Data analysis method, electronic equipment, storage medium and program product

Abstract

The invention provides a data analysis method, electronic equipment, a storage medium and a program product, relates to the technical field of computers, and is used for united business data analysis in a multi-cloud scene. The method comprises the following steps: receiving business demand information from a cloud service consumer; the business demand information comprises to-be-analyzed business data; sending the initial model to at least one target cloud service provider in response to the business demand information of the cloud service consumer; the storage data in the target cloud service provider and the to-be-analyzed business data meet a data category matching condition; receiving intermediate model parameters correspondingly sent by at least one target cloud service provider, and aggregating the intermediate model parameters to obtain a target model; the intermediate model parameters are obtained by pre-training the initial model by the corresponding target cloud service provider based on the storage data; and analyzing the to-be-analyzed business data through the target model, and determining a data analysis result.

Description

Technical Field

[0001] The present application relates to the field of computer technology, and in particular to a data analysis method, electronic equipment, storage medium and program product. Background Art

[0002] With the continuous development of big data and artificial intelligence technologies, data has become an important basis for corporate decision-making. Enterprises are paying more and more attention to the collection, analysis and utilization of data to formulate more accurate and effective business strategies. Therefore, enterprises have increasing requirements for data storage, processing and analysis capabilities, and need more efficient, secure and scalable data storage and service deployment solutions.

[0003] The popularity of multi-cloud environments is an effective strategy for enterprises to cope with current complex and changing business needs and technical challenges. Multi-cloud environments provide higher flexibility, reliability and scalability, helping enterprises optimize resource allocation, avoid provider lock-in and improve business continuity.

[0004] However, multi-cloud scenarios also bring about problems such as inconsistent data types across platforms and data leakage during data transmission. Summary of the invention

[0005] The present application provides a data analysis method, electronic device, storage medium and program product for joint data analysis in a multi-cloud scenario.

[0006] In a first aspect, the present application provides a data analysis method, which is applied to a multi-cloud manager (Cloud Service Partner, CSN), comprising: receiving business demand information from a cloud service consumer (Cloud Service Customer, CSC); the business demand information includes business data to be analyzed; in response to the business demand information of the cloud service consumer, sending an initial model to at least one target cloud service provider (Cloud Service Provider, CSP); the storage data in the target cloud service provider and the business data to be analyzed meet the data category matching condition; receiving intermediate model parameters sent by each of the at least one target cloud service provider, and aggregating the intermediate model parameters to obtain a target model; the intermediate model parameters are obtained by pre-training the initial model based on the storage data of the corresponding target cloud service provider; analyzing the business data to be analyzed by using the target model to determine the data analysis result.

[0007] It can be understood that by receiving business demand information from cloud service consumers, the target cloud service provider can be quickly determined, thereby improving business response speed. By setting data category matching conditions, the cloud consumption provider can be flexibly selected based on business demand information, thereby improving the flexibility of the data analysis method. The target model obtained through multiple intermediate model parameters can average the deviations between multiple intermediate model parameters, thereby improving the accuracy of data analysis.

[0008] In a possible implementation, the data category matching condition includes: the data type of the stored data in the cloud service provider is the same as the data type of the business data to be analyzed.

[0009] In another possible implementation, the data category matching condition includes: the business type of the data stored in the cloud service provider is the same as the business type of the business data to be analyzed.

[0010] Another possible implementation manner is that the target cloud service provider is a cloud service provider that has passed security certification.

[0011] In another possible implementation manner, the method further includes: obtaining the analysis task duration of the business data to be analyzed; when the analysis task duration is greater than a duration threshold, periodically performing security authentication on the target cloud service provider at a preset interval.

[0012] Another possible implementation method is that the business demand information also includes the analysis task start time and analysis task deadline of the business data to be analyzed; obtaining the analysis task duration of the business data to be analyzed includes: determining the analysis task duration based on the analysis task start time and analysis task deadline of the business data to be analyzed.

[0013] In another possible implementation, the method further includes: performing security authentication on the cloud service provider in the following manner: sending an authentication request message to the cloud service provider, the authentication request message including a group element randomly selected from a cyclic group; the modulus of the cyclic group is a prime number; receiving authentication response information from the cloud service provider; the authentication response information includes a first parameter and a second parameter; wherein the first parameter is the result of a first modular exponentiation operation; the base of the first modular exponentiation operation is a generator of the cyclic group; the exponent of the first modular exponentiation operation is an integer randomly selected by the cloud service provider; the modulus of the first modular exponentiation operation is the modulus of the cyclic group ; The second parameter is the result of the second modular exponentiation operation; the base of the second modular exponentiation operation is the product of the integer key held by the cloud service provider and the first parameter; the exponent of the second modular exponentiation operation is 1; the modulus of the second modular exponentiation operation is the modulus of the cyclic group minus 1; a group element randomly selected from the cyclic group and the integer key satisfy the third modular exponentiation operation; the result of the third modular exponentiation operation is the group element; the base of the third modular exponentiation operation is the generator of the cyclic group; the exponent of the third modular exponentiation operation is the integer key; the modulus of the third modular exponentiation operation is the modulus of the cyclic group; based on the first parameter and the second parameter, determine whether the cloud service provider has passed the security authentication.

[0014] Another possible implementation method is to determine whether the cloud service provider has passed the security authentication based on the first parameter and the second parameter, including: calculating the verification value according to the fourth modular exponentiation operation based on the first parameter and the second parameter; wherein the base of the fourth modular exponentiation operation is the third parameter; the exponent of the fourth modular exponentiation operation is 1, and the modulus of the fourth modular exponentiation operation is the modulus of the cyclic group; the third parameter is the product of the fourth parameter and the fifth parameter; the base of the fourth parameter is a group element randomly selected from the cyclic group; the exponent of the fourth parameter is the first parameter; the base of the fifth parameter is the first parameter; the exponent of the fifth parameter is the second parameter; when the verification value is equal to the first parameter, it is determined that the cloud service provider has passed the security authentication; when the verification value is not equal to the first parameter, it is determined that the cloud service provider has not passed the security authentication.

[0015] Another possible implementation is to aggregate the intermediate model parameters to obtain the target model, including: taking a weighted average based on each intermediate model parameter and the corresponding weight value to obtain the target model parameters; and determining the target model based on the target model parameters.

[0016] Another possible implementation method is to determine the target model based on the target model parameters, including: substituting the target model parameters into the initial model to obtain a reference model; and determining the reference model as the target model when the reference model meets the training stop condition.

[0017] In another possible implementation, the training stop condition includes at least one of the following: the analysis task deadline indicated by the business demand information is reached at the current moment; the number of times the target model is substituted into the initial model to obtain the reference model reaches a number threshold; the output error of the reference model is less than or equal to the error threshold.

[0018] Another possible implementation method is to analyze the business data to be analyzed through the target model to determine the data analysis results, including: analyzing the business data to be analyzed through the target model to determine the initial analysis results; performing data processing on the initial analysis results to determine the data analysis results; the data processing includes at least one of the following: data correction, result desensitization, generating reports according to preset templates, and visual data conversion.

[0019] In the second aspect, the present application provides a data analysis device, including: a data security analysis module, a task scheduling module and a result processing module, the data security analysis module is used to receive business demand information from cloud service consumers; the business demand information includes business data to be analyzed; in response to the business demand information of the cloud service consumer, an initial model is sent to at least one target cloud service provider; the storage data in the target cloud service provider and the business data to be analyzed meet the data category matching conditions; the task scheduling module is used to receive intermediate model parameters sent by each corresponding one of the at least one target cloud service provider; the intermediate model parameters are aggregated to obtain a target model; the intermediate model parameters are obtained by pre-training the initial model based on the storage data by the corresponding target cloud service provider; the result processing module is used to analyze the business data to be analyzed through the target model to determine the data analysis results.

[0020] In a possible implementation, the data category matching condition includes: the data type of the stored data in the cloud service provider is the same as the data type of the business data to be analyzed.

[0021] In a possible implementation, the data category matching condition includes: the business type of the data stored in the cloud service provider is the same as the business type of the business data to be analyzed.

[0022] Another possible implementation manner is that the target cloud service provider is a cloud service provider that has passed security certification.

[0023] Another possible implementation method is that the data security analysis module is also used to obtain the analysis task duration of the business data to be analyzed; when the analysis task duration is greater than the duration threshold, the target cloud service provider is periodically security authenticated at a preset interval.

[0024] Another possible implementation method is that the business demand information also includes the analysis task start time and analysis task deadline of the business data to be analyzed; the data security analysis module is specifically used to determine the analysis task duration based on the analysis task start time and analysis task deadline of the business data to be analyzed.

[0025] In another possible implementation, the data security analysis module is specifically used to perform security authentication on the cloud service provider in the following manner: sending an authentication request message to the cloud service provider, the authentication request message includes a group element randomly selected from a cyclic group; the modulus of the cyclic group is a prime number; receiving authentication response information from the cloud service provider; the authentication response information includes a first parameter and a second parameter; wherein the first parameter is the result of a first modular exponentiation operation; the base of the first modular exponentiation operation is a generator of the cyclic group; the exponent of the first modular exponentiation operation is an integer randomly selected by the cloud service provider; the modulus of the first modular exponentiation operation is a cyclic group The modulus of the second modular exponentiation operation; the second parameter is the result of the second modular exponentiation operation; the base of the second modular exponentiation operation is the product of the integer key held by the cloud service provider and the first parameter; the exponent of the second modular exponentiation operation is 1; the modulus of the second modular exponentiation operation is the modulus of the cyclic group minus 1; a group element randomly selected from the cyclic group and the integer key satisfy the third modular exponentiation operation; the result of the third modular exponentiation operation is the group element; the base of the third modular exponentiation operation is the generator of the cyclic group; the exponent of the third modular exponentiation operation is the integer key; the modulus of the third modular exponentiation operation is the modulus of the cyclic group; based on the first parameter and the second parameter, determine whether the cloud service provider has passed the security authentication.

[0026] In another possible implementation, the data security analysis module is specifically used to calculate the verification value according to a fourth modular exponentiation operation based on the first parameter and the second parameter; wherein the base of the fourth modular exponentiation operation is the third parameter; the exponent of the fourth modular exponentiation operation is 1, and the modulus of the fourth modular exponentiation operation is the modulus of the cyclic group; the third parameter is the product of the fourth parameter and the fifth parameter; the base of the fourth parameter is a group element randomly selected from the cyclic group; the exponent of the fourth parameter is the first parameter; the base of the fifth parameter is the first parameter; the exponent of the fifth parameter is the second parameter; when the verification value is equal to the first parameter, it is determined that the cloud service provider has passed the security authentication; when the verification value is not equal to the first parameter, it is determined that the cloud service provider has not passed the security authentication.

[0027] In another possible implementation, the task scheduling module is specifically used to perform weighted averaging based on each intermediate model parameter and the corresponding weight value to obtain the target model parameter; based on the target model parameter, determine the target model.

[0028] Another possible implementation method is that the task scheduling module is specifically used to substitute the target model parameters into the initial model to obtain a reference model; when the reference model meets the training stop condition, the reference model is determined as the target model.

[0029] In another possible implementation, the training stop condition includes at least one of the following: the analysis task deadline indicated by the business demand information is reached at the current moment; the number of times the target model is substituted into the initial model to obtain the reference model reaches a number threshold; the output error of the reference model is less than or equal to the error threshold.

[0030] Another possible implementation method is that the result processing module is specifically used to analyze the business data to be analyzed through the target model to determine the initial analysis results; perform data processing on the initial analysis results to determine the data analysis results; data processing includes at least one of the following: data correction, result desensitization, report generation according to preset templates and visual data conversion.

[0031] In a third aspect, the present application provides an electronic device, comprising: a processor and a memory; the memory stores instructions executable by the processor; when the processor is configured to execute the instructions, the electronic device implements the method of the first aspect above.

[0032] In a fourth aspect, the present application provides a computer-readable storage medium, which includes: computer software instructions; when the computer software instructions are executed in an electronic device, the electronic device implements the method of the first aspect above.

[0033] In a fifth aspect, the present application provides a computer program product, which includes a computer program; when the computer program runs in an electronic device, the electronic device implements the method of the first aspect.

[0034] The beneficial effects of the second to fifth aspects mentioned above refer to the corresponding description of the first aspect and will not be repeated here. BRIEF DESCRIPTION OF THE DRAWINGS

[0035] Figure 1 A schematic diagram of an application environment of a data analysis method provided in this application;

[0036] Figure 2 A flowchart of a data analysis method provided in this application;

[0037] Figure 3 A schematic diagram of another data analysis method provided for this application;

[0038] Figure 4 A flowchart of a security authentication method provided for this application;

[0039] Figure 5 A schematic diagram of the composition of a data analysis device provided in this application;

[0040] Figure 6 A schematic diagram of the structure of an electronic device provided in this application. DETAILED DESCRIPTION

[0041] A data analysis method provided by the present application will be described in detail below with reference to the accompanying drawings.

[0042] The term "and / or" in this article is merely a description of the association relationship of associated objects, indicating that three relationships may exist. For example, A and / or B can mean: A exists alone, A and B exist at the same time, and B exists alone.

[0043] The terms "first" and "second" and the like in the specification and drawings of this application are used to distinguish different objects, or to distinguish different processing of the same object, rather than to describe a specific order of objects.

[0044] In addition, the terms "including" and "having" and any variations thereof mentioned in the description of the present application are intended to cover non-exclusive inclusions. For example, a process, method, system, product or device that includes a series of steps or units is not limited to the listed steps or units, but may optionally include other steps or units that are not listed, or may optionally include other steps or units that are inherent to these processes, methods, products or devices.

[0045] It should be noted that, in the embodiments of the present application, words such as "exemplary" or "for example" are used to indicate examples, illustrations or descriptions. Any embodiment or design described as "exemplary" or "for example" in the embodiments of the present application should not be interpreted as being more preferred or more advantageous than other embodiments or designs. Specifically, the use of words such as "exemplary" or "for example" is intended to present related concepts in a specific way.

[0046] In order to facilitate the clear description of the technical solutions of the embodiments of the present application, in the embodiments of the present application, the words "first", "second", etc. are used to distinguish the same items or similar items with basically the same functions and effects. Those skilled in the art can understand that the words "first", "second", etc. are not limiting the quantity and execution order.

[0047] In the description of the present application, unless otherwise specified, “plurality” means two or more.

[0048] With the continuous advancement of technology and the increase in data volume, data analysis methods in multi-cloud scenarios play an important role in more fields. The main application scenarios include data integration and management, real-time data analysis, business intelligence reporting, and security monitoring.

[0049] In the intelligent data collection and analysis system and method for a multi-cloud environment, the process of deploying pre-developed test applications to various target cloud platforms requires obtaining multiple parameters of these cloud platforms, including data storage capacity, computing capacity, security capacity, and processing capacity, and centrally analyzing the obtained parameters to automatically provide decision-making recommendations for the enterprise.

[0050] In a data security analysis method and system, firstly, a first feature vector of the data to be analyzed is extracted from the log of the data security management system, and then network data is obtained from the historical data set and a second feature vector is extracted. Then, the two feature vectors are clustered using a clustering algorithm to determine the degree of association between them, and finally a data security analysis result for the data to be analyzed is obtained based on the degree of association.

[0051] However, deploying test applications on multiple cloud platforms and collecting performance parameters involves complex configuration and management, which increases maintenance costs and operational difficulty. There is also the problem of inconsistent data across platforms. Determining data security analysis results based on the degree of correlation of feature vectors does not consider the security of data transmission in multi-cloud scenarios.

[0052] In response to the above technical problems, the present application provides a data analysis method, which is applied to a multi-cloud manager. The idea is to receive business demand information from cloud service consumers, quickly determine the target cloud service provider, and improve the business response speed. In response to the business demand information of cloud service consumers, send an initial model to at least one target cloud service provider; the storage data in the target cloud service provider and the business data to be analyzed meet the data category matching conditions, set the data category matching conditions, and flexibly select the cloud consumption provider according to the business demand information, thereby improving the flexibility of the data analysis method. Receive the intermediate model parameters sent by at least one target cloud service provider, aggregate the intermediate model parameters, and obtain the target model. The deviations between multiple intermediate model parameters can be averaged to improve the accuracy of data analysis. The business data to be analyzed is analyzed by the target model, and the data analysis results are determined to help cloud service consumers make more informed decisions.

[0053] The embodiments provided in this application are described in detail below in conjunction with the accompanying drawings.

[0054] Figure 1 Schematic diagram of the application environment involved in the data analysis method provided in this application. Figure 1 As shown, the application environment includes: a cloud service consumer, a cloud service provider, and a multi-cloud manager.

[0055] Among them, cloud service consumers and cloud service providers are connected through a multi-cloud manager.

[0056] In some embodiments, the cloud service consumer is used to send a request for business requirement information to the multi-cloud manager and receive data analysis results.

[0057] Exemplarily, the cloud service consumer may be one or more, connected to the multi-cloud manager. The cloud service consumer may use a terminal device with wireless transceiver function to issue a service request, such as a mobile phone, tablet computer, wearable device, vehicle-mounted device, augmented reality (AR) / virtual reality (VR) device, laptop computer, ultra-mobile personal computer (UMPC), netbook, personal digital assistant (PDA), etc.

[0058] In some embodiments, the multi-cloud manager has a data analysis device, including a data security analysis module, a task scheduling module and a result processing module.

[0059] In some embodiments, the data security analysis module is used to receive business demand information of cloud service consumers.

[0060] Exemplarily, the data security analysis module is further used to send the initial model to at least one target cloud service provider.

[0061] Exemplarily, the data security analysis module may also be used to perform security authentication on the target cloud service provider.

[0062] In some embodiments, according to specific implementation scenarios and business requirements, a variety of different initial models are configured in the multi-cloud manager, which is not limited here.

[0063] For example, assuming that there is a feature matching model in the multi-cloud manager, by building a correlation model of historical data, the feature matching technology is used to evaluate the similarity between the business data set to be analyzed (A) and the recognition result data set (B) that has been formed, mainly by calculating the Jaccard distance. The smaller the Jaccard distance, the higher the feature matching degree between the original data and the recognition result data. It can be considered that the two data sets are very similar in feature space. The Jaccard distance satisfies the following formula (1):

[0064]

[0065] Among them, the recognition result dataset that has been formed needs to be continuously updated and trained to ensure the flexibility and accuracy of the multi-cloud manager.

[0066] For example, assuming that there is a machine learning model in the multi-cloud manager, first define an objective function (Obj) that satisfies the following formula (2):

[0067]

[0068] in, is the loss function, measuring the predicted value With the actual label y i The error between k ) is a regularization term used to control the complexity of the machine learning model and avoid overfitting. The training data set of the model (C) C = {(x1, y1), (x2, y2), …, (x w ,y w )},x i is the sample feature, y i For the corresponding label.

[0069] For example, for the i-th sample, the corresponding predicted value is The following formula (3) is satisfied:

[0070]

[0071] Among them, f k (x i ) represents the model's response to the input feature x i The predicted value of . K is used to represent the number of trees in the model. Aggregate the predictions from all trees to form the final prediction value

[0072] In some embodiments, combined with the predicted value With the actual label y i Construct a loss function. The choice of the specific loss function depends on the specific task and will not be elaborated here.

[0073] In some embodiments, the task scheduling module is used to receive at least one intermediate model parameter.

[0074] Exemplarily, the task scheduling module is also used to aggregate the intermediate model parameters to obtain the target model parameters.

[0075] Exemplarily, the result processing module is used to analyze the business data to be analyzed through the target model to determine the data analysis results.

[0076] Exemplarily, the multi-cloud manager may be a server, such as a server cluster consisting of multiple servers, or a single server, or a computer, or a processor or processing chip in a server or computer, etc.

[0077] In some embodiments, a cloud service provider is used to receive the initial model.

[0078] Exemplarily, the cloud service provider is also used to pre-train the initial model based on the stored data to obtain intermediate model parameters.

[0079] Exemplarily, the cloud service provider is further configured to send the intermediate model parameters to the multi-cloud manager.

[0080] Exemplarily, the cloud service provider may be one or more, connected to the multi-cloud manager, and the cloud service provider may be a server, such as a server cluster consisting of multiple servers, or a single server, or a computer, or a processor or processing chip in a server or computer, etc.

[0081] It should be noted that the system architecture described in the embodiments of the present application is for more clearly illustrating the technical solutions of the embodiments of the present application, and does not constitute a limitation on the technical solutions provided in the embodiments of the present application. A person of ordinary skill in the art can know that with the evolution of the system architecture, the technical solutions provided in the embodiments of the present application are also applicable to similar technical problems.

[0082] See also Figure 2 , is a flow chart of a data analysis method provided in an embodiment of the present application. Figure 2 As shown, the data analysis method provided by the present application specifically includes the following steps S201 to S203.

[0083] S201. In response to received business requirement information, send an initial model to at least one target cloud service provider.

[0084] The business demand information includes the business data to be analyzed, which is proposed by the cloud service consumer. The storage data in the target cloud service provider and the business data to be analyzed meet the data category matching condition.

[0085] For example, as shown in Table 1, the business demand information may include at least one of the following: mission start time (Mission / Task Start Time, mol_sta), business data type to be analyzed (Data Type for Analysis, dat_lat), memory required for the business to be analyzed (Data Capacity, dat_cap), necessary end condition required for the task (Mission / Task End Necessity, mol_end_ness), and task deadline (Mission / Task End Time, mol_endt), and business data to be analyzed. Among them, the business data to be analyzed of the cloud service consumer depends on the specific business scenario and is not limited here.

[0086] Table 1 Business requirement information

[0087] Task start time mol_sta Type of business data to be analyzed dat_lat Memory required for the business to be analyzed dat_cap Necessary conditions for the task to end mol_end_ness Task deadline mol_endt Business data to be analyzed

[0088] The storage data in the target cloud service provider and the business data to be analyzed meet the data category matching condition.

[0089] In some embodiments, the initial model is internally configured in the multi-cloud manager. The specific initial model type is determined according to actual conditions and is not limited here.

[0090] Exemplarily, the multi-cloud manager receives business demand information from cloud service consumers, determines at least one target cloud service provider based on data category matching conditions and a data storage record table, and sends an initial model to the target cloud service provider; wherein the data storage record table includes a data storage type record table and a business type record table.

[0091] It can be understood that sending the initial model to the target cloud service provider that meets the data category matching conditions improves the accuracy of data analysis, while the matching of data types speeds up the model training time and the speed of data analysis.

[0092] S202: Receive intermediate model parameters sent by at least one target cloud service provider, and aggregate the intermediate model parameters to obtain a target model.

[0093] In some embodiments, the target cloud service provider is a cloud service provider that has passed security authentication.

[0094] Among them, the intermediate model parameters are obtained by pre-training the initial model based on the storage data by the corresponding target cloud service provider.

[0095] S203: Analyze the business data to be analyzed by using the target model to determine the data analysis results.

[0096] It is understandable that the data analysis method provided by this application can quickly select the target cloud service provider according to the data category matching conditions after receiving the business needs of the cloud service consumer, thereby accelerating the speed of data analysis and business response time. At the same time, the diversity of cloud service provider selection enhances the flexibility and versatility of the data analysis method. Aggregating multiple intermediate model parameters avoids the problem of inconsistent data between different cloud service providers and increases the accuracy of the data analysis method.

[0097] In some embodiments, the data category matching condition in S201 above may include the following two conditions:

[0098] Condition 1: The data type of the data stored in the cloud service provider is the same as the data type of the business data to be analyzed.

[0099] In some embodiments, the multi-cloud manager can determine, based on the data type of the business data to be analyzed and the first correspondence, a cloud service provider whose data type of the stored data in the first correspondence is the same as the data type of the business data to be analyzed as the target cloud service provider; the first correspondence is used to represent the correspondence between the cloud service provider and the data type of the stored data.

[0100] Exemplarily, before the cloud service consumer and the cloud service provider access the multi-cloud manager, a record table corresponding to the data storage type has been formed and configured in the multi-cloud manager. The multi-cloud manager determines the target cloud service provider based on the record table corresponding to the data storage type, but the specific record table corresponding to the data storage type can be changed according to the specific situation and is not limited here.

[0101] For example, as shown in Table 2, it is assumed that the data types of the services to be analyzed are structured data (such as database records) and unstructured data (such as images, audio, video, etc.). For example, for structured data, a cloud service provider that provides efficient database query, analysis, and reporting functions is selected; for unstructured data, a cloud service provider with image recognition, video analysis, natural language processing, and other functions is selected.

[0102] Table 2 Data storage type record table

[0103]

[0104] Condition 2: The business type of the data stored in the cloud service provider is the same as the business type of the business data to be analyzed.

[0105] In some embodiments, according to the business type of the business data to be analyzed and the second correspondence, a cloud service provider whose business type of the data stored in the second correspondence is the same as the business type of the business data to be analyzed is determined as the target cloud service provider; the second correspondence is used to represent the correspondence between the cloud service provider and the business type of the stored data. The second correspondence is set according to the specific scenario requirements and is not limited here.

[0106] Exemplarily, as shown in Table 3, the correspondence between business types and cloud service providers, assuming that the business types to be analyzed include enterprise resource planning, customer relationship management, data storage and backup, big data and analysis, and software development and testing, each business type has different requirements for the cloud service provider.

[0107] Table 3 Business type record table

[0108]

[0109]

[0110] In some embodiments, the step S202 of “aggregating the intermediate model parameters to obtain the target model” can be specifically implemented as the following steps S2021 to S2022:

[0111] S2021. Perform weighted averaging based on each intermediate model parameter and the corresponding weight value to obtain the target model parameter.

[0112] In some embodiments, the weight value corresponding to each intermediate model parameter can be configured according to specific business requirements and internal resources of the cloud service provider, which is not limited here.

[0113] Exemplarily, the internal resources of a cloud service provider include data quality and reliability, computing power, network latency and bandwidth resources, geographic location and coverage, and historical performance and stability.

[0114] For example, according to the business requirements, cloud service providers with high matching degree are selected and given higher weights. For example, for businesses with high real-time requirements, cloud service providers that provide low-latency services are given higher weights.

[0115] In some embodiments, the target model parameters (ω global ) satisfies the following formula (4):

[0116]

[0117] Among them, ω i Used to represent the weight value corresponding to the i-th intermediate model parameter, It is used to represent the i-th intermediate model parameter, and n is used to represent the number of intermediate models.

[0118] It should be noted that there is no unique way to aggregate the intermediate model parameters, which can be determined by the actual situation and is not limited here.

[0119] S2022. Determine the target model based on the target model parameters.

[0120] It can be understood that by determining the target model parameters by weighted averaging, the weights can be freely configured according to business needs, which increases the flexibility and accuracy of the data analysis method. At the same time, multiple intermediate model parameters are aggregated to improve the response speed and reliability of the data analysis method.

[0121] In some embodiments, the above step S2022 can be specifically implemented as steps a1 to a2:

[0122] a1. Substitute the target model parameters into the initial model to obtain the reference model.

[0123] Exemplarily, the initial model type is determined according to business requirements, which is not limited here.

[0124] a2. When the reference model meets the training stop condition, the reference model is determined as the target model.

[0125] Exemplarily, the training stop condition includes at least one of the following:

[0126] Condition 1: The current moment reaches the analysis task deadline indicated by the business requirement information.

[0127] In some embodiments, the deadline for the analysis task is the latest time for the completion of the data analysis, which is determined based on specific business needs and will not be elaborated here.

[0128] Condition 2: The number of times the target model is substituted into the initial model to obtain the reference model reaches the number threshold.

[0129] For example, the number threshold is determined based on the specific business and the initial model, and is not limited here. For example, in the early stage of training, as the number of initial model iterations increases, the performance of the initial model tends to improve significantly, but when the performance of the initial model reaches a certain level, continued iteration may lead to overfitting.

[0130] Condition 3: The output error of the reference model is less than or equal to the error threshold.

[0131] In some embodiments, the error threshold is a preset critical value used to evaluate the magnitude of the reference model output error. For a reference model whose output error is less than or equal to the error threshold, it generally meets the business requirements to be analyzed.

[0132] Exemplarily, the error threshold is determined based on specific business requirements and is not limited here.

[0133] In some embodiments, the above step S203 may be specifically implemented as steps S2031 to S2032:

[0134] S2031. Analyze the business data to be analyzed through the target model to determine the initial analysis results.

[0135] S2032. Perform data processing on the initial analysis results to determine the data analysis results; the data processing includes at least one of the following: data correction, result desensitization, report generation according to a preset template, and visual data conversion.

[0136] Exemplarily, data correction is used to correct errors or abnormal data in the initial analysis results, such as correcting input errors, processing missing values, identifying and correcting duplicate data, etc.

[0137] Exemplarily, result desensitization refers to hiding or replacing sensitive information in the initial analysis results to protect personal privacy or business secrets.

[0138] For example, generating a report according to a preset template According to the needs and purposes of data analysis, a preset report template can be designed, including tables, charts, texts, etc. The preset template is used to format the analysis results and generate a report with a clear structure and easy to understand.

[0139] Exemplarily, visual data conversion refers to converting data analysis results into visual forms such as graphics and images to more intuitively display data features and trends.

[0140] In some embodiments, before the step of "sending the initial model to at least one target cloud service provider", the data analysis method provided by the present application further includes S204-S205:

[0141] S204: Obtain the analysis task duration of the business data to be analyzed.

[0142] In some embodiments, the business demand information also includes the analysis task start time and analysis task deadline of the business data to be analyzed. The above step S204 is specifically implemented as follows: determining the analysis task duration according to the analysis task start time and analysis task deadline of the business data to be analyzed.

[0143] Exemplarily, the analysis task start time is the time when the multi-cloud manager receives the service to be analyzed.

[0144] The duration of the analysis task is the time period from the start time of the analysis task to the end time of the analysis task.

[0145] S205: When the duration of the analysis task is greater than the duration threshold, periodically perform security authentication on the target cloud service provider at a preset interval.

[0146] Exemplarily, the duration threshold is determined according to specific business requirements and is not limited here. For the case where the analysis task market is less than or equal to the duration threshold, it is only necessary to perform security authentication on the target cloud service provider before the multi-cloud manager first sends the initial model to the target cloud service provider.

[0147] For example, the preset interval can be determined according to specific business requirements, and is not limited here. For business requirements with strong confidentiality, the preset interval can be appropriately reduced.

[0148] As an example, the multi-cloud manager may perform security authentication only for the selected target cloud service provider, or the multi-cloud manager may perform security authentication for all cloud service providers. This embodiment of the present application is not limited to this. The specific process of the multi-cloud manager performing security authentication can refer to the introduction at S301-S303 in the following embodiment, which will not be repeated here.

[0149] It is understandable that security authentication of the target cloud service provider can avoid data leakage when data is transmitted between multiple cloud service providers, improve the security protection of data, and enhance the user experience of cloud service consumers.

[0150] In some embodiments, before aggregating the intermediate model parameters to obtain the target model, the multi-cloud manager may also perform security authentication on the cloud service provider in the following manner:

[0151] S301. Randomly select a group element from the cyclic group and send it to the cloud service provider.

[0152] Among them, the modulus of the cyclic group is a prime number.

[0153] Exemplarily, the multi-cloud manager selects a cyclic group G over a finite field, whose generator is g, and a prime number p to define the group operations. For example, in the multiplication group of integers modulo p In this case, p is a prime number and g is a multiplicative group. A generator of .

[0154] Exemplarily, the multi-cloud manager randomly selects a group element (y) from the cyclic group and sends it to the cloud service provider.

[0155] S302: Receive authentication response information from the cloud service provider.

[0156] The authentication response information includes a first parameter and a second parameter.

[0157] Among them, the first parameter is the result of the first modular exponentiation operation; the base of the first modular exponentiation operation is the generator of the cyclic group; the exponent of the first modular exponentiation operation is an integer randomly selected by the cloud service provider; and the modulus of the first modular exponentiation operation is the modulus of the cyclic group.

[0158] Exemplarily, the first parameter r satisfies the following calculation formula (5):

[0159] r = g k modp formula (5)

[0160] Here, k is an integer randomly selected by the cloud service provider.

[0161] In some embodiments, the second parameter is the result of a second modular exponentiation operation; the base of the second modular exponentiation operation is the product of an integer key held by the cloud service provider and the first parameter; the exponent of the second modular exponentiation operation is 1; and the modulus of the second modular exponentiation operation is the modulus of the cyclic group minus 1.

[0162] Exemplarily, the second parameter s satisfies the following calculation formula (6):

[0163] s=k=xr mod(p-1) Formula (6)

[0164] Among them, x is the integer key held by the cloud service provider, and the cloud service provider needs to prove the possession of the integer key to the multi-cloud manager.

[0165] In some embodiments, a group element randomly selected from a cyclic group and an integer key satisfy a third modular exponentiation operation; a result of the third modular exponentiation operation is a group element; a base of the third modular exponentiation operation is a generator of the cyclic group; an exponent of the third modular exponentiation operation is the integer key; and a modulus of the third modular exponentiation operation is a modulus of the cyclic group.

[0166] Exemplarily, the third modular exponentiation operation satisfies the following formula (7):

[0167] y=g x mod p Formula (7)

[0168] Here, y is a randomly selected group element in the cyclic group, and x is an integer key held by the cloud service provider.

[0169] S303: Determine whether the cloud service provider has passed security authentication based on the first parameter and the second parameter.

[0170] In some embodiments, the above step S303 is specifically implemented as the following steps b1 to b3:

[0171] b1. Based on the first parameter and the second parameter, calculate the verification value according to the fourth modular exponentiation operation.

[0172] In some embodiments, the base of the fourth modular exponentiation operation is the third parameter; the exponent of the fourth modular exponentiation operation is 1, the modulus of the fourth modular exponentiation operation is the modulus of the cyclic group; and the third parameter is the product of the fourth parameter and the fifth parameter.

[0173] The base of the fourth parameter is a group element randomly selected from the cyclic group; and the exponent of the fourth parameter is the first parameter.

[0174] The base of the fifth parameter is the first parameter; the exponent of the fifth parameter is the second parameter.

[0175] For example, the fourth parameter is y r , the fifth parameter is r s , the third parameter is yr ·r s , the fourth modular exponentiation operation satisfies the following formula (8):

[0176] Verification value = y r ·r s mod p Formula (8)

[0177] Exemplarily, the verification value is used to verify that the cloud service provider can participate in the data analysis.

[0178] b2. When the verification value is equal to the first parameter, determine that the cloud service provider passes the security authentication.

[0179] Exemplarily, when the verification value is equal to r, it proves that the cloud service provider owns the integer key, and the cloud service provider has passed the security authentication and is qualified to participate in data analysis.

[0180] b3. When the verification value is not equal to the first parameter, it is determined that the cloud service provider has not passed the security authentication.

[0181] Exemplarily, when the verification value is not equal to r, it proves that the cloud service provider does not possess the integer key and cannot participate in the data analysis process.

[0182] The data analysis method of the present application embodiment is introduced below with a specific embodiment. The specific implementation process of the method is as follows: Figure 3 shown.

[0183] c1. After receiving the business requirement information, the initial model is sent to at least one target cloud service provider.

[0184] In some embodiments, the initial model type is determined according to specific business requirement information, and the initial model is sent to a target cloud service provider that has passed security authentication.

[0185] Exemplarily, the target cloud service provider is determined based on a matching condition between the stored data and the data category.

[0186] For example, Figure 4 The multi-cloud manager sends a random element as a challenge, the cloud service provider receives the random element, calculates the first element (that is, the first parameter mentioned above) and the second element (that is, the second parameter mentioned above) according to the random element, and the cloud service provider sends the first element and the second element to the multi-cloud manager. The multi-cloud manager verifies whether the first element and the second element satisfy the fourth modular exponentiation operation. The cloud service provider that satisfies the fourth modular exponentiation operation passes the security authentication.

[0187] c2. Receive intermediate model parameters sent by at least one target cloud service provider.

[0188] In some embodiments, the target cloud service provider pre-trains the initial model based on the stored data to obtain intermediate model parameters.

[0189] c3. Aggregate the intermediate model parameters to obtain the reference model.

[0190] Exemplarily, at least one intermediate model parameter is aggregated by weighted averaging to obtain a reference model.

[0191] c4. Whether the reference model meets the training stop condition.

[0192] Exemplarily, if the reference model meets the training stop condition, the reference model is the target model, and the process goes to step c5; if the reference model does not meet the training stop condition, the process goes to step c1.

[0193] c5. Analyze the business data to be analyzed through the target model and determine the data analysis results.

[0194] Exemplarily, the business data to be analyzed is analyzed through the target model to determine the initial analysis results; the initial analysis results are processed to determine the data analysis results; the data processing includes at least one of the following: data correction, result desensitization, report generation according to a preset template, and visual data conversion.

[0195] It can be seen that the above mainly introduces the solution provided by the embodiment of the present application from the perspective of the method. In order to achieve the above functions, the embodiment of the present application provides a hardware structure and / or software module corresponding to each function. It should be easily appreciated by those skilled in the art that, in combination with the modules and algorithm steps of each example described in the embodiment disclosed herein, the embodiment of the present application can be implemented in the form of hardware or a combination of hardware and computer software. Whether a function is executed in the form of hardware or computer software driving hardware depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to exceed the scope of the present invention.

[0196] The embodiment of the present application can divide the functional modules of the data analysis device according to the above method example. For example, each functional module can be divided according to each function, or two or more functions can be integrated into one processing module. The above integrated module can be implemented in the form of hardware or in the form of software functional modules. Optionally, the division of modules in the embodiment of the present application is schematic and is only a logical functional division. There may be other division methods in actual implementation.

[0197] In some embodiments, the present application further provides a data analysis device, which may include one or more functional modules for implementing the data analysis method of the above method embodiment.

[0198] For example, Figure 5 A schematic diagram of the composition of a data analysis device provided in an embodiment of the present application. Figure 5 As shown, the data analysis device 800 includes: a data security analysis module 801, a task scheduling module 802 and a result processing module 803. The data security analysis module 801 is used to receive business demand information from cloud service consumers; the business demand information includes business data to be analyzed; in response to the business demand information of the cloud service consumer, an initial model is sent to at least one target cloud service provider; the storage data in the target cloud service provider and the business data to be analyzed meet the data category matching condition; the task scheduling module 802 is used to receive the intermediate model parameters sent by each corresponding one of the at least one target cloud service provider; the intermediate model parameters are aggregated to obtain the target model; the intermediate model parameters are obtained by the corresponding target cloud service provider through pre-training the initial model based on the storage data; the result processing module 803 is used to analyze the business data to be analyzed through the target model to determine the data analysis result.

[0199] In some embodiments, the data category matching condition includes: the data type of the stored data in the cloud service provider is the same as the data type of the business data to be analyzed.

[0200] In other embodiments, the data category matching condition includes: the business type of the data stored in the cloud service provider is the same as the business type of the business data to be analyzed.

[0201] In some further embodiments, the target cloud service provider is a cloud service provider that has passed security authentication.

[0202] In some other embodiments, the data security analysis module 801 is also used to obtain the analysis task duration of the business data to be analyzed; when the analysis task duration is greater than a duration threshold, security authentication is performed periodically on the target cloud service provider at a preset interval.

[0203] In some other embodiments, the business demand information also includes the analysis task start time and analysis task deadline of the business data to be analyzed; the data security analysis module 801 is specifically used to determine the analysis task duration based on the analysis task start time and analysis task deadline of the business data to be analyzed.

[0204] In some other embodiments, the data security analysis module 801 is specifically used to perform security authentication on the cloud service provider in the following manner: sending an authentication request message to the cloud service provider, the authentication request message including a group element randomly selected from a cyclic group; the modulus of the cyclic group is a prime number; receiving authentication response information from the cloud service provider; the authentication response information includes a first parameter and a second parameter; wherein the first parameter is the result of a first modular exponentiation operation; the base of the first modular exponentiation operation is a generator of the cyclic group; the exponent of the first modular exponentiation operation is an integer randomly selected by the cloud service provider; the modulus of the first modular exponentiation operation is a cyclic group The modulus of the second modular exponentiation operation; the second parameter is the result of the second modular exponentiation operation; the base of the second modular exponentiation operation is the product of the integer key held by the cloud service provider and the first parameter; the exponent of the second modular exponentiation operation is 1; the modulus of the second modular exponentiation operation is the modulus of the cyclic group minus 1; a group element randomly selected from the cyclic group and the integer key satisfy the third modular exponentiation operation; the result of the third modular exponentiation operation is the group element; the base of the third modular exponentiation operation is the generator of the cyclic group; the exponent of the third modular exponentiation operation is the integer key; the modulus of the third modular exponentiation operation is the modulus of the cyclic group; based on the first parameter and the second parameter, determine whether the cloud service provider has passed the security authentication.

[0205] In some other embodiments, the data security analysis module 801 is specifically used to calculate the verification value according to the fourth modular exponentiation operation based on the first parameter and the second parameter; wherein the base of the fourth modular exponentiation operation is the third parameter; the exponent of the fourth modular exponentiation operation is 1, and the modulus of the fourth modular exponentiation operation is the modulus of the cyclic group; the third parameter is the product of the fourth parameter and the fifth parameter; the base of the fourth parameter is a group element randomly selected from the cyclic group; the exponent of the fourth parameter is the first parameter; the base of the fifth parameter is the first parameter; the exponent of the fifth parameter is the second parameter; when the verification value is equal to the first parameter, it is determined that the cloud service provider has passed the security authentication; when the verification value is not equal to the first parameter, it is determined that the cloud service provider has not passed the security authentication.

[0206] In some other embodiments, the task scheduling module 802 is specifically used to perform weighted averaging based on each intermediate model parameter and the corresponding weight value to obtain the target model parameter; and determine the target model based on the target model parameter.

[0207] In some other embodiments, the task scheduling module 802 is specifically used to substitute the target model parameters into the initial model to obtain a reference model; when the reference model meets the training stop condition, the reference model is determined as the target model.

[0208] In some further embodiments, the training stop condition includes at least one of the following: the analysis task deadline indicated by the business demand information is reached at the current moment; the number of times the target model is substituted into the initial model to obtain the reference model reaches a number threshold; the output error of the reference model is less than or equal to an error threshold.

[0209] In some other embodiments, the result processing module 803 is specifically used to analyze the business data to be analyzed through the target model to determine the initial analysis results; perform data processing on the initial analysis results to determine the data analysis results; the data processing includes at least one of the following: data correction, result desensitization, generating reports according to preset templates and visual data conversion.

[0210] In the case of implementing the functions of the above-mentioned integrated modules in the form of hardware, the embodiment of the present invention provides a possible structural diagram of the electronic device involved in the above-mentioned embodiment. Figure 6 As shown, the electronic device 900 includes: a processor 902 , a communication interface 903 , and a bus 904 . Optionally, the electronic device 900 may further include a memory 901 .

[0211] The processor 902 may be a processor that implements or executes various exemplary logic blocks, modules, and circuits described in conjunction with the disclosure of the present application. The processor 902 may be a central processing unit, a general-purpose processor, a digital signal processor, an application-specific integrated circuit, a field programmable gate array, or other programmable logic devices, transistor logic devices, hardware components, or any combination thereof. It may implement or execute various exemplary logic blocks, modules, and circuits described in conjunction with the disclosure of the present application. The processor 902 may also be a combination that implements computing functions, such as a combination of one or more microprocessors, a combination of a DSP and a microprocessor, and the like.

[0212] The communication interface 903 is used to connect with other devices via a communication network, such as Ethernet, wireless access network, wireless local area network (WLAN), etc.

[0213] The memory 901 may be a read-only memory (ROM) or other types of static storage devices that can store static information and instructions, a random access memory (RAM) or other types of dynamic storage devices that can store information and instructions, or an electrically erasable programmable read-only memory (EEPROM), a disk storage medium or other magnetic storage device, or any other medium that can be used to carry or store desired program code in the form of instructions or data structures and can be accessed by a computer, but is not limited thereto.

[0214] As a possible implementation, the memory 901 may exist independently of the processor 902, and the memory 901 may be connected to the processor 902 via a bus 904 for storing instructions or program codes. When the processor 902 calls and executes the instructions or program codes stored in the memory 901, the data analysis method provided in the embodiment of the present invention can be implemented.

[0215] In another possible implementation, the memory 901 may also be integrated with the processor 902 .

[0216] The bus 904 may be an extended industry standard architecture (EISA) bus, etc. The bus 904 may be divided into an address bus, a data bus, a control bus, etc. For ease of representation, Figure 6 Only one thick line is used in the diagram, but this does not mean that there is only one bus or only one type of bus.

[0217] Through the description of the above implementation methods, technical personnel in the relevant field can clearly understand that for the convenience and simplicity of description, only the division of the above-mentioned functional modules is used as an example. In actual applications, the above-mentioned functions can be assigned to different functional modules as needed, that is, the internal structure of the service calling device can be divided into different functional modules to complete all or part of the functions described above.

[0218] The embodiment of the present application also provides a computer-readable storage medium. All or part of the processes in the above method embodiments can be completed by computer instructions to instruct the relevant hardware, and the program can be stored in the above computer-readable storage medium. When the program is executed, it may include the processes of the above method embodiments. The computer-readable storage medium can be the memory or memory of any of the above embodiments. The above computer-readable storage medium can also be an external storage device of the above service calling device, such as a plug-in hard disk, a smart memory card (smart media card, SMC), a secure digital (secure digital, SD) card, a flash card (flash card), etc. equipped on the above service calling device. Further, the above computer-readable storage medium can also include both the internal storage unit of the above service calling device and an external storage device. The above computer-readable storage medium is used to store the above computer program and other programs and data required by the above service calling device. The above computer-readable storage medium can also be used to temporarily store data that has been output or is to be output.

[0219] An embodiment of the present application also provides a computer program product, which includes a computer program. When the computer program product is run on a computer, the computer is enabled to execute any one of the data analysis methods provided in the above embodiments.

[0220] The above are only specific implementations of the present application, but the protection scope of the present application is not limited thereto. Any changes or substitutions within the technical scope disclosed in the present application should be included in the protection scope of the present application. Therefore, the protection scope of the present application should be based on the protection scope of the claims.

Claims

1. A data analysis method, characterized in that: Applies to multi-cloud managers, including: Receiving business demand information from a cloud service consumer; the business demand information includes business data to be analyzed; In response to the business demand information of the cloud service consumer, an initial model is sent to at least one target cloud service provider; the stored data in the target cloud service provider and the business data to be analyzed meet a data category matching condition; Receiving intermediate model parameters sent by at least one of the target cloud service providers, and aggregating the intermediate model parameters to obtain a target model; the intermediate model parameters are obtained by pre-training the initial model based on the storage data by the corresponding target cloud service provider; The business data to be analyzed is analyzed through the target model to determine the data analysis result.

2. The method according to claim 1, characterized in that The data category matching condition includes: the data type of the stored data in the cloud service provider is the same as the data type of the business data to be analyzed.

3. The method according to claim 1, characterized in that The data category matching condition includes: the business type of the data stored in the cloud service provider is the same as the business type of the business data to be analyzed.

4. The method according to claim 1, characterized in that: The target cloud service provider is a cloud service provider that has passed security certification.

5. The method according to claim 4, characterized in that The method further comprises: Obtaining the analysis task duration of the business data to be analyzed; When the duration of the analysis task is greater than the duration threshold, security authentication is periodically performed on the target cloud service provider at a preset interval.

6. The method according to claim 5, characterized in that The business demand information also includes the analysis task start time and analysis task deadline of the business data to be analyzed; The obtaining of the analysis task duration of the service data to be analyzed includes: The duration of the analysis task is determined according to the analysis task start time and the analysis task deadline of the business data to be analyzed.

7. The method according to claim 4, characterized in that The method further comprises: Perform security authentication on cloud service providers as follows: Sending an authentication request message to the cloud service provider, wherein the authentication request message includes a group element randomly selected from a cyclic group; the modulus of the cyclic group is a prime number; Receiving authentication response information from a cloud service provider; the authentication response information includes a first parameter and a second parameter; Among them, the first parameter is the result of the first modular exponentiation operation; the base of the first modular exponentiation operation is the generator of the cyclic group; the exponent of the first modular exponentiation operation is an integer randomly selected by the cloud service provider; the modulus of the first modular exponentiation operation is the modulus of the cyclic group; the second parameter is the result of the second modular exponentiation operation; the base of the second modular exponentiation operation is the product of the integer key held by the cloud service provider and the first parameter; the exponent of the second modular exponentiation operation is 1; the modulus of the second modular exponentiation operation is the modulus of the cyclic group minus 1; a group element randomly selected from the cyclic group and the integer key satisfy a third modular exponentiation operation; the result of the third modular exponentiation operation is the group element; the base of the third modular exponentiation operation is the generator of the cyclic group; the exponent of the third modular exponentiation operation is the integer key; the modulus of the third modular exponentiation operation is the modulus of the cyclic group; Based on the first parameter and the second parameter, determine whether the cloud service provider passes the security authentication.

8. The method according to claim 7, characterized in that The determining, based on the first parameter and the second parameter, whether the cloud service provider passes the security authentication includes: Calculate a verification value according to a fourth modular exponentiation operation based on the first parameter and the second parameter; The base of the fourth modular exponentiation operation is the third parameter; the exponent of the fourth modular exponentiation operation is 1, and the modulus of the fourth modular exponentiation operation is the modulus of the cyclic group; the third parameter is the product of the fourth parameter and the fifth parameter; the base of the fourth parameter is a group element randomly selected from the cyclic group; the exponent of the fourth parameter is the first parameter; the base of the fifth parameter is the first parameter; and the exponent of the fifth parameter is the second parameter. When the verification value is equal to the first parameter, determining that the cloud service provider passes the security authentication; When the verification value is not equal to the first parameter, it is determined that the cloud service provider has failed the security authentication.

9. The method according to claim 1, characterized in that: The step of aggregating the intermediate model parameters to obtain the target model includes: Based on each intermediate model parameter and the corresponding weight value, a weighted average is calculated to obtain the target model parameter; Based on the target model parameters, the target model is determined.

10. The method according to claim 9, characterized in that The step of determining the target model based on the target model parameters comprises: Substituting the target model parameters into the initial model to obtain a reference model; When the reference model satisfies a training stop condition, the reference model is determined as the target model.

11. The method according to claim 10, characterized in that The training stop condition includes at least one of the following: The current moment reaches the analysis task deadline indicated by the business requirement information; The number of times the target model is substituted into the initial model to obtain a reference model reaches a number threshold; The output error of the reference model is less than or equal to an error threshold.

12. The method according to claim 1, characterized in that Analyzing the business data to be analyzed by using the target model to determine the data analysis result includes: Analyze the business data to be analyzed by using the target model to determine an initial analysis result; The initial analysis results are processed to determine the data analysis results; the data processing includes at least one of the following: data correction, result desensitization, report generation according to a preset template, and visual data conversion.

13. An electronic device, characterized in that: It comprises a processor and a memory, wherein the processor is coupled to the memory; the memory is used to store computer instructions, and the computer instructions are loaded and executed by the processor to enable a computer device to implement the data analysis method as described in any one of claims 1 to 12.

14. A computer-readable storage medium, characterized in that: The computer-readable storage medium includes computer-executable instructions, and when the computer-executable instructions are executed on a computer, the computer is enabled to execute the data analysis method according to any one of claims 1 to 12.

15. A computer program product, characterized in that The computer program product comprises a computer program, and when the computer program is run on an electronic device, the electronic device executes the data analysis method according to any one of claims 1 to 12.

Images