Interaction data processing method and device, equipment and storage medium

Abstract

The invention provides an interactive data processing method and device, equipment and a storage medium, which can be applied to various scenes such as cloud technology, artificial intelligence, map field, smart traffic, auxiliary driving and vehicle-mounted, and the method comprises the following steps: obtaining a preset amount of interactive data; the interaction data comprises interaction test data and interaction contrast data; screening out candidate interaction test data from the interaction test data to obtain target interaction test data, and screening out candidate interaction contrast data from the interaction contrast data to obtain target interaction contrast data; and performing performance evaluation processing on the target interaction test data and the target interaction contrast data to obtain performance evaluation data. The data obtained after screening is used for performance evaluation processing, the obtained performance evaluation data can have higher accuracy, then the service system conducts updating according to the performance evaluation data, the response efficiency of the service system can be improved, server computing resource consumption of the service system is reduced, and the server running speed is increased.

Description

Technical Field

[0001] The present application belongs to the field of computer technology, and specifically relates to an interactive data processing method, device, equipment and storage medium. Background Art

[0002] With the rapid development of Internet technology, various business systems are often continuously optimized and updated. When optimizing and updating some business systems, it is usually necessary to use some testing methods to test the optimization effects of new versions or new functions. For example, A / B testing is used to test the optimization effects of new versions or new functions. A / B testing is to formulate two plans for the same goal, let some terminal objects use plan A, and let the other terminal objects use plan B. By defining a series of business indicators based on the interactive data fed back by the terminal objects, the effects of plan A and plan B are evaluated.

[0003] In the related art, in order to improve the test efficiency, when conducting A / B experiments, multiple experimental layers are usually defined, and different experimental layers can reuse the business requests of the same terminal object, so that multiple tests can be performed at the same time. However, when the same parameters are configured in different experimental layers, if the parameter values ​​in different experimental layers are different, when the business requests of the same terminal object are reused, there are multiple alternative values ​​for the parameter, and each parameter can only take effect on one value for the business request. This will result in the value of the parameter taking effect is not necessarily the value configured by the experimental plan. In this case, the interaction data fed back by the terminal object is not accurate, and accurate test results cannot be obtained. The business system is updated according to the test results, which may result in low response efficiency and increase the server computing resource consumption of the business system. Summary of the invention

[0004] In order to solve the above technical problems, the present application provides an interactive data processing method, device, equipment and storage medium.

[0005] On the one hand, an embodiment of the present application provides an interactive data processing method, the method comprising:

[0006] Obtaining interaction data corresponding to a preset number of business requests; the interaction data is interaction data between a preset terminal object and a business processing result; the business processing result is obtained by processing a preset number of business requests based on a first business layer and a second business layer; the first business layer and the second business layer contain the same control parameters; the first business layer includes a first test business and a first comparison business; the second business layer includes a second test business and a second comparison business; the interaction data includes interaction test data and interaction comparison data; the interaction test data includes data corresponding to the first test business and the second test business, and data corresponding to the first test business and the second comparison business; the interaction comparison data includes data corresponding to the first comparison business and the second test business, and data corresponding to the first comparison business and the second comparison business;

[0007] Determine, from the interaction test data, interaction data corresponding to the service request processed by both the first test service and the second test service to obtain candidate interaction test data; and determine, from the interaction comparison data, interaction data corresponding to the service request processed by both the comparison service and the second test service to obtain candidate interaction comparison data;

[0008] Screening out candidate interaction test data in the interaction test data to obtain target interaction test data, and screening out candidate interaction control data in the interaction control data to obtain target interaction control data;

[0009] The target interaction test data and the target interaction control data are subjected to performance evaluation processing to obtain performance evaluation data.

[0010] On the other hand, an embodiment of the present application further provides an interactive data processing device, the device comprising:

[0011] An acquisition module is used to acquire interaction data corresponding to a preset number of business requests; the interaction data is interaction data between a preset terminal object and a business processing result; the business processing result is obtained by processing a preset number of business requests based on a first business layer and a second business layer; the first business layer and the second business layer contain the same control parameters; the first business layer includes a first test business and a first comparison business; the second business layer includes a second test business and a second comparison business; the interaction data includes interaction test data and interaction comparison data; the interaction test data includes data corresponding to the first test business and the second test business, and data corresponding to the first test business and the second comparison business; the interaction comparison data includes data corresponding to the first comparison business and the second test business, and data corresponding to the first comparison business and the second comparison business;

[0012] A candidate interaction data determination module is used to determine, from the interaction test data, interaction data corresponding to the service request processed by both the first test service and the second test service to obtain candidate interaction test data, and to determine, from the interaction comparison data, interaction data corresponding to the service request processed by both the comparison service and the second test service to obtain candidate interaction comparison data;

[0013] A screening module is used to screen out candidate interactive test data in the interactive test data to obtain target interactive test data, and to screen out candidate interactive control data in the interactive control data to obtain target interactive control data;

[0014] The performance evaluation data determination module is used to perform performance evaluation processing on the target interaction test data and the target interaction control data to obtain performance evaluation data.

[0015] On the other hand, an embodiment of the present application also provides an electronic device for interactive data processing, the electronic device comprising a processor and a memory, the memory storing at least one instruction or at least one program, and the at least one instruction or at least one program is loaded and executed by the processor to implement the interactive data processing method as described above.

[0016] On the other hand, an embodiment of the present application also provides a computer-readable storage medium, in which at least one instruction or at least one program is stored, and the at least one instruction or the at least one program is loaded and executed by a processor to implement the interactive data processing method as described above.

[0017] On the other hand, an embodiment of the present application further provides a computer program product, which, when executed by a processor, implements the interactive data processing method as described above.

[0018] The interactive data processing method, device, equipment and storage medium proposed in the embodiment of the present application obtain candidate interactive test data by determining the interactive data corresponding to the business request processed by the first test business and the second test business at the same time from the interactive test data, and obtain candidate interactive comparison data by determining the interactive data corresponding to the business request processed by the comparison business and the second test business at the same time from the interactive comparison data, so that the candidate interactive test data and the candidate interactive comparison data can be screened out to prevent the candidate interactive test data and the candidate interactive comparison data from interfering with the test results of the first business layer. Using the target interactive test data and the target interactive comparison data obtained after screening for performance evaluation processing can make the obtained performance evaluation data have higher accuracy, so that the optimization effect of the updated content can be more accurately evaluated, and then the business system is updated according to the performance evaluation data, which can improve the response efficiency of the business system. At the same time, the server computing resource consumption of the business system is reduced, and the server operation speed of the business system is improved. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] In order to more clearly illustrate the technical solutions and advantages in the embodiments of the present application or the prior art, the drawings required for use in the embodiments or the prior art descriptions are briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0020] Figure 1 It is a schematic diagram of an implementation environment of an interactive data processing method according to an exemplary embodiment.

[0021] Figure 2 It is a flowchart of an interactive data processing method according to an exemplary embodiment.

[0022] Figure 3 The present invention is a flowchart of a method for determining interaction data according to an exemplary embodiment.

[0023] Figure 4 The present invention is a schematic diagram of a process of determining a business processing result according to an exemplary embodiment.

[0024] Figure 5 The diagram is a logic diagram of an interactive data processing method according to an exemplary embodiment.

[0025] Figure 6 The figure is a block diagram of an interactive data processing device according to an exemplary embodiment.

[0026] Figure 7 It is a hardware structure block diagram of a server of an interactive data processing method provided according to an exemplary embodiment. DETAILED DESCRIPTION

[0027] The following will be combined with the drawings in the embodiments of the present application to clearly and completely describe the technical solutions in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of this application.

[0028] It should be noted that the terms "first", "second", etc. in the specification and claims of the embodiment of the present application and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that the data used in this way can be interchanged where appropriate, so that the embodiments of the embodiment of the present application described here can be implemented in an order other than those illustrated or described here. Thus, the features defined as "first" and "second" can expressly or implicitly include one or more of the features. In the description of the present embodiment, unless otherwise specified, the meaning of "multiple" is two or more. In addition, the terms "including" and "having" and any of their variations are intended to cover non-exclusive inclusions, for example, a process, method, system, product or server comprising a series of steps or units is not necessarily limited to those steps or units clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.

[0029] In order to make the purpose, technical solution and advantages disclosed in the embodiments of the present application more clearly understood, the embodiments of the present application are further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the embodiments of the present application and are not used to limit the embodiments of the present application.

[0030] In order to facilitate understanding of the above-mentioned technical solutions and the technical effects produced by the embodiments of the present application, the embodiments of the present application first explain the relevant professional terms:

[0031] A / B experiment: A method for testing the effect of an updated solution online. The experimental samples (e.g., the terminal objects displayed by the receiving object) are randomly divided into two mutually exclusive sample groups, the experimental group and the control group. The control group samples keep the original solution unchanged, while the experimental group uses the new solution. Finally, the difference between the two groups of results is compared to obtain the updated effect of the solution.

[0032] Layer domain model: used to manage and allocate business requests for A / B experiments. Business requests for A / B experiments can be divided into several subsets according to "domains", and business requests corresponding to different "domains" are mutually exclusive. In the same "domain", the "layers" corresponding to several modules are divided according to the business system modules, and each "layer" can reuse all business requests in the "domain".

[0033] Experimental domain: The business system divides the business requests within a certain period of time into subsets. The business requests corresponding to different experimental domains are mutually exclusive. Several sub-experimental domains or experimental layers can be nested in the same experimental domain.

[0034] Experimental layer: Different experiments are divided into multiple experimental layers according to information such as solution type or effective module. Each experimental layer stores the configuration information of several experiments and is a collection of experiments of the same type. Each experimental layer corresponds to the full amount of business requests. Business requests allocated to different solutions in the same experimental layer are mutually exclusive, and business requests in different experimental layers can be orthogonal to each other.

[0035] Experimental parameters: used to control the configuration of the logic of a module in the business system. By configuring different values ​​of the same parameter for the control group and the experimental group of the A / B experiment, different scheme logics can be effective for different groups. For example, an experimental parameter can be the font color of the display interface. The experimental group and the control group can set different values. For example, the control group can be (0, 255, 0), and the experimental group can be (0, 0, 255).

[0036] Parameter coverage: In the layer domain model, for several cross-experimental layer business requests that are hit at the same time, if the same parameter is configured in these experimental layers but with different values, then there are multiple alternative values ​​for the parameter in the business request. Since only one value can be effective for each parameter in a business request, it will eventually follow a certain rule to take the value and take effect. For the groups in other experimental layers, although they are hit by the business request, the values ​​of their parameters are not effective as expected. This is parameter coverage.

[0037] Consumption: refers to the increase in resources that can be brought to the display platform after the display object is released.

[0038] Exposure: refers to the terminal object being able to browse the display object after the display object is delivered to the terminal object.

[0039] Click: refers to the terminal object clicking on the display object after the display object is delivered to the terminal object.

[0040] Conversion: refers to the expected impact on the terminal object after the display object is delivered, such as downloading the application displayed in the display object, activating the application displayed in the display object, placing an order for the product displayed in the display object, etc. Conversion can refer to different behaviors in different scenarios.

[0041] In the layer domain model of the A / B experiment, the same experimental parameters may be configured in different experimental layers. When the business requests assigned by the layer domain model are reused in different experimental layers, the values ​​of these experimental parameters will conflict, so that the experimental parameter values ​​in each experiment cannot take effect as expected. This leads to the pollution of the experimental effective traffic, which directly affects the effect evaluation, resulting in a certain difference between the directly statistical experimental effect and the actual situation.

[0042] In view of this, the embodiments of the present application propose an interactive data processing method, device, equipment and storage medium, by determining the interactive data corresponding to the business request processed by the first test business and the second test business at the same time from the interactive test data to obtain candidate interactive test data, and determining the interactive data corresponding to the business request processed by the control business and the second test business at the same time from the interactive comparison data to obtain candidate interactive comparison data, so that the candidate interactive test data and the candidate interactive comparison data can be screened out to prevent the candidate interactive test data and the candidate interactive comparison data from interfering with the test results of the first business layer. Using the target interactive test data and the target interactive comparison data obtained after screening for performance evaluation processing can make the obtained performance evaluation data have higher accuracy, so that the optimization effect of the updated content can be more accurately evaluated, and then the business system is updated according to the performance evaluation data, which can improve the response efficiency of the business system. At the same time, the server computing resource consumption of the business system is reduced, and the server operation speed of the business system is improved.

[0043] Figure 1 FIG. 1 is a schematic diagram of an implementation environment of an interactive data processing method according to an exemplary embodiment. Figure 1 As shown, the implementation environment may include at least a terminal 01 and a server 02. The terminal 01 and the server 2 may be directly or indirectly connected via wired or wireless communication, and this application does not impose any limitation thereto.

[0044] Specifically, the terminal 01 can be used to send a service request to the server 02. Optionally, the terminal 01 can be a smart phone, a tablet computer, a laptop computer, a desktop computer, a smart speaker, an intelligent voice interaction device, a smart home appliance, a smart watch, a vehicle terminal, an aircraft, etc., but is not limited thereto. The embodiments of the present application can be applied to various scenarios, including but not limited to the sending of service requests involved in various scenarios such as cloud technology, artificial intelligence, smart transportation, and assisted driving.

[0045] Specifically, the server 02 can be used to obtain interaction data corresponding to a preset number of business requests; the interaction data is obtained by processing each business request based on the first business layer and the second business layer; the interaction data includes interaction test data corresponding to the first test business and interaction comparison data corresponding to the comparison business. It can also be used to determine the interaction data corresponding to the business requests processed by the first test business and the second test business at the same time from the interaction test data to obtain candidate interaction test data, and to determine the interaction data corresponding to the business requests processed by the comparison business and the second test business at the same time from the interaction comparison data to obtain candidate interaction comparison data. It can also be used to screen out candidate interaction test data in the interaction test data to obtain target interaction test data, and to screen out candidate interaction comparison data in the interaction comparison data to obtain target interaction comparison data. It can also be used to perform performance evaluation processing on the target interaction test data and the target interaction comparison data to obtain performance evaluation data. Optionally, the server 02 can be an independent physical server, or a server cluster or distributed system composed of multiple physical servers, or a cloud server that provides basic cloud computing services such as cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communications, middleware services, domain name services, security services, content delivery networks (CDN), as well as big data and artificial intelligence platforms.

[0046] It should be noted that Figure 1 This is just an example. In other scenarios, other implementation environments may also be included. In addition, the relevant information (including but not limited to feature information) and data (including but not limited to interactive data) of the terminal object involved in the embodiments of the present application are all information and data authorized by the terminal object or fully authorized by all parties.

[0047] Figure 2 is a flowchart of an interactive data processing method according to an exemplary embodiment. The method can be used Figure 1 In the implementation environment of the embodiment. This specification provides the method operation steps as described in the embodiment or flowchart, but may include more or fewer operation steps based on routine or non-creative work. The order of steps listed in the embodiment is only one way of executing the steps among many steps, and does not represent the only execution order. When the actual system or server product is executed, it can be executed in sequence or in parallel (for example, in a parallel processor or multi-threaded processing environment) according to the method shown in the embodiment or the drawings. Specifically, Figure 2 As shown, the method may include:

[0048] S101: Acquire interaction data corresponding to a preset number of business requests; the interaction data includes interaction test data and interaction comparison data.

[0049] In the embodiment of the present application, when the business system is updated, the effect of the updated content can be tested by designing a test plan. The business system can be a platform or application that provides a certain network service, such as a content publishing platform, a live broadcast platform, a social service platform, an e-commerce platform, an instant messaging application, a game application, an audio and video application, etc. The content of the business system update includes but is not limited to the update of business functions, the update of the human-computer interaction interface, the update of the background response strategy, the update of data storage, etc.

[0050] In an embodiment of the present application, when testing the optimization effect of the updated content, a preset number of business requests can be used to test the business logic containing the updated content and the original business logic of the business system. Specifically, the server of the business system can obtain a preset number of business requests. The content requested by the business request is the service content that the business system can provide, such as search and recall content, recommended article information, live broadcast, commodity, audio and video, searched account, group chat, etc. After the server of the business system obtains the preset number of business requests, it uses the business logic containing the updated content and the original business logic of the business system to process the business requests respectively, and obtains the business processing results corresponding to each business request. The business processing result can be the content requested by the business request, or it can be the associated content of the content requested by the business request. For example, search for recall content, or advertisements recommended based on search recall content. The server of the business system obtains the interactive processing result corresponding to each business request, and obtains the interactive data according to the interactive processing result, and then evaluates the optimization effect of the updated content according to the interactive data corresponding to each business request.

[0051] Specifically, the interaction data may be obtained by processing a preset number of business requests based on the first business layer and the second business layer. The first business layer and the second business layer are both experimental layers in the same experimental domain in the layer domain model. The first business layer and the second business layer may reuse all business requests corresponding to the experimental domain. In other words, the same business request may be processed by the business logic in the first business layer and the second business layer at the same time. Different types of experiments are respectively arranged in the first business layer and the second business layer. The first business layer may include a first test business and a first control business. The first test business is an experimental group in the first business layer, which includes a business logic to be tested, and the business logic to be tested is a business logic containing an updated content. The first control business is a control group in the first business layer, which includes a business logic used as a comparison of the first test business. Optionally, the business logic used as a comparison may be the original business logic of the business system, or may be a business logic obtained by updating the original business logic of the business system. The second business layer includes a second test business, and the second test business is an experimental group in the second business layer, which includes another business logic to be tested, and the business logic to be tested is a business logic containing another updated content. Correspondingly, the second business layer also includes a second comparison business, which includes a business logic for comparison with the second test business. Similarly, the business logic for comparison can be the original business logic of the business system, or the business logic updated from the original business logic of the business system.

[0052] It should be noted that the above description of the layer domain model is only a partial structure of the layer domain model, and does not mean that the layer domain model in the embodiment of the present application only includes the first business layer and the second business layer. In some embodiments, the layer domain model may include at least one experimental domain, each experimental domain may include at least one experimental layer, each experimental layer corresponds to an experiment, and each experimental layer may include a control group and at least one experimental group.

[0053] In an embodiment of the present application, the first service layer and the second service layer include the same control parameter. That is, the first test service, the first control service, the second test service and the second control service all include the same control parameter. The control parameter is an experimental parameter, which may have different values ​​in different service logics. In particular, the first test service, the first control service, the second test service and the second control service all include service logic that can independently process service requests. Without considering the content of the experiment, the first test service, the control service, the second test service and the second control service can all output the service processing results corresponding to the service request. As an example, the first service layer may be an experimental layer for the font color of the display object. The second service layer may be an experimental layer for other types of experiments. The control parameter may be the font color of the display object. Optionally, the display object may be an advertisement, a short video title, a web page, etc. In the first test service, the value of the control parameter may be (0, 0, 255), and in the first control service and the second test service, the value of the control parameter may be (0, 0, 0). Assume that the first test service, the first comparison service and the second test service each output a service processing result corresponding to the service request, the font color of the display object output by the first test service is blue, and the font color of the display object output by the first comparison service and the second test service is black.

[0054] In the embodiment of the present application, the interactive data is the business system processing the business request according to the preset business logic after receiving the business request sent by the preset terminal object, thereby obtaining the business processing result, and then sending the business processing result to the preset terminal object. The business system obtains the interactive processing result of the preset terminal object on the business processing result, thereby obtaining the interactive data corresponding to the business request. By using the above method to determine the interactive data corresponding to each business request, it can be ensured that the interactive data corresponding to the obtained business request is true and reliable, so that the optimization effect of the updated content can be accurately evaluated. Figure 3 is a flow chart of a method for determining interaction data according to an exemplary embodiment. Figure 3 As shown, the method for determining the interaction data includes:

[0055] S201: Receive a preset number of service requests.

[0056] In the embodiment of the present application, when determining the interactive data, the business system first needs to receive a preset number of business requests. The preset number of business requests can be requests sent to the business system by a number of preset terminal objects within a preset time period, such as business requests sent to the business system by all preset terminal objects within 3 days. The terminal object can be the object of all clients using the business system, and the preset terminal object can be part of all terminal objects. The terminal object can register an account in the business system to use the services provided by the business system. The account corresponding to the preset terminal object can be an account selected according to a certain strategy for testing the effect of business updates. For example, the business system is a game application. Before officially updating the game content, some game accounts are often selected to test the updated content. These accounts are the accounts corresponding to the preset terminal objects. The business request sent by the preset terminal object refers to a type of request that can trigger the business logic in the first business layer and the second business layer. When the business system receives these business requests, it will trigger the business request allocation logic configured in the layer domain model and allocate it to the first business layer and the second business layer for processing.

[0057] S203: Based on the first business layer and the second business layer, each business request is processed respectively to obtain a business processing result corresponding to each business request.

[0058] In an embodiment of the present application, for each business request, the business system will process it based on the first business layer and the second business layer respectively, and then determine the business processing result corresponding to each business request based on the results of the first business layer and the second business layer processing the business request respectively.

[0059] Specifically, the first business layer may include a first test business and a first comparison business. The second business layer may include a second test business and a second comparison business. Optionally, the business logic of the first comparison business and the second comparison business may be the same. When the business system receives a business request, the server of the business system sends the business request to the first business layer and the second business layer for processing. Among them, the first test business or the first comparison business in the first business layer processes the business request according to its business logic. For example, the business request is parsed to obtain the request information carried in the business request, and then the request content is determined according to the request information. Then, according to the values ​​of the control parameter configuration in the business, the values ​​of each control parameter of the request content are determined, so as to obtain the business processing result output by the first business layer. Similarly, the second test business or the second comparison business in the second business layer will also process the business request according to its business logic and output the business processing result of the second business layer.

[0060] As an example, for a business request for a display object sent by a terminal object, when the business system receives the business request, the business system sends the business request to the first business layer and the second business layer for processing. Among them, the first test business or the first comparison business in the first business layer processes the business request according to its business logic. Specifically, the first test business or the first comparison business parses the business request to obtain the display object information carried in the business request, such as the display object type, the display object identifier, etc., and then determines the corresponding display object according to the display object information, and then determines the values ​​of each control parameter of the display object according to the values ​​of the control parameter configuration in the business, thereby obtaining the business processing result output by the first business layer. Similarly, the second test business or the second comparison business in the second business layer will also process the business request according to its business logic and output the business processing result of the second business layer.

[0061] In an embodiment of the present application, for a preset number of service requests, the layer domain model can be allocated according to the grouping of the first service layer and the second service layer. For the first service layer, when the layer domain model allocates the preset number of service requests, since the same service request cannot be processed simultaneously by different groups in the same experimental layer, that is, the same service request cannot be processed by the first test service and the first control service at the same time, the preset number of service requests can be divided into the first number of first test service requests and the second number of first control service requests. The first test service request is processed by the first test service, and the first control service request is processed by the first control service. Similarly, for the second service layer, when the layer domain model allocates the preset number of service requests, the preset number of service requests can be divided into the third number of second test service requests and the fourth number of second control service requests. The second test service request is processed by the second test service, and the second control service request is processed by the second control service. In some embodiments, the first service layer and the second service layer may also include other groups. In this case, the layer domain model also needs to divide the service requests processed by these groups from the preset number of service requests.

[0062] In the embodiment of the present application, since the service requests in different experimental layers can be reused, there will be a situation where the service requests are cross-utilized. Table 1 is a distribution cross-relationship table of a preset number of service requests according to an exemplary embodiment. As shown in Table 1, T1 represents the first test service request, C1 represents the first control service request, and (1-T1-C1) represents the service requests corresponding to other groups in the first service layer. T2 represents the second test service request, C2 represents the second control service request, and (1-T2-C2) represents the service requests corresponding to other groups in the second service layer. T1*T2 represents the service request reused by the first test service and the second test service. Assuming that the number of service requests allocated to T1 and T2 in the preset number of service requests is 20%, the number of service requests reused by the first test service and the second test service is 4%. T1*C2 represents the service request reused by the first test service and the second control service. T1*(1-T2-C2) represents the service request reused by the first test service and other groups in the second service layer. C1*T2 represents the service request reused by the first control service and the second test service. C1*C2 indicates a service request multiplexed by the first control service and the second control service. C1*(1-T2-C2) indicates a service request multiplexed by the first control service and other groups of the second service layer. (1-T1-C1)*T2 indicates a service request multiplexed by other groups of the first service layer and the second test service. (1-T1-C1)*C2 indicates a service request multiplexed by other groups of the first service layer and the second control service. (1-T1-C1)*(1-T2-C2) indicates a service request multiplexed by other groups of the first service layer and other groups of the second service layer.

[0063] Table 1 A cross-relation table of the distribution of a preset number of business requests

[0064] T2 C2 (1-T2-C2) T1 T1*T2 T1*C2 T1*(1-T2-C2) C1 C1*T2 C1*C2 C1*(1-T2-C2) (1-T1-C1) (1-T1-C1)*T2 (1-T1-C1)*C2 (1-T1-C1)*(1-T2-C2)

[0065] Based on the above results of allocating a preset number of business requests, each business request can be processed. Specifically, each first test business request is processed based on the first test business, and a first test processing result corresponding to each first test business request is obtained. Each first comparison business request is processed based on the first control business, and a first comparison processing result corresponding to each first comparison business request is obtained. Each second test business request is processed based on the second test business, and a second test processing result corresponding to each second test business request is obtained. Each second comparison business request is processed based on the second control business, and a second comparison processing result corresponding to each second comparison business request is obtained. By allocating a preset number of business requests in the first business layer and the second business layer respectively, business requests can be reused to achieve simultaneous testing of multiple update contents, shorten the test cycle, and improve test efficiency.

[0066] In an embodiment of the present application, after obtaining the first test processing result, the second test processing result, the first control processing result and the second control processing result, the business processing result corresponding to each business request can be determined based on each first test processing result, each second test processing result, each first control processing result and each second control processing result.

[0067] In actual applications, when a preset terminal object sends a service request to the business system, the business system determines the business logic for processing the service request according to the layer domain model. The service request can be processed by different experimental layers at the same time, but for the same experimental layer, the service request can only be processed by one business logic. For example, when the service request is processed by the business logic corresponding to the first test service in the first business layer, the service request can no longer be processed by the control service in the first business layer, but the service request can be processed by the second test service in the second business layer. When the service request is processed by the business logic in the first business layer and the second business layer at the same time, the processing results corresponding to the two business logics can be obtained respectively. Then, the business processing result finally sent to the preset terminal object is determined according to the processing results corresponding to the two business logics. For example, the first test processing result and the second test processing result corresponding to the service request processed by the first test service and the second test service at the same time can determine the first type of processing result. The first type of processing result is used to send to the preset terminal object corresponding to the service request to obtain the interactive test data corresponding to the service request. For another example, for a service request processed by both the first comparison service and the second test service, the first comparison processing result and the second test processing result corresponding to the service request can determine the second type of processing result. The second type of processing result is used to send to the preset terminal object corresponding to the service request to obtain the interactive comparison data corresponding to the service request.

[0068] Specifically, Figure 4 is a schematic diagram of a process for determining a business processing result according to an exemplary embodiment. Figure 4 As shown, the third number of second test service requests may include first-class service requests and second-class service requests. Among them, the first-class service request is a service request processed by the first test service and the second test service at the same time, and the second-class service request is a service request processed by the first comparison service and the second test service at the same time. Obviously, the first-class service request is a part of the service request in the first test request, and the second-class service request is a part of the service request in the first comparison service request. When determining the service processing result corresponding to each service request, it can be specifically determined based on the first test processing result and the second test processing result corresponding to each first-class service request. The first-class processing result is used to send to the preset terminal object corresponding to the first-class service request to obtain the interactive test data corresponding to the first-class service request. Based on the comparison processing result and the second test processing result corresponding to each second-class service request, the second-class processing result corresponding to each second-class service request is determined. The second-class processing result is used to send to the preset terminal object corresponding to the second-class service request to obtain the interactive comparison data corresponding to the second-class service request. After obtaining the first-class processing result and the second-class processing result, each first-class processing result and each second-class processing result can be determined as the service processing result corresponding to each service request. By dividing the second test business request into first-category business requests and second-category business requests, we can obtain processing results corresponding to different types of business requests, so that each business request can obtain the corresponding business processing result, thereby ensuring the normal operation of the business system while conducting multiple experiments at the same time.

[0069] In the embodiment of the present application, since the first test service and the second test service have the same control parameter. For a service request that is processed by the first test service and the second test service at the same time, the value corresponding to the parameter in the corresponding service processing result can only be one. In this case, when determining the first type of processing result, it is necessary to set the effectiveness policy of the control parameter to ensure the normal operation of the service system. Specifically, the first test processing result corresponding to each first type of service request includes the first control data corresponding to the control parameter in the first test service. The second test processing result corresponding to each first type of service request includes the second control data corresponding to the control parameter in the second test service. When determining the first type of processing result corresponding to each first type of service request, the first type of processing result corresponding to each first type of service request can be determined based on the priority of each first control data and the priority of each second control data. That is to say, which control data has a higher effectiveness priority in the first control data and the second control data, which control data is effective in the first type of processing result. By determining the first type of processing result according to the effectiveness priority of the first control data and the second control data, it is ensured that the first type of service request can obtain the corresponding service processing result, and the normal operation of the service system is guaranteed.

[0070] Similarly, for a service request that is processed by both the first control service and the second test service, the value corresponding to the parameter in the corresponding service processing result can only be one. Therefore, when determining the second-class processing result, the second-class processing result can also be determined according to the effectiveness policy of the above-mentioned control parameters. Specifically, the control processing result corresponding to each second-class service request includes the third control data corresponding to the control parameter in the control service. The second test processing result corresponding to each second-class service request includes the fourth control data corresponding to the control parameter in the second test service. When determining the second-class processing result corresponding to each second-class service request, the second-class processing result corresponding to each second-class service request is determined based on the priority of each third control data and the priority of each fourth control data. That is to say, which control data has a higher effectiveness priority among the third control data and the fourth control data, which control data is effective in the second-class processing result. By determining the second-class processing result according to the effectiveness priority of the third control data and the fourth control data, it is ensured that the second-class service request can obtain the corresponding service processing result, and the normal operation of the service system is guaranteed.

[0071] In the embodiment of the present application, the business processing result corresponding to each business request finally obtained may be the content requested by the business request, or may be the content associated with the content requested by the business request. For example, the content requested by the business request may be a request to the business system to display a fitness video. The business processing result obtained after the business system processes the business request may be a fitness video recalled by the business system according to the search conditions in the business request, or may be a display object related to fitness, such as an advertisement for fitness equipment.

[0072] S205: Send the business processing result corresponding to each business request to the preset terminal object corresponding to each business request, so as to obtain the interactive processing result between the preset terminal object corresponding to each business request and the business processing result corresponding to each business request.

[0073] In an embodiment of the present application, after obtaining the business processing result corresponding to each business request, the business processing result can be sent to the preset terminal object corresponding to each business request. The preset terminal object that receives the business processing result is the account that sends the business request. After the preset terminal object receives the business processing result, the business processing result can be interactively processed. Depending on the business processing result, the preset terminal object can interact with the business processing result in different ways. For example, the business processing result is a display object, and the preset terminal object can interact with the business processing result by interactive processing methods such as clicking, browsing, downloading, purchasing, etc. Of course, the preset terminal object can also directly close or delete the business processing result. Regardless of how the preset terminal object interacts with the business processing result, a corresponding interactive processing result will be generated.

[0074] S207: Based on the interaction processing result between the preset terminal object corresponding to each business request and the business processing result corresponding to each business request, the interaction data corresponding to each business request is obtained.

[0075] In an embodiment of the present application, for any business request, after the interaction processing result between the preset terminal object that sends the business request and the business processing result is obtained, the interaction data corresponding to the business request can be determined according to the interaction processing result. The interaction data may be data directly associated with the interaction processing result, such as exposure, click, conversion and other data. As an example, if the interaction processing result is that the preset terminal object clicks on the business processing result corresponding to the business request, then the interaction data corresponding to the business request is click. Correspondingly, if the interaction processing result is that the preset terminal object closes the business processing result corresponding to the business request, then the interaction data corresponding to the business request is not clicked. The interaction data may also be data indirectly associated with the interaction processing result, such as consumption, resource transfer quantity and other data.

[0076] In an embodiment of the present application, the interaction data may include interaction test data and interaction comparison data. The interaction test data and the interaction comparison data are used to determine the performance evaluation data. Since the performance evaluation data corresponds to the experimental layer, the interaction test data and the interaction comparison data are also relative to a certain experimental layer. For example, for the first service layer, the interaction test data includes data corresponding to the first test service and the second test service, and data corresponding to the first test service and the second comparison service. The interaction comparison data includes data corresponding to the first comparison service and the second test service, and data corresponding to the first comparison service and the second comparison service.

[0077] S103: Determine, from the interaction test data, the interaction data corresponding to the service requests processed by both the first test service and the second test service to obtain candidate interaction test data; and determine, from the interaction comparison data, the interaction data corresponding to the service requests processed by both the comparison service and the second test service to obtain candidate interaction comparison data.

[0078] In the embodiment of the present application, after obtaining the interaction data corresponding to the preset number of service requests, the optimization effect of each test service can be evaluated based on the interaction data. In order to ensure the accuracy of the evaluation results, before evaluating the optimization effect of each test service, the interaction data corresponding to the parameter coverage situation needs to be eliminated to prevent the parameter coverage from interfering with the evaluation results.

[0079] In general, the business logic corresponding to the control service mostly adopts the original business logic in the business system, and the effectiveness priority of the control parameters of the control service is often lower than the control parameter priority of the test service. Even in the case of parameter coverage, the control service will not affect the interaction data obtained by other experimental layers. However, the test service may affect the interaction data obtained by other experimental layers. For example, when the effectiveness priority of the control parameters in the second test service is higher than that of the first test service, the control parameters in the first test service cannot take effect as expected, and the interaction data corresponding to the first test service will be affected by the parameter coverage. In this case, when evaluating the optimization effect of the first test service, it is necessary to eliminate the influence of the parameter coverage on the interaction test data and the interaction control data. When the effectiveness priority of the control parameters in the first test service is higher than that of the second test service, the control parameters in the first test service can take effect as expected, and the interaction data corresponding to the first test service will not be affected by the parameter coverage, but the interaction control data obtained by the first control service will be affected by the parameter coverage because the effectiveness priority of the control parameters is lower than that of the second test service. In this case, when evaluating the optimization effect of the first test service, it is necessary to eliminate the influence of the parameter coverage on the interaction control data. Similarly, the interaction data corresponding to the second service layer will also be affected by the first test service. The situation where the first test service affects the interaction data corresponding to the second service layer is the same as the situation where the second test service affects the interaction data corresponding to the first service layer.

[0080] In an embodiment of the present application, for any experimental layer, when evaluating the optimization effect of the test service in the experimental layer, it is necessary to find the interaction data corresponding to the service requests that are not reused by another experimental layer in the interaction data. Taking the first service layer as an example, according to the service request allocation method shown in Table 1, the service requests that are not reused by the second test service layer are: T1*C2, T1*(1-T2-C2), C1*C2, C1*(1-T2-C2). The optimization effect can be represented by performance evaluation data. When calculating the performance evaluation data corresponding to the first test service, the interaction test data corresponding to the first test service can be divided by the interaction control data corresponding to the first control service, and then subtracted from 1 to obtain. Assuming that the service request represents the corresponding interaction data, the performance evaluation data corresponding to the first test service (Lift ExpGroup 1 ) can be calculated according to the following formula:

[0081]

[0082] According to formula (1), to calculate the performance evaluation data of the first test service, it is necessary to exclude the interaction data corresponding to the service requests processed by both the first test service and the second test service, and the interaction data corresponding to the service requests processed by both the first control service and the second test service. Conversely, the method for calculating the performance evaluation data of the second test service is the same as the method for calculating the performance evaluation data corresponding to the first test service.

[0083] In an embodiment of the present application, for any experimental layer, when eliminating the interactive data affected by parameter coverage, the interactive data obtained by the test services in other experimental layers can be determined first. Continuing with the first service layer as an example, after a preset number of service requests are processed by the first service layer, the interactive data obtained may include interactive test data corresponding to the first test service and interactive comparison data corresponding to the first comparison service. Among them, the interactive test data corresponding to the first test service includes data processed by both the first test service and the second test service. And the interactive comparison data corresponding to the first comparison service includes data processed by both the first comparison service and the second test service. The service processing results corresponding to these interactive data are affected by parameter coverage, and the values ​​of the control parameters in the service processing results are not the values ​​of the control parameters configured in the first test service or the first comparison service. Therefore, the interactive data obtained based on this situation cannot accurately measure the optimization effect of the first test service. When evaluating the optimization effect of the first test service, it is necessary to determine this part of the interactive data and eliminate it.

[0084] In an embodiment of the present application, when determining the interactive data obtained by the influence of parameter coverage in the interactive test data and the interactive comparison data, it can be determined according to the request log data corresponding to the interactive test data and the interactive comparison data. Specifically, first obtain the request log data corresponding to each business request, the request log data includes the business logic executed by the business request and the obtained interactive data. Then, based on the request log data corresponding to each business request, determine the interactive data corresponding to the business request processed by the first test business and the second test business at the same time from the interactive test data to obtain candidate interactive test data, and determine the interactive data corresponding to the business request processed by the first comparison business and the second test business at the same time from the interactive comparison data to obtain candidate interactive comparison data. By obtaining the request log data corresponding to each business request, the candidate interactive test data and the candidate interactive test data can be quickly determined from the interactive data, thereby improving the efficiency of eliminating the interactive data obtained by the influence of parameter coverage, and at the same time improving the accuracy of the test business evaluation.

[0085] S105: Screening out candidate interaction test data in the interaction test data to obtain target interaction test data, and screening out candidate interaction comparison data in the interaction comparison data to obtain target interaction comparison data.

[0086] In an embodiment of the present application, after determining candidate interaction test data from the interaction test data and determining candidate interaction comparison data from the interaction comparison data, the candidate interaction test data and the candidate interaction test data can be filtered and removed from the interaction test data and the interaction comparison data, respectively, so as not to affect the accuracy of the optimization effect evaluation.

[0087] S107: Perform performance evaluation processing on the target interaction test data and the target interaction control data to obtain performance evaluation data.

[0088] In an embodiment of the present application, after the candidate interaction test data and the candidate interaction test data are filtered out and removed from the interaction test data and the interaction control data respectively, the obtained target interaction test data and target interaction control data are pure interaction data that are not affected by parameter coverage. Therefore, the optimization effect of the first test service can be evaluated based on the target interaction test data and the target interaction control data.

[0089] Generally speaking, for any experimental layer, the business requests processed by the experimental group and the control group in the experimental layer are allocated according to certain feature dimensions. Specifically, the business sent by the preset terminal object is marked according to its feature information, and then the preset number of business requests are divided into different groups according to the dimension of feature information. For example, the preset number is 200, and these 200 business requests correspond to feature information of whether to like apples. According to the feature information of liking apples and not liking apples, the business requests can be divided into two categories, of which 40 like apples and 160 do not like apples. When allocating these businesses between the experimental group and the control group, in order to ensure the accuracy of the experimental results, business requests of different feature dimensions can be evenly allocated between the experimental group and the control group. For example, among the business requests processed by the experimental group, 80 business requests have the feature information of not liking apples, and 20 business requests have the feature information of liking apples. Similarly, among the business requests processed by the control group, 80 business requests have the feature information of not liking apples, and 20 business requests have the feature information of liking apples.

[0090] After eliminating the interaction data affected by parameter coverage, the data structure of the experimental design may be destroyed. The characteristic dimension ratio of the business request corresponding to the target interaction test data and the target interaction control data may be different from the characteristic dimension ratio of the business request corresponding to the interaction test data and the interaction control data. In this case, the Simpson's paradox problem will be introduced, resulting in inaccurate performance evaluation data finally calculated. Therefore, in order to avoid this situation, after eliminating some interaction data, it is necessary to restore its data volume according to the characteristic dimensions of the corresponding business request to avoid affecting the accuracy of the final performance evaluation data.

[0091] Specifically, before the target interactive test data and the target interactive control data are subjected to performance evaluation processing, it is necessary to first perform data volume restoration processing on the target interactive test data to obtain interactive test restoration data, and to perform data volume restoration processing on the target interactive control data to obtain interactive control restoration data. Then, performance evaluation processing is performed based on the interactive test restoration data and the interactive control restoration data to obtain performance evaluation data. As an example, the data volume of the interactive test data is 100, and the target interactive test data obtained after screening out the candidate interactive test data is 90. In order not to introduce the Simpson's paradox problem, before calculating the performance evaluation data, the target interactive test data can be divided by 90% to restore the data volume of the target interactive test data. Similarly, the restoration method of the target interactive control data can also refer to the above method. By performing data volume restoration processing on the target interactive test data and the target interactive control data, the Simpson's paradox problem introduced after the candidate interactive test data and the candidate interactive control data are eliminated is avoided, thereby ensuring the accuracy of the performance evaluation data finally calculated.

[0092] In the embodiment of the present application, the interactive test data and the interactive control data are usually obtained by the business request of at least one dimension. Therefore, when restoring the data volume of the target interactive test data and the target interactive control data, the data volume can be restored according to each dimension respectively, so as to improve the accuracy of data restoration. Specifically, the target interactive test data includes the target interactive test sub-data of at least one dimension, and the target interactive control data includes the target interactive control sub-data of at least one dimension. When performing data volume restoration processing, the target interactive test sub-data of each dimension is subjected to data volume restoration processing to obtain the target interactive test restoration sub-data of each dimension, and the target interactive control sub-data of each dimension is subjected to data volume restoration processing to obtain the target interactive control restoration sub-data of each dimension. Then, the target interactive test restoration sub-data of each dimension is fused to obtain the interactive test restoration data, and the target interactive control restoration sub-data of each dimension is fused to obtain the interactive control restoration data. By restoring the target interactive test sub-data and the target interactive control sub-data corresponding to each dimension, and then fusing the target interactive test restoration sub-data and the target interactive control restoration sub-data of each dimension, the obtained interactive test restoration data and the interactive control restoration data can be consistent with the data composition structure of the interactive test data and the interactive control data, so as to realize the accurate calculation of the performance evaluation data.

[0093] As an example, taking the restoration of target interaction test data as an example, it is assumed that the interaction test data includes target interaction test data corresponding to liking apples and target interaction test data corresponding to not liking apples. When restoring the data volume of the target interaction test data, not only the data volume of the target interaction test data corresponding to liking apples needs to be restored, but also the data volume of the target interaction test data corresponding to not liking apples needs to be restored. Then, the target interaction test restoration sub-data obtained by restoring the target interaction test data corresponding to liking apples is added to the target interaction test restoration sub-data obtained by restoring the target interaction test data corresponding to not liking apples, so as to obtain the interaction test restoration data corresponding to the target interaction test data. Similarly, the restoration method of the target interaction control data can be the same as the restoration method of the target interaction test data.

[0094] In an embodiment of the present application, the interactive test data may include interactive test sub-data corresponding to each dimension. When the target interactive test sub-data of each dimension is subjected to data volume restoration processing, for any dimension, the proportion of the target interactive test sub-data in the dimension in the interactive test sub-data in the corresponding dimension can be determined first, and then the target interactive test restoration sub-data of the dimension can be determined. Specifically, when the target interactive test sub-data of each dimension is subjected to data volume restoration processing, the first target proportion of the target interactive test sub-data of each dimension can be determined based on the interactive test sub-data corresponding to each dimension and the target interactive test sub-data of each dimension. Among them, the first target proportion is used to characterize the proportion of the target interactive test sub-data of each dimension in the interactive test sub-data corresponding to each dimension. Then, based on the first target proportion of the target interactive test sub-data of each dimension and the target interactive test sub-data of each dimension, the target interactive test restoration sub-data of each dimension is determined. By determining the first target proportion of the target interactive test sub-data of each dimension, and determining the target interactive test restoration sub-data of each dimension based on the first target proportion and the target interactive test sub-data of each dimension, the determination of the target interactive test restoration sub-data of each dimension can be realized quickly and accurately, while ensuring the accurate evaluation of the optimization effect of the test business, the evaluation efficiency is improved.

[0095] As an example, suppose that the number of interactive test data corresponding to loving apples in the interactive test data is 80, and the number of interactive test data corresponding to not liking apples is 20, and the number of interactive test data corresponding to loving apples in the target interactive test data is 60, and the number of interactive test data corresponding to not liking apples is 10. Then for the dimension of loving apples, the first target accounts for 3 / 4. When restoring the target interactive test sub-data corresponding to this dimension, 60 can be divided by 3 / 4 to achieve the restoration of the target interactive test sub-data in the dimension of loving apples. When restoring the target interactive test sub-data in the dimension of not liking apples, the target interactive test sub-data in the dimension of not liking apples can be used, divided by the first target proportion corresponding to the dimension of not liking apples, to obtain the target interactive test restoration sub-data corresponding to the dimension of not liking apples.

[0096] In an embodiment of the present application, the interactive comparison data may include interactive comparison sub-data corresponding to each dimension. When the target interactive comparison sub-data of each dimension is subjected to data volume restoration processing, for any dimension, the proportion of the target interactive comparison sub-data in the interactive comparison sub-data in the corresponding dimension can be determined first, and then the target interactive comparison restoration sub-data of the dimension can be determined. Specifically, when the target interactive comparison sub-data of each dimension is subjected to data volume restoration processing, the second target proportion of the target interactive comparison sub-data of each dimension can be determined based on the interactive comparison sub-data corresponding to each dimension and the target interactive comparison sub-data of each dimension. Among them, the second target proportion is used to characterize the proportion of the target interactive comparison sub-data of each dimension in the interactive comparison sub-data corresponding to each dimension. Then, based on the second target proportion of the target interactive comparison sub-data of each dimension and the target interactive comparison sub-data of each dimension, the target interactive comparison restoration sub-data of each dimension is determined. By determining the first target proportion of the target interactive comparison sub-data of each dimension, and determining the target interactive comparison restoration sub-data of each dimension based on the first target proportion and the target interactive comparison sub-data of each dimension, the determination of the target interactive comparison restoration sub-data of each dimension can be realized quickly and accurately, while ensuring the accurate evaluation of the optimization effect of the test business, the evaluation efficiency is improved.

[0097] As an example, suppose that the number of interactive control data corresponding to liking apples in the interactive control data is 80, and the number of interactive control data corresponding to not liking apples is 20, and the number of interactive control data corresponding to liking apples in the target interactive control data is 40, and the number of interactive control data corresponding to not liking apples is 10. Then for the dimension of liking apples, the first target accounts for 50%. When restoring the target interactive control sub-data corresponding to this dimension, 40 can be divided by 50%, thereby realizing the restoration of the target interactive control sub-data in the dimension of liking apples. When restoring the target interactive control sub-data in the dimension of not liking apples, the target interactive control sub-data in the dimension of not liking apples can be used, divided by the first target proportion corresponding to the dimension of not liking apples, so as to obtain the target interactive control restored sub-data corresponding to the dimension of not liking apples.

[0098] In an embodiment of the present application, after obtaining the interactive test restoration data and the interactive comparison restoration data, the restoration process restores only the amount of data, not the real interactive data. Therefore, after obtaining the interactive test restoration data and the interactive comparison restoration data, it is also necessary to determine the number of different interactive processing results in the interactive test restoration data and the interactive comparison restoration data according to the proportion of different interactive processing results in the target interactive test data and the target interactive comparison data. Specifically, the target interactive test data includes at least two different types of interactive data, and the proportion of each type of interactive data in the target interactive test data is determined, and then the number of each type of interactive data in the interactive test restoration data is determined based on the proportion of each type of interactive data in the target interactive test data. Similarly, the target interactive comparison data includes at least two different types of interactive data, and the proportion of each type of interactive data in the target interactive comparison data is determined, and then the number of each type of interactive data in the interactive comparison restoration data is determined based on the proportion of each type of interactive data in the target interactive comparison data.

[0099] As an example, the target interaction test data includes two types of interaction data: click and non-click. Assuming that the data volume of the target interaction test data is 100, and the data volume of the interaction test restoration data is 200, the number of click-type interaction data in the target interaction test data is 90, and the number of non-click-type interaction data is 10. Then the proportion of click-type interaction data in the target interaction test data is 90%, and the proportion of non-click-type interaction data in the target interaction test data is 10%. The data volume of click-type interaction data in the interaction test restoration data is 200*90%=180. The data volume of non-click-type interaction data in the interaction test restoration data is 200*10%=20. Similarly, the method for determining the data volume of each type in the interaction control restoration data is the same as the method for determining the data volume of each type in the interaction test restoration data.

[0100] In the embodiment of the present application, after the amount of each type of data in the interactive test restoration data and the interactive comparison restoration data is calculated, the performance evaluation data can be calculated based on the data of each type in the interactive test restoration data and the interactive comparison restoration data. In some embodiments, if there is a difference between the number of business requests processed by the test business and the number of business requests processed by its corresponding comparison business, before calculating the performance evaluation data, it is also necessary to convert the amount of each type of data in the interactive test restoration data and the interactive comparison restoration data according to the ratio of the number of business requests processed by the test business to the number of business requests processed by its corresponding comparison business, and then calculate the performance evaluation data based on the conversion result.

[0101] As an example, assume that the ratio of the number of business requests processed by the test business to the number of business requests processed by its corresponding control business is 2:1, the number of click-type interaction data in the interactive test restoration data is 500, and the number of click-type interaction data in the interactive control restoration data is 200. Before calculating the performance evaluation data, it is necessary to convert the number of click-type interaction data in the interactive test restoration data and the interactive control restoration data according to the ratio of business requests. Specifically, the conversion method of click-type interaction data in the interactive test restoration data is: 500 / (2 / 3)=750, and the conversion method of click-type interaction data in the interactive control restoration data is: 200 / (1 / 3)=600. Then, the performance evaluation data is calculated using the converted data, specifically: 750 / 600-1=0.25.

[0102] In the embodiment of the present application, Figure 5 is a logic diagram of an interactive data processing method according to an exemplary embodiment. Figure 5 As shown, when the interactive data processing method is specifically implemented, the candidate interactive test data can be first screened out from the interactive test data to obtain the target interactive test data. At the same time, the candidate interactive control data can be screened out from the interactive control data to obtain the target interactive control data. Then, the data volume of the target interactive test data and the target interactive control data is restored to obtain the interactive test restored data and the interactive control restored data. Then, according to the ratio of the number of business requests processed by the test business to the number of business requests processed by its corresponding control business, the data volume of each type in the interactive test restored data and the interactive control restored data is converted, so that the data volume of each type after conversion is in a directly comparable dimension. Finally, the converted data can be used to calculate the performance evaluation data.

[0103] The interactive data processing method described in the embodiment of the present application, by determining candidate interactive test data and candidate interactive control data and screening these data, effectively eliminates the interference of parameter coverage on experimental evaluation, restores the real improvement of the experiment, and makes the evaluation conclusion more real and credible.

[0104] In the embodiment of the present application, after actual testing, the performance evaluation data obtained by the interactive data processing method described in the embodiment of the present application is more accurate than the performance evaluation data obtained by not excluding the influence of parameter coverage. Table 2 is a performance evaluation data comparison table shown according to an exemplary embodiment. As shown in Table 2, consumption is used as interactive data, the real improvement refers to the performance evaluation data obtained when the experiment is run alone, the original improvement refers to the performance evaluation data obtained by not excluding the influence of parameter coverage, and the pure improvement refers to the performance evaluation data obtained by using the interactive data processing method described in the embodiment of the present application. The experiment is not interfered by parameter coverage, and the consumption is improved by 0.11%, and it is significant. In the case of parameter coverage in the experiment, the consumption is improved by 0.06%, and it is not significant. However, the interactive data processing method described in the embodiment of the present application is used to exclude the influence of parameter coverage when calculating the performance evaluation data, and the consumption obtained is improved by 0.12%, and it is significant. It can be seen that the performance evaluation data obtained by the interactive data processing method described in the embodiment of the present application is closer to the real performance evaluation data obtained without interference from parameter coverage.

[0105] Table 2 A performance evaluation data comparison table

[0106]

[0107]

[0108] In the embodiments of the present application, it was found through actual testing that the existence of parameter coverage will affect the accuracy of the experimental evaluation conclusions. The experimental evaluation conclusions obtained when the influence of parameter coverage is excluded may cause the experimental evaluation conclusions obtained under parameter coverage to be reversed. Table 3 is a statistical table of experimental evaluation conclusion reversals according to an exemplary embodiment. As shown in Table 3, the table shows the number of experiments in which the experimental evaluation conclusions are reversed. The interactive data processing method described in the embodiments of the present application can provide more realistic and reliable experimental evaluation conclusions, thereby providing a more accurate and reliable basis for the update of the business system.

[0109] Table 3. A statistical table of experimental evaluation conclusion reversal

[0110]

[0111] In the embodiment of the present application, after obtaining the performance evaluation data, it can be determined whether to update the business system according to the performance evaluation data. Specifically, for a certain experiment, after obtaining the performance evaluation data corresponding to the test business, it is determined whether there is an improvement relative to the current business logic in the business system, and whether the degree of improvement exceeds the threshold. If there is an improvement and the degree of improvement exceeds the threshold, the business logic of the test business in the experiment can be used to update the current business logic in the business system. For example, for the experiment of the font color of the display object, it is assumed that the performance evaluation data corresponding to the test business is clicks, and the improvement threshold is 0.5%. If the clicks of the test business are improved by 1%, it can be determined that the optimization effect of the test business is significantly improved, and the test business can be used to update the current business logic in the business system. If the clicks of the test business are improved by 0.2%, it can be determined that the optimization effect of the test business is not significantly improved. If the test business is used to update the current business logic in the business system, the cost is high. In this case, it is not suitable to use the test business to update the current business logic in the business system.

[0112] The embodiment of the present application also provides an interactive data processing device. Figure 6 is a block diagram of an interactive data processing device according to an exemplary embodiment. Figure 6 As shown, the device may include at least:

[0113] The acquisition module 301 is used to acquire interaction data corresponding to a preset number of business requests; the interaction data is interaction data between a preset terminal object and a business processing result; the business processing result is obtained by processing a preset number of business requests based on the first business layer and the second business layer; the first business layer and the second business layer contain the same control parameters; the first business layer includes a first test business and a first comparison business; the second business layer includes a second test business and a second comparison business; the interaction data includes interaction test data and interaction comparison data; the interaction test data includes data corresponding to the first test business and the second test business, and data corresponding to the first test business and the second comparison business; the interaction comparison data includes data corresponding to the first comparison business and the second test business, and data corresponding to the first comparison business and the second comparison business;

[0114] The candidate interaction data determination module 303 is used to determine the interaction data corresponding to the service request processed by the first test service and the second test service from the interaction test data to obtain the candidate interaction test data, and to determine the interaction data corresponding to the service request processed by the reference service and the second test service from the interaction comparison data to obtain the candidate interaction comparison data;

[0115] A screening module 305 is used to screen out candidate interaction test data in the interaction test data to obtain target interaction test data, and to screen out candidate interaction comparison data in the interaction comparison data to obtain target interaction comparison data;

[0116] The performance evaluation data determination module 307 is used to perform performance evaluation processing on the target interaction test data and the target interaction control data to obtain performance evaluation data.

[0117] In some optional embodiments, the device further includes an interaction data determination module, and the interaction data determination module includes:

[0118] A service request receiving submodule, used to receive a preset number of service requests;

[0119] A business processing submodule, used to process each business request based on the first business layer and the second business layer, and obtain a business processing result corresponding to each business request;

[0120] The interactive processing result acquisition submodule is used to send the business processing result corresponding to each business request to the preset terminal object corresponding to each business request, so as to obtain the interactive processing result between the preset terminal object corresponding to each business request and the business processing result corresponding to each business request;

[0121] The interaction data determination submodule is used to obtain the interaction data corresponding to each business request based on the interaction processing result between the preset terminal object corresponding to each business request and the business processing result corresponding to each business request.

[0122] In some optional embodiments, the preset number of service requests includes a first number of first test service requests, a second number of control service requests, and a third number of second test service requests; and the service processing submodule includes:

[0123] A first test processing result determining unit is used to process each first test service request based on the first test service to obtain a first test processing result corresponding to each first test service request; the first test processing result is used to determine a first type of processing result, and the first type of processing result is used to send to a preset terminal object corresponding to the first test service request to obtain interactive test data corresponding to the first test service request;

[0124] A comparison processing result determination unit is used to process each comparison service request based on the comparison service to obtain a comparison processing result corresponding to each comparison service request; the comparison processing result is used to determine a second type of processing result, and the second type of processing result is used to send to a preset terminal object corresponding to the comparison service request to obtain interactive comparison data corresponding to the comparison service request;

[0125] A second test processing result determining unit, configured to process each second test service request based on the second test service to obtain a second test processing result corresponding to each second test service request;

[0126] The business processing result determining unit is used to determine the business processing result corresponding to each business request based on each first test processing result, each comparison processing result and each second test processing result.

[0127] In some optional embodiments, the third number of second test service requests includes first-category service requests and second-category service requests, wherein the first-category service requests are service requests processed by both the first test service and the second test service; the second-category service requests are service requests processed by both the control service and the second test service; and the service processing result determination unit includes:

[0128] A first-category processing result determination subunit, configured to determine the first-category processing result corresponding to each first-category service request based on the first test processing result and the second test processing result corresponding to each first-category service request;

[0129] A second-category processing result determination subunit, configured to determine the second-category processing result corresponding to each second-category service request based on the comparison processing result corresponding to each second-category service request and the second test processing result;

[0130] The business processing result determination subunit is used to determine each first-category processing result and each second-category processing result as a business processing result corresponding to each business request.

[0131] In some optional embodiments, the first test processing result corresponding to each first-type service request includes first control data corresponding to a control parameter in the first test service; the second test processing result corresponding to each first-type service request includes second control data corresponding to a control parameter in the second test service; the first-type processing result determination subunit includes:

[0132] The first-category processing result determination component is used to determine the first-category processing result corresponding to each first-category service request based on the priority of each first control data and the priority of each second control data.

[0133] In some optional embodiments, the comparison processing result corresponding to each second-type service request includes third control data corresponding to the control parameter in the comparison service; the second test processing result corresponding to each second-type service request includes fourth control data corresponding to the control parameter in the second test service; the second-type processing result determination subunit includes:

[0134] The second-type processing result determination component is used to determine the second-type processing result corresponding to each second-type service request based on the priority of each third control data and the priority of each fourth control data.

[0135] In some optional embodiments, the candidate interaction data determination module includes:

[0136] The request log data acquisition submodule is used to obtain the request log data corresponding to each business request; the request log data includes the business logic executed by the business request and the obtained interaction data;

[0137] The candidate interaction data determination submodule is used to determine, based on the request log data corresponding to each business request, from the interaction test data, the interaction data corresponding to the business requests processed by both the first test business and the second test business, to obtain candidate interaction test data, and to determine, from the interaction comparison data, the interaction data corresponding to the business requests processed by both the comparison business and the second test business, to obtain candidate interaction comparison data.

[0138] In some optional embodiments, the performance evaluation data determination module includes:

[0139] The restoration processing submodule is used to perform data volume restoration processing on the target interactive test data to obtain interactive test restoration data, and to perform data volume restoration processing on the target interactive comparison data to obtain interactive comparison restoration data;

[0140] The performance evaluation data determination submodule is used to perform performance evaluation processing on the interactive test restoration data and the interactive control restoration data to obtain performance evaluation data.

[0141] In some optional embodiments, the target interaction test data includes target interaction test sub-data of at least one dimension; the target interaction comparison data includes target interaction comparison sub-data of at least one dimension; and the restoration processing sub-module includes:

[0142] A target interaction test restored sub-data determining unit is used to perform data volume restoration processing on the target interaction test sub-data of each dimension to obtain the target interaction test restored sub-data of each dimension;

[0143] A target interactive comparison restoration sub-data determining unit is used to perform data volume restoration processing on the target interactive comparison sub-data of each dimension to obtain the target interactive comparison restoration sub-data of each dimension;

[0144] The restored data determination unit is used to fuse the target interactive test restored sub-data of each dimension to obtain interactive test restored data, and to fuse the target interactive comparison restored sub-data of each dimension to obtain interactive comparison restored data.

[0145] In some optional embodiments, the interactive test data includes interactive test sub-data corresponding to each dimension; and the target interactive test restoration sub-data determining unit includes:

[0146] A first target proportion determination subunit is used to determine a first target proportion of the target interaction test subdata of each dimension based on the interaction test subdata corresponding to each dimension and the target interaction test subdata of each dimension; the first target proportion is used to characterize the proportion of the target interaction test subdata of each dimension in the interaction test subdata corresponding to each dimension;

[0147] The target interaction test restoration sub-data determining sub-unit is used to determine the target interaction test restoration sub-data of each dimension based on the first target proportion of the target interaction test sub-data of each dimension and the target interaction test sub-data of each dimension.

[0148] In some optional embodiments, the interactive comparison data includes interactive comparison sub-data corresponding to each dimension; and the target interactive comparison restoration sub-data determination unit includes:

[0149] The second target proportion determining subunit is used to determine the second target proportion of the target interactive control subdata of each dimension based on the interactive control subdata corresponding to each dimension and the target interactive control subdata of each dimension; the second target proportion is used to represent the proportion of the target interactive control subdata of each dimension in the interactive control subdata corresponding to each dimension;

[0150] The target interaction comparison restoration sub-data determining sub-unit is used to determine the target interaction comparison restoration sub-data of each dimension based on the second target proportion of the target interaction comparison sub-data of each dimension and the target interaction comparison sub-data of each dimension.

[0151] It should be noted that the interactive data processing device embodiment provided in the embodiments of the present application and the interactive data processing method embodiment described above are based on the same inventive concept.

[0152] An embodiment of the present application also provides an electronic device for interactive data processing, which includes a processor and a memory, wherein the memory stores at least one instruction or at least one program, and the at least one instruction or at least one program is loaded and executed by the processor to implement an interactive data processing method provided in any of the above embodiments.

[0153] An embodiment of the present application also provides a computer-readable storage medium, which can be set in a terminal to store at least one instruction or at least one program for implementing an interactive data processing method in a method embodiment, and the at least one instruction or at least one program is loaded and executed by a processor to implement the interactive data processing method provided in the above method embodiment.

[0154] Optionally, in the embodiment of this specification, the storage medium may be located in at least one of the multiple network servers of the computer network. Optionally, in this embodiment, the above storage medium may include but is not limited to: a USB flash drive, a read-only memory (ROM), a random access memory (RAM), a mobile hard disk, a magnetic disk or an optical disk, and other media that can store program codes.

[0155] The memory of the embodiment of this specification can be used to store software programs and modules, and the processor executes various functional applications and data processing by running the software programs and modules stored in the memory. The memory may mainly include a program storage area and a data storage area, wherein the program storage area may store an operating system, applications required for functions, etc.; the data storage area may store data created according to the use of the device, etc. In addition, the memory may include a high-speed random access memory, and may also include a non-volatile memory, such as at least one disk storage device, a flash memory device, or other volatile solid-state storage device. Accordingly, the memory may also include a memory controller to provide the processor with access to the memory.

[0156] The embodiment of the present application also provides a computer program product or a computer program, which includes a computer instruction stored in a computer-readable storage medium. The processor of the computer device reads the computer instruction from the computer-readable storage medium, and the processor executes the computer instruction, so that the computer device executes the interactive data processing method provided by the above method embodiment.

[0157] The method embodiments provided in the embodiments of the present application can be executed in a terminal, a computer terminal, a server or a similar computing device. Taking running on a server as an example, Figure 7 FIG. 1 is a hardware structure diagram of a server of an interactive data processing method provided according to an exemplary embodiment. Figure 7As shown, the server 400 may have relatively large differences due to different configurations or performances, and may include one or more central processing units (CPU) 410 (the central processing unit 410 may include but is not limited to a processing device such as a microprocessor MCU or a programmable logic device FPGA), a memory 430 for storing data, and one or more storage media 420 (such as one or more mass storage devices) for storing application programs 423 or data 422. Among them, the memory 430 and the storage medium 420 can be short-term storage or permanent storage. The program stored in the storage medium 420 may include one or more modules, each of which may include a series of instruction operations on the server. Furthermore, the central processing unit 410 can be configured to communicate with the storage medium 420 and execute a series of instruction operations in the storage medium 420 on the server 400. The server 400 may also include one or more power supplies 460, one or more wired or wireless network interfaces 450, one or more input and output interfaces 440, and / or one or more operating systems 421, such as Windows ServerTM, Mac OS XTM, UnixTM, LinuxTM, FreeBSDTM, etc.

[0158] The input / output interface 440 may be used to receive or send data via a network. The specific example of the network may include a wireless network provided by a communication provider of the server 400. In one example, the input / output interface 440 includes a network adapter (Network Interface Controller, NIC), which may be connected to other network devices via a base station so as to communicate with the Internet. In one example, the input / output interface 440 may be a radio frequency (RF) module, which is used to communicate with the Internet wirelessly.

[0159] It can be understood by those skilled in the art that Figure 7 The structure shown is only for illustration and does not limit the structure of the above electronic device. Figure 7 More or fewer components as shown, or with Figure 7 Different configurations shown.

[0160] It should be noted that the above-mentioned sequence of the embodiments of the present application is for description only and does not represent the advantages and disadvantages of the embodiments. The above-mentioned specific embodiments of this specification are described. Other embodiments are within the scope of the attached claims. In some cases, the actions or steps recorded in the claims can be performed in an order different from that in the embodiments and still achieve the desired results. In addition, the processes depicted in the drawings do not necessarily require the specific order or continuous order shown to achieve the desired results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0161] Each embodiment in this specification is described in a progressive manner, and the same or similar parts between the embodiments can be referred to each other, and each embodiment focuses on the differences from other embodiments. In particular, for the device and server embodiments, since they are basically similar to the method embodiments, the description is relatively simple, and the relevant parts can be referred to the partial description of the method embodiments.

[0162] A person skilled in the art will understand that all or part of the steps to implement the above embodiments may be accomplished by hardware, or by a program to instruct the relevant hardware, and the program may be stored in a computer-readable storage medium, and the above-mentioned storage medium may be a read-only memory, a disk or an optical disk, etc.

[0163] The above are only preferred embodiments of the present application and are not intended to limit the present application. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present application should be included in the protection scope of the present application.

Claims

1. An interactive data processing method, characterized in that: The method comprises: Obtaining interaction data corresponding to a preset number of business requests; the interaction data is interaction data between a preset terminal object and a business processing result; the business processing result is obtained by processing a preset number of business requests based on a first business layer and a second business layer; the first business layer and the second business layer contain the same control parameters; the first business layer includes a first test business and a first comparison business; the second business layer includes a second test business and a second comparison business; the interaction data includes interaction test data and interaction comparison data; the interaction test data includes data corresponding to the first test business and the second test business, and data corresponding to the first test business and the second comparison business; the interaction comparison data includes data corresponding to the first comparison business and the second test business, and data corresponding to the first comparison business and the second comparison business; Determine, from the interaction test data, interaction data corresponding to the service request processed by both the first test service and the second test service to obtain candidate interaction test data; and determine, from the interaction comparison data, interaction data corresponding to the service request processed by both the comparison service and the second test service to obtain candidate interaction comparison data; Screening out the candidate interaction test data in the interaction test data to obtain target interaction test data, and screening out the candidate interaction comparison data in the interaction comparison data to obtain target interaction comparison data; The target interaction test data and the target interaction control data are subjected to performance evaluation processing to obtain performance evaluation data.

2. The method according to claim 1, characterized in that The method for determining the interaction data includes: receiving a preset number of business requests; Based on the first business layer and the second business layer, each of the business requests is processed respectively to obtain a business processing result corresponding to each of the business requests; Sending the business processing result corresponding to each of the business requests to the preset terminal object corresponding to each of the business requests, so as to obtain the interactive processing result between the preset terminal object corresponding to each of the business requests and the business processing result corresponding to each of the business requests; Based on the interactive processing result between the preset terminal object corresponding to each of the service requests and the service processing result corresponding to each of the service requests, the interactive data corresponding to each of the service requests is obtained.

3. The method according to claim 2, characterized in that The preset number of service requests includes a first number of first test service requests, a second number of control service requests, and a third number of second test service requests; and processing each of the service requests based on the first service layer and the second service layer to obtain a service processing result corresponding to each of the service requests includes: Processing each of the first test service requests based on the first test service to obtain a first test processing result corresponding to each of the first test service requests; the first test processing result is used to determine a first type of processing result, and the first type of processing result is used to send to a preset terminal object corresponding to the first test service request to obtain interactive test data corresponding to the first test service request; Processing each of the control service requests based on the control service to obtain a control processing result corresponding to each of the control service requests; the control processing result is used to determine a second type of processing result, and the second type of processing result is sent to a preset terminal object corresponding to the control service request to obtain interactive control data corresponding to the control service request; Process each of the second test service requests based on the second test service to obtain a second test processing result corresponding to each of the second test service requests; Based on each of the first test processing results, each of the comparison processing results and each of the second test processing results, a business processing result corresponding to each of the business requests is determined.

4. The method according to claim 3, characterized in that The third number of second test service requests includes first-category service requests and second-category service requests, wherein the first-category service requests are service requests processed by both the first test service and the second test service; the second-category service requests are service requests processed by both the control service and the second test service; and determining the service processing result corresponding to each service request based on each first test processing result, each control processing result, and each second test processing result, includes: Determine a first-category processing result corresponding to each of the first-category service requests based on a first test processing result and a second test processing result corresponding to each of the first-category service requests; Determine the second type of processing result corresponding to each of the second type of service requests based on the comparison processing result corresponding to each of the second type of service requests and the second test processing result; Each of the first-category processing results and each of the second-category processing results are determined as business processing results corresponding to each of the business requests.

5. The method according to claim 4, characterized in that The first test processing result corresponding to each of the first-type service requests includes first control data corresponding to the control parameters in the first test service; the second test processing result corresponding to each of the first-type service requests includes second control data corresponding to the control parameters in the second test service; The determining, based on the first test processing result and the second test processing result corresponding to each of the first-category service requests, the first-category processing result corresponding to each of the first-category service requests comprises: Based on the priority of each first control data and the priority of each second control data, a first-type processing result corresponding to each first-type service request is determined.

6. The method according to claim 4, characterized in that The comparison processing result corresponding to each of the second-type service requests includes the third control data corresponding to the control parameter in the comparison service; the second test processing result corresponding to each of the second-type service requests includes the fourth control data corresponding to the control parameter in the second test service; The determining, based on the comparison processing result and the second test processing result corresponding to each of the second-category service requests, the second-category processing result corresponding to each of the second-category service requests comprises: Based on the priority of each of the third control data and the priority of each of the fourth control data, a second type of processing result corresponding to each of the second type of service requests is determined.

7. The method according to any one of claims 1 to 6, characterized in that: The step of determining, from the interaction test data, interaction data corresponding to a service request processed by both the first test service and the second test service to obtain candidate interaction test data, and determining, from the interaction comparison data, interaction data corresponding to a service request processed by both the comparison service and the second test service to obtain candidate interaction comparison data, includes: Obtaining request log data corresponding to each of the business requests; the request log data includes the business logic executed by the business request and the obtained interaction data; Based on the request log data corresponding to each of the business requests, the candidate interaction test data is determined from the interaction test data, and the candidate interaction comparison data is determined from the interaction comparison data.

8. The method according to any one of claims 1 to 6, characterized in that: The performing performance evaluation processing on the target interaction test data and the target interaction control data to obtain performance evaluation data includes: Performing data volume restoration processing on the target interactive test data to obtain interactive test restoration data, and performing data volume restoration processing on the target interactive comparison data to obtain interactive comparison restoration data; The interactive test restoration data and the interactive control restoration data are subjected to performance evaluation processing to obtain performance evaluation data.

9. The method according to claim 8, characterized in that The target interaction test data includes target interaction test sub-data of at least one dimension; the target interaction comparison data includes target interaction comparison sub-data of at least one dimension; performing data volume restoration processing on the target interaction test data to obtain interaction test restoration data, and performing data volume restoration processing on the target interaction comparison data to obtain interaction comparison restoration data, include: Performing data volume restoration processing on the target interaction test sub-data of each dimension to obtain the target interaction test restored sub-data of each dimension; Performing data volume restoration processing on the target interaction comparison sub-data of each dimension to obtain the target interaction comparison restoration sub-data of each dimension; The target interaction test restoration sub-data of each dimension are integrated to obtain the interaction test restoration data; the target interaction comparison restoration sub-data of each dimension are integrated to obtain the interaction comparison restoration data.

10. The method according to claim 9, characterized in that The interactive test data includes interactive test sub-data corresponding to each dimension; The data volume restoration process is performed on the target interaction test sub-data of each dimension to obtain the target interaction test restoration sub-data of each dimension, including: Based on the interaction test sub-data corresponding to each dimension and the target interaction test sub-data of each dimension, determining a first target proportion of the target interaction test sub-data of each dimension; the first target proportion is used to characterize the proportion of the target interaction test sub-data of each dimension in the interaction test sub-data corresponding to each dimension; Based on the first target proportion of the target interaction test sub-data of each dimension and the target interaction test sub-data of each dimension, the target interaction test restoration sub-data of each dimension is determined.

11. The method according to claim 9 or 10, characterized in that: The interactive comparison data includes interactive comparison sub-data corresponding to each dimension; the target interactive comparison sub-data of each dimension is subjected to data volume reduction processing to obtain the target interactive comparison reduction sub-data of each dimension, including: Based on the interactive control sub-data corresponding to each dimension and the target interactive control sub-data of each dimension, determining a second target proportion of the target interactive control sub-data of each dimension; the second target proportion is used to represent the proportion of the target interactive control sub-data of each dimension in the interactive control sub-data corresponding to each dimension; Based on the second target proportion of the target interaction comparison sub-data of each dimension and the target interaction comparison sub-data of each dimension, the target interaction comparison restoration sub-data of each dimension is determined.

12. An interactive data processing device, characterized in that: The device comprises: An acquisition module, used to acquire interaction data corresponding to a preset number of business requests; the interaction data is interaction data between a preset terminal object and a business processing result; the business processing result is obtained by processing a preset number of business requests based on a first business layer and a second business layer; the first business layer and the second business layer contain the same control parameters; the first business layer includes a first test business and a first comparison business; the second business layer includes a second test business and a second comparison business; the interaction data includes interaction test data and interaction comparison data; the interaction test data includes data corresponding to the first test business and the second test business, and data corresponding to the first test business and the second comparison business; the interaction comparison data includes data corresponding to the first comparison business and the second test business, and data corresponding to the first comparison business and the second comparison business; a candidate interaction data determination module, configured to determine, from the interaction test data, interaction data corresponding to a service request processed by both the first test service and the second test service, to obtain candidate interaction test data, and to determine, from the interaction comparison data, interaction data corresponding to a service request processed by both the comparison service and the second test service, to obtain candidate interaction comparison data; a screening module, configured to screen out the candidate interaction test data in the interaction test data to obtain target interaction test data, and to screen out the candidate interaction comparison data in the interaction comparison data to obtain target interaction comparison data; The performance evaluation data determination module is used to perform performance evaluation processing on the target interaction test data and the target interaction control data to obtain performance evaluation data.

13. An electronic device for interactive data processing, characterized in that: The device includes a processor and a memory, wherein the memory stores at least one instruction or at least one program, and the at least one instruction or the at least one program is loaded by the processor and executed by the interactive data processing method according to any one of claims 1-11.

14. A computer-readable storage medium, characterized in that: The storage medium stores at least one instruction or at least one program, and the at least one instruction or at least one program is loaded and executed by the processor to implement the interactive data processing method according to any one of claims 1-11.

15. A computer program product comprising a computer program, characterized in that When the computer program is executed by a processor, the interactive data processing method according to any one of claims 1 to 11 is implemented.

Images