, the rule accuracy corresponding to the target attribute anomaly determination rule and the model accuracy corresponding to the target attribute anomaly generation model are determined. The rule accuracy and the model accuracy are both numerical values between 0 and 1.

[0077] In the second step, the information of each product attribute information in the product attribute information set is enhanced to obtain an enhanced product attribute information set.

[0078] Thirdly, in response to determining that the model accuracy is higher than or equal to the rule accuracy, the target attribute anomaly determination rule is used to determine the first anomaly information corresponding to each product attribute information in the enhanced product attribute information set, to obtain a first anomaly information set. The first anomaly information can represent information that the attribute content corresponding to the product attribute information is abnormal content. In practice, the first anomaly information can be numerical information or label information.

[0079] Fourthly, the first anomaly information in the first anomaly information set and the product attribute information in the product attribute information set are correspondingly bound to generate bound information, to obtain a bound information set.

[0080] Fifthly, the bound information set and the target attribute anomaly information generation model are sent to a target node in the graphics processor cluster, so that the target node executes the following steps:

[0081] Firstly, the target attribute anomaly information generation model and the bound information set are obtained.

[0082] Secondly, at least one similar model corresponding to the target attribute anomaly information generation model is obtained from the model storage library according to the remaining computing resources corresponding to the target node. The sum of the application resources corresponding to the at least one similar model is less than or equal to the remaining computing resources. The similar model can be a neural network model similar to the target attribute anomaly information generation model in model accuracy and network structure.

[0083] Thirdly, for each similar model in the at least one similar model, each bound information in the bound information set is input into the similar model to generate second anomaly information, to obtain a second anomaly information set.

[0084] Fourthly, the obtained at least one second anomaly information set is sent to the generation node corresponding to the attribute anomaly standard information.

[0085] Sixthly, each bound information in the bound information set is input into the target attribute anomaly generation model to generate third anomaly information, to obtain a third anomaly information set.

[0086] Seventhly, for each bound information in the bound information set, the following determination steps are executed:

[0087] Sub-step 1, determining the third anomaly information and the at least one second anomaly information corresponding to the bound information.

[0088] Sub-step 2, determine the comprehensive abnormal information for the third abnormal information and the at least one second abnormal information according to the at least one model weight corresponding to the at least one similar model and the model weight corresponding to the target attribute abnormal information. The model weight in the at least one model weight has a one-to-one correspondence with the similar model in the at least one similar model. The model weight can represent the importance of the model output content.

[0089] As an example, the execution subject can perform weighted processing on the third abnormal information and the at least one second abnormal information according to the at least one model weight corresponding to the at least one similar model and the model weight corresponding to the target attribute abnormal information to generate the comprehensive abnormal information.

[0090] Step 8, filter the comprehensive abnormal information representing the existence of an abnormality from the obtained comprehensive abnormal information set as a target comprehensive abnormal information, and obtain at least one target comprehensive abnormal information.

[0091] Step 9, determine the attribute information range corresponding to the at least one target comprehensive abnormal information as a target attribute information range.

[0092] As an example, the execution subject can determine at least one product attribute information corresponding to the at least one target comprehensive abnormal information as at least one target product attribute information. Then, determine the attribute information range corresponding to the at least one target product attribute information as the target attribute information range.

[0093] Step 10, set attribute abnormal standard information representing that attribute information is abnormal information within the target attribute information range and is non-abnormal information outside the target attribute information range.

[0094] Step 11, in response to determining that the model accuracy is less than the rule accuracy, determine the fourth abnormal information corresponding to each product attribute information in the enhanced product attribute information set using the target attribute abnormal generation model to obtain a fourth abnormal information set.

[0095] Step 12, for each product attribute information in the product attribute information set, perform the following generation steps:

[0096] Sub-step 1, determine the first abnormal information and the fourth abnormal information corresponding to the product attribute information as the first target abnormal information and the fourth target abnormal information, respectively.

[0097] Sub-step 2, in response to determining that the first target abnormal information and the fourth target abnormal information both represent an abnormality, generate judgment information representing that the product attribute information has an abnormality.

[0098] Sub-step 3, in response to determining that the first target abnormal information and the fourth target abnormal information exist abnormal information representing no abnormality, generating judgment information representing that the above product attribute information does not exist abnormality.

[0099] Thirteenth step, according to the obtained judgment information, determining at least one product attribute information representing that there exists abnormality as at least one abnormal product attribute information.

[0100] Fourteenth step, determining the attribute information range corresponding to the above at least one abnormal product attribute information as an abnormal attribute information range.

[0101] Fifteenth step, generating attribute abnormality standard information according to the above abnormal attribute information range.

[0102] The above "in some optional implementations of some embodiments", as one of the invention points, solves another technical problem of the present disclosure "cannot fully utilize the target attribute abnormality determination rule and the corresponding abnormal information generation resource of the above target attribute abnormality generation model, resulting in inaccurate attribute abnormality standard information generation". Based on this, the present disclosure first determines the generation method of the corresponding attribute abnormality standard information by determining the respective accuracy of the target attribute abnormality determination rule and the target attribute abnormality generation model. Then, on the basis that the accuracy of the target attribute abnormality determination rule is higher than or equal to that of the target attribute abnormality generation model, the abnormal information set corresponding to the enhanced product attribute information set after attribute information enhancement is determined by using the target attribute abnormality determination rule first (that is, the first abnormal information set). Then, considering the limited resources of the node currently generating the attribute abnormality standard information, at least one second abnormal information set corresponding to the binding information set generated based on the first abnormal information set is determined by at least one similar node at the target node to further supplement the information amount of the abnormal information. Based on this, the comprehensive abnormal information corresponding to each binding information can be accurately generated. On the basis of determining to generate at least one comprehensive abnormal information, the comprehensive abnormal information existing abnormality is determined by screening to determine the corresponding target attribute information range. On the basis that the accuracy of the target attribute abnormality determination rule is less than that of the target attribute abnormality generation model, the judgment information of whether the product attribute information exists abnormality can be determined by comprehensively considering the abnormal information corresponding to the output of the target attribute abnormality determination rule and the abnormal information corresponding to the output of the target attribute abnormality generation model. Finally, according to the obtained judgment information set, the abnormal attribute information range can be accurately determined to further determine the attribute abnormality standard information.

[0103] Step 105, generating the first standard information portrait corresponding to the obtained at least one attribute abnormality standard information.

[0104] In some embodiments, the execution subject can generate a first standard information portrait corresponding to the resulting at least one attribute anomaly standard information. The first standard information portrait and the object portrait are of the same tree structure. The first standard portrait can represent attribute anomaly judgment standards corresponding to each product use attribute under at least one product use attribute.

[0105] Step 106, store the first standard information portrait and the removed object portrait set in a target storage terminal.

[0106] In some embodiments, the execution subject can store the first standard information portrait and the removed object portrait set in a target storage terminal. The target storage terminal can be a pre-determined storage terminal.

[0107] In some optional implementations of some embodiments, after step 107, the steps further include:

[0108] First, in response to determining that the object data representing the request to apply for the target virtual product is received, the object data is pre-processed to obtain pre-processed object data. The data preprocessing can include but is not limited to one of the following: data blur removal, data missing removal.

[0109] Second, generate a target object portrait of the same structure format as the removed object portrait according to the pre-processed object data. Third, match the content corresponding to the target object portrait with the standard content corresponding to the first standard information portrait to generate a matching result portrait.

[0110] Fourth, obtain a current attribute emphasis direction selected from the attribute emphasis direction set. Different attribute emphasis directions correspond to different attribute scoring portraits. The attribute scoring portrait corresponds to a tree structure that is the same as the tree structure corresponding to the first standard information portrait. The current attribute emphasis direction can represent each attribute that is emphasized for monitoring. That is, the current attribute emphasis direction can include each key attribute that is emphasized for monitoring. The current attribute emphasis direction can represent each attribute that is emphasized for attention.

[0111] Fifth, determine the attribute scoring portrait corresponding to the current attribute emphasis direction as a first target attribute scoring portrait.

[0112] Sixth, score the matching result portrait according to the first target attribute scoring portrait to generate a first anomaly score.

[0113] As an example, first, the above execution subject can score each attribute information in the matching result image according to the attribute anomaly standard information corresponding to each attribute in the first target attribute scoring image, to generate each anomaly score. Then, the anomaly scores are weighted to generate a first anomaly score.

[0114] In the seventh step, a first product use anomaly report corresponding to the target object is generated according to the first anomaly score and the first target attribute scoring image. The target object corresponds to the object data. The first product use anomaly report can be an anomaly report of the target product use.

[0115] In the eighth step, the first product use anomaly report is sent to the anomaly auditing terminal in an encrypted form for further product use anomaly auditing.

[0116] In some optional implementations of some embodiments, the steps further include:

[0117] In the first step, in response to determining that the selected attribute emphasis direction does not exist in the attribute emphasis direction set, an attribute image corresponding to the first standard information image is displayed on a target attribute filtering interface for a target person to manually arrange a combination of attributes on the target attribute filtering interface. The attribute image is a tree structure image including each attribute. The arrangement of the combination of attributes is a component dragging manner in the interface. The target attribute filtering interface can be an interface for displaying the attribute image for attribute filtering. The unimportant attributes in the attribute image can be removed through the component dragging manner.

[0118] In the second step, the attribute combination and combination naming information selected by the target person on the target attribute filtering interface are obtained. The combination naming information can be a naming identifier corresponding to the attribute combination.

[0119] In the third step, an attribute scoring image corresponding to the attribute combination is determined as a second target attribute scoring image.

[0120] In the fourth step, the matching result image is scored according to the second target attribute scoring image to generate a second anomaly score. Details are not repeated, and refer to the determination of the first anomaly score.

[0121] In the fifth step, a second product use anomaly report corresponding to the target object is generated according to the second anomaly score and the second target attribute scoring image.

[0122] In the sixth step, the second product use anomaly report is sent to the anomaly auditing terminal in an encrypted form for further product use anomaly auditing.

[0123] In a seventh step, the above attribute group and the above combination naming information are added to the above attribute focus direction set in the form of key-value pairs for storage.

[0124] The above various embodiments of the present disclosure have the following beneficial effects: through the virtual product standard portrait storage method of some embodiments of the present disclosure, the standard information portrait for abnormal judgment of the target virtual product can be efficiently and accurately generated in the case of adapting to the current remaining computing resources. Specifically, the reason for the insufficient accuracy and timeliness of the related abnormal judgment is that the product usage dataset is often large in data volume, resulting in a waste of a large amount of computing time when directly input into the multi-attribute abnormal standard information output model, and causing the multi-attribute abnormal standard information output model to learn more redundant features. And in the case that the multi-attribute abnormal standard information output model needs a large amount of computing resources, if the terminal cannot effectively supply computing resources, it will cause the output to be interrupted and affect the generation of abnormal judgment standards. Based on this, the virtual product standard portrait storage method of some embodiments of the present disclosure first acquires the product usage dataset of the target virtual product within a predetermined time period, which is used as the basic dataset for abnormal judgment of each product usage attribute. Based on this, the accurate generation of the virtual product standard portrait can be realized. Then, for each product usage object in the product usage object set, the object portrait corresponding to the product usage object can be accurately generated according to the product usage data corresponding to the product usage object. Through the object portrait, the object portrait of each usage object corresponding to the target virtual product is determined, which facilitates the subsequent query of the product usage object related information. Next, the product usage attribute set corresponding to the obtained object portrait set is subjected to attribute and attribute information pruning to obtain at least one product usage attribute and a removed object portrait set, so as to determine at least one product usage attribute for key monitoring and a removed object portrait set including key product usage attribute information. Then, for each product usage attribute in the at least one product usage attribute, a distributed graphics processor cluster is used to perform the following generation steps: first, determine the product attribute information set in the object portrait set corresponding to the product usage attribute, so as to facilitate the generation of attribute abnormal standard information corresponding to the product usage attribute. In addition, the distributed graphics processor cluster can effectively provide computing resources available for computing. Second, acquire the pre-stored attribute abnormal determination rule set and attribute abnormal generation model set corresponding to the product usage attribute, so as to jointly determine the abnormal judgment standard from the rule and neural network model angles, which can greatly improve the accuracy of the judgment. Third, according to the remaining computing resources corresponding to the graphics processor cluster, the target attribute abnormal determination rule and the target attribute abnormal generation model are selected from the attribute abnormal determination rule set and the attribute abnormal generation model set, respectively, so as to effectively select the rules and models for subsequent real-time generation of attribute abnormal standard information in the case of current computing resources. Fourth, according to the product attribute information set, the target attribute abnormal determination rule and the target attribute abnormal generation model can be used to accurately generate attribute abnormal standard information for the product usage attribute.Further, a first standard information portrait corresponding to the generated at least one attribute anomaly standard information is generated, wherein the first standard information portrait and the object portrait are the same tree structure. Here, through the form of the first standard information portrait, the efficiency of the anomaly determination can be improved and the determination time is saved during subsequent anomaly determination. Finally, the first standard information portrait and the object portrait set after the removal are stored in a target storage terminal, so as to facilitate the execution of subsequent anomaly determination and the periodic update of the first standard information portrait.

[0125] Further reference Figure 2 , as an implementation of the method shown in the above figures, the disclosure provides some embodiments of a virtual product standard portrait storage device, which corresponds to the method embodiments shown in Figure 1 , the virtual product standard portrait storage device can be applied to various electronic devices.

[0126] As Figure 2 shown, a virtual product standard portrait storage device 200 includes an acquisition unit 201, a first generation unit 202, a clipping unit 203, an execution unit 204, a second generation unit 205, and a storage unit 206. Wherein, the acquisition unit 201 is configured to acquire a product use data set of a target virtual product within a predetermined time period; the first generation unit 202 is configured to, for each product use object in the product use object set, generate an object portrait corresponding to the product use object according to the product use data corresponding to the product use object; the clipping unit 203 is configured to perform attribute and attribute information clipping on the product use attribute set corresponding to the obtained object portrait set, to obtain at least one product use attribute and an object portrait set after removal; the execution unit 204 is configured to, for each product use attribute in the at least one product use attribute, execute the generation step by using a distributed graphics processor cluster: determine the product attribute information set in the object portrait set corresponding to the product use attribute; acquire the pre-stored attribute anomaly determination rule set and attribute anomaly generation model set corresponding to the product use attribute; according to the calculation remaining resources corresponding to the graphics processor cluster, the target attribute anomaly determination rule and the target attribute anomaly generation model are respectively screened from the attribute anomaly determination rule set and the attribute anomaly generation model set; using the target attribute anomaly determination rule and the target attribute anomaly generation model, generate attribute anomaly standard information for the product use attribute; the second generation unit 205 is configured to generate a first standard information portrait corresponding to the generated at least one attribute anomaly standard information, wherein the first standard information portrait and the object portrait are the same tree structure; the storage unit 206 is configured to store the first standard information portrait and the object portrait set after the removal in a target storage terminal.

[0127] It is understood that the units described in the virtual product standard image storage device 200 correspond to the reference Figure 1 The individual steps in the described method correspond. Thus, the operations, features and resulting advantages described above for the method equally apply to the virtual product standard image storage device 200 and the units contained therein, which are not described again here.

[0128] Reference is made below to Figure 3 which shows a structural schematic diagram of an electronic device (e.g., electronic device) 300 suitable for implementing some embodiments of the present disclosure. Figure 3 The electronic device shown is merely an example and should not bring any limitation to the function and scope of use of the embodiments of the present disclosure.

[0129] As shown in Figure 3 , the electronic device 300 can include a processing device (e.g., central processor, graphics processor, etc.) 301 that can perform various appropriate actions and processes according to programs stored in a read-only memory (ROM) 302 or programs loaded from a storage device 308 into a random access memory (RAM) 303. In the RAM 303, various programs and data required for the operation of the electronic device 300 are also stored. The processing device 301, the ROM 302, and the RAM 303 are connected to each other through a bus 304. An input / output (I / O) interface 305 is also connected to the bus 304.

[0130] In general, the following devices can be connected to the I / O interface 305: input devices 306 including, for example, a touch screen, a touchpad, a keyboard, a mouse, a camera, a microphone, an accelerometer, a gyroscope, etc.; output devices 307 including, for example, a liquid crystal display (LCD), a speaker, a vibrator, etc.; storage devices 308 including, for example, a magnetic tape, a hard disk, etc.; and communication devices 309. The communication devices 309 can allow the electronic device 300 to communicate wirelessly or wired with other devices to exchange data. Although Figure 3 The electronic device 300 is shown with various devices, but it is understood that not all of the devices shown are required to be implemented or provided. More or fewer devices can alternatively be implemented or provided. Figure 3 Each block shown in

[0131] In particular, according to some embodiments of the present disclosure, the processes described above with reference to the flowcharts can be implemented as a computer software program. For example, some embodiments of the present disclosure include a computer program product comprising a computer program carried on a computer readable medium, the computer program containing program code for performing the methods illustrated by the flowcharts. In some such embodiments, the computer program can be downloaded and installed from a network via the communication device 309, or installed from the storage device 308, or installed from the ROM 302. When the computer program is executed by the processing device 301, the above-mentioned functions defined in the methods of some embodiments of the present disclosure are performed.

[0132] It should be noted that the computer readable medium mentioned above in some embodiments of the present disclosure can be a computer readable signal medium or a computer readable storage medium or any combination of the two. The computer readable storage medium may, for example, but not limited to, an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, device or apparatus, or any combination of the above. More specific examples of the computer readable storage medium can include, but are not limited to, an electrical connection having one or more wires, a portable computer diskette, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the above. In some embodiments of the present disclosure, the computer readable storage medium can be any tangible medium containing or storing a program that can be used by or in conjunction with an instruction execution system, device or apparatus. While in some embodiments of the present disclosure, the computer readable signal medium can include a data signal carried in a baseband or as a part of a carrier wave, which carries computer readable program code. Such a propagated data signal can take many forms, including but not limited to an electromagnetic signal, an optical signal or any suitable combination of the above. The computer readable signal medium can also be any computer readable medium other than the computer readable storage medium that can send, propagate or transmit a program for use by or in connection with an instruction execution system, device or apparatus. The program code contained on the computer readable medium can be transmitted by any suitable medium, including but not limited to a wire, an optical fiber, an RF (radio frequency) or the like, or any suitable combination of the above.

[0133] In some embodiments, the client, server, and other components can communicate using any known or later developed form of computer readable media, including a wireless medium, over a propagation medium that facilitates wired or wireless electronic communication transmission, over a local area network, a wide area network, the Internet, or any other form of communication medium. In some embodiments, the client, server, and other components can utilize various forms of communication protocols, including HTTP (HyperText Transfer Protocol), to communicate over the Internet.

[0134] The computer readable medium described above can be included in the electronic device described above; or can exist separately from the electronic device and be not assembled into the electronic device. The computer readable medium described above carries one or more programs, when the one or more programs are executed by the electronic device, cause the electronic device to: obtain a product use data set of a target virtual product in a predetermined time period; for each product use object in the product use object set, generate an object portrait corresponding to the product use object according to the product use data corresponding to the product use object; perform attribute and attribute information pruning on the product use attribute set corresponding to the obtained object portrait set, to obtain at least one product use attribute and a post-removal object portrait set; for each product use attribute in the at least one product use attribute, utilize a distributed graphics processor cluster to perform a generation step: determine a product attribute information set corresponding to the product use attribute in the object portrait set; obtain a pre-stored attribute anomaly determination rule set and an attribute anomaly generation model set corresponding to the product use attribute; according to the computing remaining resources corresponding to the graphics processor cluster, screen out a target attribute anomaly determination rule and a target attribute anomaly generation model from the attribute anomaly determination rule set and the attribute anomaly generation model set, respectively; according to the product attribute information set, utilize the target attribute anomaly determination rule and the target attribute anomaly generation model to generate attribute anomaly standard information for the product use attribute; generate a first standard information portrait corresponding to the obtained at least one attribute anomaly standard information, wherein the first standard information portrait and the object portrait are of the same tree structure; and store the first standard information portrait and the post-removal object portrait set in a target storage terminal.

[0135] Computer program code for carrying out operations of some embodiments of the present disclosure can be written in any combination of one or more programming languages, including an object oriented programming language such as Java, Smalltalk, C++ or the like, and conventional procedural programming languages, such as the "C" programming language or similar programming languages. The program code can execute entirely on the user's computer, partly on the user's computer, as a stand-alone software package, partly on the user's computer and partly on a remote computer or entirely on the remote computer or server. In the latter scenario, the remote computer can be connected to the user's computer through any type of network, including a local area network (LAN) or a wide area network (WAN), or the connection can be made to an external computer (for example, through the Internet using an Internet Service Provider).

[0136] The computer program instructions can also be loaded onto a computer or other programmable information processing apparatus to cause a series of operations to be performed on the computer or other programmable information processing apparatus to produce a computer implemented process such that the instructions which execute on the computer or other programmable information processing apparatus implement the functions / acts specified in the flowchart and / or block diagram block or blocks.

[0137] The units described in some embodiments of the present disclosure can be implemented by software, or can be implemented by hardware. The described units can also be arranged in a processor, for example, can be described as: a processor includes an acquisition unit, a first generation unit, a clipping unit, an execution unit, a second generation unit and a storage unit. Among them, the name of these units does not constitute the limitation of the unit itself in some cases, for example, the acquisition unit can also be described as "a unit for acquiring the product use data set of the target virtual product in a predetermined time period".

[0138] The functionality described herein above can be performed, at least in part, by one or more hardware logic components. For example, and without limitation, illustrative types of hardware logic components that can be used include Field-programmable Gate Arrays (FPGAs), Program- specific Integrated Circuits (ASICs), Program- specific Standard Products (ASSPs), System-on-a-chip systems (SOCs), Complex Programmable Logic Devices (CPLDs), etc.

[0139] The above description is merely exemplary of the disclosure and the application made use of the principles of the technology. It is to be understood that the application is not limited in scope to the described technical features, and that the application can be practiced with modification other than those described in the embodiment discussed in the context of the above description of the best mode. For example, it will be appreciated that those skilled in the art, on the basis of the principles described herein, can effect modifications or alterations, such as but not limited to a combination of technical features described above or equivalent technical features, without departing from the scope of the application as defined by the appended claims.

Claims

1. A virtual product standard image storage method, comprising: obtaining a product use data set of a target virtual product within a predetermined time period; for each product use object in the product use object set, generating an object image corresponding to the product use object according to the product use data corresponding to the product use object; performing attribute and attribute information pruning on the product use attribute set corresponding to the obtained object image set to obtain at least one product use attribute and a post-removal object image set; for each product use attribute in the at least one product use attribute, using a distributed graphics processor cluster to perform the following generation steps: determining a product attribute information set corresponding to the product use attribute in the object image set; obtaining a pre-stored attribute anomaly determination rule set and an attribute anomaly generation model set corresponding to the product use attribute; determining a calculation remaining resource sequence of the graphics processor cluster within a target future time period; determining an attribute division resource sequence corresponding to the product use attribute from the calculation remaining resource sequence; determining a collocation combination set between the attribute anomaly determination rule set and the attribute anomaly generation model set; for each collocation combination in the collocation combination set, performing the following generation steps: determining a pre-estimated consumption time length and a pre-estimated consumption resource corresponding to the attribute anomaly determination rule and the attribute anomaly generation model in the collocation combination; determining whether the attribute division resource sequence meets the consumption requirements corresponding to the pre-estimated consumption time length and the pre-estimated consumption resource to obtain a satisfaction information; from the obtained satisfaction information set, screening out satisfaction information representing that the consumption requirements corresponding to the pre-estimated consumption time length and the pre-estimated consumption resource are met to obtain at least one satisfaction information; from the at least one collocation combination, screening out a collocation combination with the highest comprehensive accuracy and the shortest pre-estimated consumption time length as a target collocation combination, wherein the target collocation combination includes a target attribute anomaly determination rule and a target attribute anomaly generation model, and the at least one satisfaction information and the at least one collocation combination have a one-to-one correspondence; generating attribute anomaly standard information for the product use attribute according to the product attribute information set, using the target attribute anomaly determination rule and the target attribute anomaly generation model; generating a first standard information image corresponding to the obtained at least one attribute anomaly standard information, wherein the first standard information image and the object image are of the same tree structure; storing the first standard information image and the post-removal object image set in a target storage terminal.

2. The method of claim 1, wherein, The attribute and attribute information pruning on the product use attribute set corresponding to the obtained object image set to obtain at least one product use attribute and a post-removal object image set, comprises: generating attribute prediction effective information corresponding to each product use attribute in the product use attribute set to obtain an attribute prediction effective information set, wherein the attribute prediction effective information is an effective numerical value representing attribute importance and attribute prediction ability strength; removing product use attributes with attribute prediction effective information less than a target value from the product use attribute set to obtain at least one product use attribute; For each object image in the object image set, remove attribute information corresponding to the at least one product use attribute from the object image to obtain a removed object image.

3. The method of claim 2, wherein, The method further comprises: In response to determining that object data representing a request to apply for the target virtual product is received, performing data preprocessing on the object data to obtain preprocessed object data; According to the preprocessed object data, generate a target object image corresponding to the removed object image in the same structural format; Match the content corresponding to the target object image with the standard content corresponding to the first standard information image to generate a matching result image; Obtain a selected current attribute emphasis direction in an attribute emphasis direction set, wherein different attribute emphasis directions correspond to different attribute score images, and the attribute score image corresponds to a tree structure identical to the tree structure corresponding to the first standard information image; Determine the attribute score image corresponding to the current attribute emphasis direction as a first target attribute score image; According to the first target attribute score image, score processing is performed on the matching result image to generate a first abnormal score; According to the first abnormal score and the first target attribute score image, generate a first product use abnormal report corresponding to the target object, wherein the target object corresponds to the object data; The first product use abnormal report is sent to an abnormality auditing terminal in an encrypted form for further product use abnormality auditing.

4. The method of claim 3, wherein, The method further comprises: In response to determining that the selected attribute emphasis direction does not exist in the attribute emphasis direction set, display the attribute image corresponding to the first standard information image on a target attribute screening interface for a target person to manually allocate combinations of attributes on the target attribute screening interface, wherein the attribute image is a tree structure corresponding to an image including each attribute, and the allocation mode of each attribute combination is a component dragging mode in the interface; Obtain the attribute combination and combination naming information selected by the target person on the target attribute screening interface; Determine the attribute score image corresponding to the attribute combination as a second target attribute score image; According to the second target attribute score image, score processing is performed on the matching result image to generate a second abnormal score; According to the second abnormal score and the second target attribute score image, generate a second product use abnormal report corresponding to the target object; The second product use abnormal report is sent to an abnormality auditing terminal in an encrypted form for further product use abnormality auditing; The attribute combination and the combination naming information are added to the attribute emphasis direction set in the form of key-value pairs for storage.

5. The method of claim 1, wherein, The method further comprises: Determine the attribute weight corresponding to each product use attribute in the at least one product use attribute to obtain at least one attribute weight; According to the at least one attribute weight, each computing residual resource in the computing residual resource sequence is correspondingly divided to generate available computing resources of the product use attribute at a corresponding time, obtaining at least one attribute divided resource sequence; From the at least one attribute divided resource sequence, an attribute divided resource sequence corresponding to the product use attribute is screened out.

6. The method of claim 1, wherein, According to the product attribute information set, the target attribute anomaly determination rule and the target attribute anomaly generation model are used to generate attribute anomaly standard information for the product use attribute, including: In response to determining that the comprehensive accuracy corresponding to the target collocation combination corresponding to the target attribute anomaly determination rule and the target attribute anomaly generation model is not the highest collocation combination in the collocation combination set, waiting for the computing residual resources corresponding to the graphics processor cluster to meet the resource requirements of the collocation combination with the highest accuracy; In response to determining that the meeting is met, according to the product attribute information set, using the computing residual resources corresponding to the graphics processor cluster, using the highest collocation combination, generating alternative attribute anomaly standard information for the product use attribute; And the method further comprises: Generating a second standard information portrait corresponding to the obtained at least one alternative attribute anomaly standard information, wherein the second standard information portrait and the object portrait are the same tree structure; Storing the second standard information portrait and the removed object portrait set in the target storage terminal to replace the first standard information portrait and the removed object portrait set.

7. A virtual product standard portrait storage device, comprising: An acquisition unit configured to acquire a product use data set of a target virtual product within a predetermined time period; A first generation unit configured to, for each product use object in a product use object set, generate an object portrait corresponding to the product use object according to product use data corresponding to the product use object; A clipping unit configured to perform attribute and attribute information clipping on a product use attribute set corresponding to the obtained object portrait set, obtaining at least one product use attribute and a removed object portrait set; An execution unit configured to, for each product use attribute in the at least one product use attribute, use a distributed graphics processor cluster to perform a generation step of determining a product attribute information set corresponding to the product use attribute in the object portrait set; obtaining a set of attribute exception determination rules and a set of attribute exception generation models corresponding to the product usage attribute; determining a sequence of remaining calculation resources of the set of graphics processors in a target future time period; determining an attribute partition resource sequence corresponding to the product usage attribute from the sequence of remaining calculation resources; determining a set of collocation combinations between the set of attribute exception determination rules and the set of attribute exception generation models; for each collocation combination in the set of collocation combinations, performing the following generation steps: determining a predicted consumption time length and a predicted consumption resource corresponding to the attribute exception determination rule and the attribute exception generation model in the collocation combination; determining whether the attribute partition resource sequence meets the consumption requirements corresponding to the predicted consumption time length and the predicted consumption resource to obtain meeting information; screening, from the obtained set of meeting information, meeting information representing that the consumption requirements corresponding to the predicted consumption time length and the predicted consumption resource are met to obtain at least one meeting information; screening, from the at least one collocation combination, a collocation combination with the highest comprehensive accuracy and the shortest predicted consumption time length as a target collocation combination, wherein the target collocation combination includes a target attribute exception determination rule and a target attribute exception generation model, and the at least one meeting information and the at least one collocation combination have a one-to-one correspondence; and generating attribute exception standard information for the product usage attribute by using the target attribute exception determination rule and the target attribute exception generation model according to the set of product attribute information; a second generation unit configured to generate a first standard information portrait corresponding to the obtained at least one attribute exception standard information, wherein the first standard information portrait and the object portrait are of the same tree structure; a storage unit configured to store the first standard information portrait and the set of removed object portraits in a target storage terminal. 8.An electronic device, comprising: one or more processors; a storage having one or more programs stored thereon, when the one or more programs are executed by the one or more processors, the one or more processors implement the method of any one of claims 1-6.

9. A computer readable medium having stored thereon a computer program, wherein, The program is executed by the processor to implement the method of any one of claims 1-6.

Images