Data testing method and device based on time slice rotation

Abstract

The embodiment of the invention provides a data testing method and device based on time slice rotation. The data testing method based on the round-robin comprises the steps of firstly determining initial testing parameters of a to-be-tested object based on historical feature data of the to-be-tested object, then determining a round-robin rule of the to-be-tested object based on the initial testing parameters and an attribute adjusting strategy of the to-be-tested object, and finally testing the to-be-tested object according to the round-robin rule based on the round-robin rule. Acquiring test feature data corresponding to the to-be-tested object, finally, based on the test feature data and historical feature data, generating a test result for evaluating the attribute adjustment strategy, calculating initial test parameters by using the historical feature data, and referring to data attributes of the historical feature data. Therefore, the test feature data better fits the test scene of the to-be-tested object, the data test efficiency and test accuracy are improved, the method can be suitable for various different application scenes, and the accuracy and applicability of data test are improved.

Description

Technical Field

[0001] Embodiments of this disclosure relate to the fields of computer technology and internet technology, and more particularly to data testing and apparatus based on time-slice rotation. Background Technology

[0002] In the current smart supply chain scenario, the internet environment is very complex, and there are very few samples available for evaluating algorithm strategies. Currently, there are two main methods commonly used in data testing: time-slice rotation and random flow distribution.

[0003] The time-slice rotation method currently mainly uses simple randomization to specify whether a sample should use the algorithm strategy, or directly applies the algorithm strategy at intervals. The randomized distribution method is used in traditional internet fields such as search, advertising, and recommendation, where the sample size is large and the partitioning is flexible. However, both methods obtain data that is not accurate enough, resulting in inaccurate test results. Summary of the Invention

[0004] Embodiments of this disclosure provide a time-slice-based data testing method, a time-slice-based data testing apparatus, an electronic device, a computer-readable medium, and a computer program product.

[0005] In a first aspect, embodiments of this disclosure provide a time-slice rotation-based data testing method, which includes: determining initial test parameters for the test object based on historical feature data of the test object; determining time-slice rotation rules for the test object based on the initial test parameters and the attribute adjustment strategy of the test object; obtaining test feature data corresponding to the test object based on the time-slice rotation rules; and generating test results for evaluating the attribute adjustment strategy based on the test feature data and historical feature data.

[0006] In some embodiments, the initial test parameters include initial period parameters and initial residual duration; and, based on the historical feature data of the test object, the initial test parameters of the test object are determined, including: performing periodic calculations based on the historical feature data of the test object to generate the initial period parameters corresponding to the test object; and determining the initial residual duration corresponding to the test object based on the test type and historical feature data corresponding to the test object.

[0007] In some embodiments, the initial test parameters further include a target randomization method and a target data quantity; and the method further includes: determining a target randomization strategy corresponding to the test object from multiple candidate randomization strategies based on the test type corresponding to the test object; and calculating the target data quantity corresponding to the test object based on the test type corresponding to the test object.

[0008] In some embodiments, determining the initial residual duration corresponding to the test object based on the test type and historical feature data of the test object includes: in response to determining that the test type is a first preset type, determining whether the historical feature data meets the data fitting conditions; in response to determining that the historical feature data meets the data fitting conditions, determining the first label mean and the first fitting curve corresponding to the test object based on the historical feature data; and determining the initial residual duration corresponding to the test object based on the first fitting curve and the first label mean.

[0009] In some embodiments, determining the initial residual duration corresponding to the test object based on the test type and historical feature data of the test object further includes: in response to determining that the historical feature data does not meet the data fitting conditions, obtaining similar feature data corresponding to the historical feature data; updating the historical feature data based on the similar feature data to obtain updated historical feature data; determining the second label mean and the second fitting curve corresponding to the test object based on the updated historical feature data; and determining the initial residual duration corresponding to the test object based on the second fitting curve and the second label mean.

[0010] In some embodiments, the method further includes: in response to determining that the test type is a second preset type, selecting an initial residual duration corresponding to the test object from a plurality of preset residual durations.

[0011] In some embodiments, the test feature data includes first test feature data and second test feature data; and, based on the time slice rotation rule, obtaining the test feature data corresponding to the object under test includes: obtaining the first test feature data corresponding to the object under test based on the time slice rotation rule; determining new test parameters for the object under test based on the first test feature data and historical feature data; updating the time slice rotation rule based on the new test parameters to obtain a new time slice rotation rule; and obtaining the second test feature data corresponding to the object under test based on the new time slice rotation rule.

[0012] In some embodiments, the new test parameters include new period parameters; and, based on the first test feature data and historical feature data, determining the new test parameters of the object to be tested includes: determining whether the test period corresponding to the first test feature data meets the period update condition; in response to determining that the test period corresponding to the first test feature data meets the period update condition, performing periodic calculations based on the first test feature data and historical feature data to generate new period parameters of the object to be tested.

[0013] In some embodiments, the new test parameters include a new residual duration; and, based on the first test feature data and historical feature data, determining the new test parameters for the test object includes: determining the third label mean and the third fitting curve corresponding to the test object based on the first test feature data and historical feature data; and determining the new residual duration of the test object based on the third label mean and the third fitting curve.

[0014] In some embodiments, based on test feature data and historical feature data, test results for evaluating attribute adjustment strategies are generated, including: determining whether the number of data corresponding to the test feature data and historical feature data meets a quantity condition; in response to determining that the number of data corresponding to the test feature data and historical feature data meets the quantity condition, determining an experimental sample set and a control sample set from the test feature data and historical feature data; performing a significance test on the experimental sample set and the control sample set to generate test results for evaluating attribute adjustment strategies.

[0015] In some embodiments, the historical characteristic data of the test object satisfies at least one of the following: For warehouse storage scenarios, the historical characteristic data of the test object includes storage data of the warehouse to be stored under adjusted warehouse attributes and storage data under unadjusted warehouse attributes; for item delivery scenarios, the historical characteristic data of the test object includes delivery data of the items to be delivered under adjusted delivery attributes and delivery data under unadjusted delivery attributes; for item sales scenarios, the historical characteristic data of the test object includes sales data of the items to be sold under adjusted item attribute information and sales data under unadjusted item attribute information.

[0016] Secondly, embodiments of this disclosure provide a time-slice rotation-based data testing apparatus, comprising: a parameter determination module configured to determine initial test parameters of the test object based on historical feature data of the test object; a rule determination module configured to determine time-slice rotation rules of the test object based on the initial test parameters and the attribute adjustment strategy of the test object; an acquisition module configured to acquire test feature data corresponding to the test object based on the time-slice rotation rules; and a generation module configured to generate test results for evaluating the attribute adjustment strategy based on the test feature data and historical feature data.

[0017] In some embodiments, the initial test parameters include initial period parameters and initial residual duration; and the parameter determination module is further configured to: perform periodic calculations based on historical feature data of the test object to generate initial period parameters corresponding to the test object; and determine the initial residual duration corresponding to the test object based on the test type and historical feature data corresponding to the test object.

[0018] In some embodiments, the initial test parameters further include a target randomization method and a target data quantity; and the parameter determination module is further configured to: determine the target randomization strategy corresponding to the test object from multiple candidate randomization strategies based on the test type corresponding to the test object; and calculate the target data quantity corresponding to the test object based on the test type corresponding to the test object.

[0019] In some embodiments, the parameter determination module is further configured to: in response to determining that the test type is a first preset type, determine whether the historical feature data meets the data fitting conditions; in response to determining that the historical feature data meets the data fitting conditions, determine the first label mean and the first fitting curve corresponding to the test object based on the historical feature data; and determine the initial residual duration corresponding to the test object based on the first fitting curve and the first label mean.

[0020] In some embodiments, the parameter determination module is further configured to: in response to determining that the historical feature data does not meet the data fitting conditions, obtain similar feature data corresponding to the historical feature data; update the historical feature data based on the similar feature data to obtain updated historical feature data; determine the second label mean and the second fitting curve corresponding to the test object based on the updated historical feature data; and determine the initial residual duration corresponding to the test object based on the second fitting curve and the second label mean.

[0021] In some embodiments, the parameter determination module is further configured to: in response to determining that the test type is a second preset type, select the initial residual duration corresponding to the test object from a plurality of preset residual durations.

[0022] In some embodiments, the test feature data includes first test feature data and second test feature data; and the acquisition module is further configured to: acquire the first test feature data corresponding to the object under test based on the time slice rotation rule; determine new test parameters for the object under test based on the first test feature data and historical feature data; update the time slice rotation rule based on the new test parameters to obtain a new time slice rotation rule; and acquire the second test feature data corresponding to the object under test based on the new time slice rotation rule.

[0023] In some embodiments, the new test parameters include new period parameters; and the acquisition module is further configured to: determine whether the test period corresponding to the first test feature data meets the period update condition; in response to determining that the test period corresponding to the first test feature data meets the period update condition, perform periodic calculations based on the first test feature data and historical feature data to generate new period parameters for the object to be tested.

[0024] In some embodiments, the new test parameters include a new residual duration; and the acquisition module is further configured to: determine the third label mean and the third fitting curve corresponding to the test object based on the first test feature data and historical feature data; and determine the new residual duration of the test object based on the third label mean and the third fitting curve.

[0025] In some embodiments, the generation module is further configured to: determine whether the number of data corresponding to the test feature data and the historical feature data meets the quantity condition; in response to determining that the number of data corresponding to the test feature data and the historical feature data meets the quantity condition, determine the experimental sample set and the control sample set from the test feature data and the historical feature data; perform a significance test on the experimental sample set and the control sample set, and generate test results for evaluating the attribute adjustment strategy.

[0026] In some embodiments, the historical characteristic data of the test object satisfies at least one of the following: For warehouse storage scenarios, the historical characteristic data of the test object includes storage data of the warehouse to be stored under adjusted warehouse attributes and storage data under unadjusted warehouse attributes; for item delivery scenarios, the historical characteristic data of the test object includes delivery data of the items to be delivered under adjusted delivery attributes and delivery data under unadjusted delivery attributes; for item sales scenarios, the historical characteristic data of the test object includes sales data of the items to be sold under adjusted item attribute information and sales data under unadjusted item attribute information.

[0027] Thirdly, embodiments of this disclosure provide an electronic device comprising: one or more processors; a storage device having one or more programs stored thereon; and, when the one or more programs are executed by the one or more processors, causing the one or more processors to implement a time-slice-based data testing method as described in any embodiment of the first aspect.

[0028] Fourthly, embodiments of this disclosure provide a computer-readable medium having a computer program stored thereon that, when executed by a processor, implements the time-slice-based data testing method as described in any embodiment of the first aspect.

[0029] Fifthly, embodiments of this disclosure provide a computer program product, including a computer program that, when executed by a processor, implements a time-slice-based data testing method as described in any embodiment of the first aspect.

[0030] The time-slice rotation-based data testing method provided in this disclosure first determines the initial test parameters of the test object based on its historical feature data. Then, based on the initial test parameters and the attribute adjustment strategy of the test object, it determines the time-slice rotation rules of the test object. Subsequently, based on the time-slice rotation rules, it obtains the corresponding test feature data of the test object. Finally, based on the test feature data and historical feature data, it generates test results for evaluating the attribute adjustment strategy. This method can predetermine the initial test parameters using the historical feature data of the test object, determine the time-slice rotation rules using the initial test parameters and attribute adjustment strategy, and obtain the test feature data. Compared to existing methods for obtaining test feature data, this method does not require traversing and simulating the entire application scenario of the test object. It can obtain test feature data using the determined time-slice rotation rules, saving computational resources and improving data testing efficiency. Furthermore, by using historical feature data to calculate the initial test parameters and referencing the data attributes of the historical feature data, the test feature data is more closely aligned with the test scenario of the test object, improving data testing efficiency and accuracy. This method is applicable to various application scenarios, enhancing the accuracy and applicability of data testing. Attached Figure Description

[0031] Other features, objects, and advantages of this disclosure will become more apparent from the following detailed description of non-limiting embodiments with reference to the accompanying drawings:

[0032] Figure 1 This is an exemplary system architecture diagram to which one embodiment of this disclosure can be applied;

[0033] Figure 2 This is a flowchart of an embodiment of the time-slice-based data testing method according to the present disclosure;

[0034] Figure 3 This is a flowchart of one embodiment of determining initial test parameters according to this disclosure;

[0035] Figure 4 This is a flowchart of one embodiment of determining the initial residual duration according to the present disclosure;

[0036] Figure 5 This is a flowchart of one embodiment of obtaining test feature data according to the present disclosure;

[0037] Figure 6 This is a flowchart of one embodiment of determining new period parameters according to this disclosure;

[0038] Figure 7 This is a flowchart of one embodiment of determining a new residual duration according to the present disclosure;

[0039] Figure 8 This is a schematic diagram of a structure of an embodiment of the time-slice rotation-based data testing apparatus according to the present disclosure;

[0040] Figure 9 This is a schematic diagram of the structure of an electronic device suitable for implementing embodiments of the present disclosure. Detailed Implementation

[0041] The present disclosure will now be described in further detail with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative of the relevant disclosure and not intended to limit it. Furthermore, it should be noted that, for ease of description, only the parts relevant to the disclosure are shown in the accompanying drawings.

[0042] It should be noted that, unless otherwise specified, the embodiments and features described in this disclosure can be combined with each other. This disclosure will now be described in detail with reference to the accompanying drawings and embodiments.

[0043] Figure 1 An exemplary system architecture 100 is shown for a time-slice-based data testing method and a time-slice-based data testing apparatus to which embodiments of the present disclosure can be applied.

[0044] like Figure 1 As shown, system architecture 100 may include terminal devices 104, 105, and 106, network 107, and servers 101, 102, and 103. Network 107 serves as the medium for providing communication links between terminal devices 104, 105, and 106 and servers 101, 102, and 103. Network 107 may include various connection types, such as wired or wireless communication links or fiber optic cables, etc.

[0045] Users can interact with servers 101, 102, and 103 belonging to the same server cluster via network 107 through terminal devices 104, 105, and 106 to receive or send information. Various applications can be installed on terminal devices 104, 105, and 106, such as data testing tools, instant messaging tools, social media platforms, search applications, shopping applications, and browser applications.

[0046] Terminal devices 104, 105, and 106 can be either hardware or software. When the terminal device is hardware, it can be any electronic device with a display screen that supports communication with the server, including but not limited to smartphones, tablets, laptops, and desktop computers. When the terminal device is software, it can be installed in the electronic devices listed above. It can be implemented as multiple software programs or software modules, or as a single software program or software module. No specific limitations are made here.

[0047] Terminal devices 104, 105, and 106 can pre-acquire historical feature data of the object under test and determine the initial test parameters of the object under test based on the historical feature data. Then, terminal devices 104, 105, and 106 can determine the time slice rotation rule of the object under test based on the initial test parameters and the attribute adjustment strategy of the object under test. After that, terminal devices 104, 105, and 106 execute the test data acquisition process according to the time slice rotation rule to obtain the test feature data corresponding to the object under test. Finally, terminal devices 104, 105, and 106 can generate test results for evaluating the attribute adjustment strategy based on the test feature data and historical feature data. The test results can characterize the usability of the attribute adjustment strategy.

[0048] Servers 101, 102, and 103 can be servers that provide various services, such as backend servers that receive requests sent by terminal devices with which they have established communication connections. The backend servers can receive and analyze the requests sent by the terminal devices and generate processing results.

[0049] It should be noted that a server can be either hardware or software. When a server is hardware, it can be any electronic device that provides various services to terminal devices. When a server is software, it can be implemented as multiple software programs or software modules providing various services to terminal devices, or it can be implemented as a single software program or software module providing various services to terminal devices. No specific limitations are made here.

[0050] Servers 101, 102, and 103 can pre-acquire historical feature data of the object under test and determine the initial test parameters based on this data. Then, servers 101, 102, and 103 can determine the time-slice rotation rules of the object under test based on the initial test parameters and the attribute adjustment strategy. After that, servers 101, 102, and 103 execute the test data acquisition process according to the time-slice rotation rules to obtain the test feature data corresponding to the object under test. Finally, servers 101, 102, and 103 can generate test results for evaluating the attribute adjustment strategy based on the test feature data and historical feature data. These test results can characterize the usability of the attribute adjustment strategy.

[0051] It should be noted that the time-slice-based data testing method provided in the embodiments of this disclosure can be executed by terminal devices 104, 105, 106 or servers 101, 102, 103. Accordingly, the time-slice-based data testing device can be set in terminal devices 104, 105, 106 or servers 101, 102, 103.

[0052] It should be understood that Figure 1The number of terminal devices, networks, and servers shown is merely illustrative. Depending on implementation needs, any number of terminal devices, networks, and servers can be included.

[0053] Figure 2 A flowchart 200 illustrates an embodiment of a time-slice-based data testing method according to the present disclosure. This time-slice-based data testing method includes the following steps:

[0054] Step 210: Determine the initial test parameters of the object under test based on its historical feature data.

[0055] Before implementing a time-slice-based round-robin data testing method, the user first needs to specify the requirements for the object to be tested. These requirements include algorithm requirements, test scenarios for algorithm and product evaluation, and technical thresholds. If the requirements are not feasible, development is abandoned. If the requirements are feasible, detailed methodologies are discussed further. Next, the time-slice algorithm is designed, developed, and locally validated to determine its feasibility. If feasible, the user is then asked to provide historical feature data and other information for the object to be tested. This historical feature data can then be input into the execution entity running the time-slice-based round-robin data testing method (e.g., ...). Figure 1 Terminal devices 104, 105, 106 or servers 101, 102, 103)

[0056] In this step, the aforementioned execution entity can receive historical feature data of the test object. This historical feature data can be the historical data of the test object under the attribute adjustment strategy within a historical time period. This historical feature data may include the test object's sample number, historical test date, historical label value, and feature information from multiple dimensions. Then, the execution entity can calculate test parameters based on the historical feature data of the test object to obtain the initial test parameters. These initial test parameters can be used to characterize the test parameters required for a controlled experiment on the test object based on the time-slice rotation rule. These initial test parameters may include the initial period parameters, initial residual duration, target randomization strategy, and target data quantity required by the time-slice rotation rule.

[0057] Specifically, the aforementioned execution entity can perform data analysis on historical feature data, analyze the periodicity of the historical feature data, obtain the initial period parameters corresponding to the historical feature data, and analyze the initial residual duration of the historical feature data that is affected by the strategy during time slice rotation. Then, based on the requirements provided by the demand side user, it can determine the target random strategy and the target data quantity required for the test object. Thus, the aforementioned execution entity can obtain initial test parameters including initial period parameters, initial residual duration, target random strategy, and target data quantity.

[0058] The historical characteristic data of the test objects shall meet at least one of the following conditions: For warehouse storage scenarios, the historical characteristic data of the test objects shall include storage data of the warehouse under adjusted warehouse attributes and storage data under unadjusted warehouse attributes; for item delivery scenarios, the historical characteristic data of the test objects shall include delivery data of the items to be delivered under adjusted delivery attributes and delivery data under unadjusted delivery attributes; for item sales scenarios, the historical characteristic data of the test objects shall include sales data of the items to be sold under adjusted item attribute information and sales data under unadjusted item attribute information.

[0059] Specifically, the aforementioned implementing entity can, for warehouse storage scenarios, collect storage data of the warehouse to be stored under adjusted warehouse attributes and without adjusted warehouse attributes through random testing. These warehouse attributes can include warehouse storage type, warehouse qualifications, etc. Similarly, for item delivery scenarios, the aforementioned implementing entity can collect delivery data of items to be delivered under adjusted delivery attributes and without adjusted delivery attributes through random testing. These delivery attributes can include delivery route, delivery vehicle, delivery time, etc. Furthermore, for item sales scenarios, the aforementioned implementing entity can collect sales data of items to be sold under adjusted item attribute information and without adjusted item attribute information through random testing. These item attributes can include item appearance, item quantity, etc. Therefore, the aforementioned test objects can include warehouses to be stored, items to be delivered, and items to be sold. Their historical characteristic data can include storage data of warehouses to be stored under adjusted warehouse attributes and without adjusted warehouse attributes, delivery data of items to be delivered under adjusted delivery attributes and without adjusted delivery attributes, and sales data of items to be sold under adjusted item attribute information and without adjusted item attribute information, etc.

[0060] The aforementioned execution entities can be tested across a wide range of application scenarios, making the methods more applicable to real-world business scenarios and improving testing accuracy and applicability across multiple scenarios.

[0061] Step 220: Determine the time slice rotation rules for the object under test based on the initial test parameters and the attribute adjustment strategy of the object under test.

[0062] In this step, after obtaining the initial test parameters for the object to be tested, the executing entity can determine the attribute adjustment strategy for the object to be tested based on the requirements provided by the user. The same object to be tested can correspond to different attribute adjustment strategies or the same attribute adjustment strategy. As an example, if the object to be tested is an item to be sold, the attribute adjustment strategy could be to adjust the appearance of the item to be sold.

[0063] Then, the aforementioned execution entity can determine the time slice rotation rule of the test object based on the initial test parameters and the attribute adjustment strategy corresponding to the test object. The time slice rotation rule is to conduct comparative testing according to the initial test parameters and attribute adjustment strategy based on the time slice. The time slice rotation rule can be to divide the time into multiple time periods, with each time period as a time slice, and conduct comparative testing on the data in each time slice according to the initial test parameters and attribute adjustment strategy.

[0064] Step 230: Based on the time slice rotation rule, obtain the test feature data corresponding to the object to be tested.

[0065] In this step, after the aforementioned execution entity determines the time slice rotation rule, it can conduct a comparative test on the test object according to the time slice rotation rule. This comparative test is a comparative test on the test object according to the attribute adjustment strategy. During the test, the test object can be compared according to the set time slice and initial test parameters, and the attribute adjustment strategy is used to generate test feature data corresponding to the test object. The data format of the test feature data is the same as the data format of the historical feature data. The test feature data may include the sample number of the test object, the test date, the test label value, and feature information of multiple different dimensions.

[0066] Step 240: Based on the test feature data and historical feature data, generate test results for evaluating the attribute adjustment strategy.

[0067] In this step, after obtaining the test feature data, the executing entity can determine the experimental sample set and control sample set corresponding to the test object from the test feature data and historical feature data. It then performs significance calculations on the experimental and control sample sets to test whether there is a significant difference between their label values. The testing principle is to use a significance test, which requires proving a significant difference between the two groups. The significance test principle is: null hypothesis H0: there is no difference between the experimental and control sample sets, i.e., μ1 = μ2; alternative hypothesis H1: there is a difference between the experimental and control sample sets, i.e., μ1 > μ2 or μ1 < μ2; the criterion for passing the test is that when p < α, a significant difference is determined between the experimental and control sample sets, and the significance test is considered passed.

[0068] The aforementioned implementing entity can then generate test results for evaluating the attribute adjustment strategy based on the significance calculation results. These test results characterize the evaluation of the attribute adjustment strategy and may include conclusions indicating whether the strategy has advantages or disadvantages. The implementing entity can then present the test results to the requesting user, allowing them to promptly understand the data performance of the tested object under the attribute adjustment strategy.

[0069] As one implementation of this disclosure, step 240 above, which generates test results for evaluating attribute adjustment strategies based on test feature data and historical feature data, may include the following steps:

[0070] The first step is to determine whether the number of data corresponding to the test feature data and the historical feature data meets the quantity requirements.

[0071] Specifically, after obtaining the test feature data, the aforementioned execution entity can obtain the data quantity of the test feature data and historical feature data, and determine whether the data quantity corresponding to the test feature data and historical feature data meets the quantity condition. This quantity condition may include the data quantity not being less than the target data quantity corresponding to the test object. That is, the aforementioned execution entity can compare the data quantity with the target data quantity corresponding to the test object to determine whether the data quantity is greater than or equal to the target data quantity.

[0072] The second step involves determining the experimental sample set and the control sample set from the test feature data and the historical feature data, in response to the determination that the number of data corresponding to the test feature data and the historical feature data meets the quantity condition.

[0073] Specifically, if the aforementioned executing entity determines that the number of data corresponding to the test feature data and the historical feature data meets the quantity condition, it can then determine the experimental sample set and the control sample set from the test feature data and the historical feature data based on whether or not the data attribute adjustment strategy is applied.

[0074] The third step is to perform significance tests on the experimental sample set and the control sample set to generate test results for evaluating the attribute adjustment strategy.

[0075] Specifically, after determining the experimental sample set and control sample set from the test feature data and historical feature data, the aforementioned implementing entity can obtain the data types of the test feature data and historical feature data, and perform significance testing on the experimental sample set and control sample set according to the significance verification method corresponding to the data type, generating test results for evaluating the attribute adjustment strategy.

[0076] If the data type is a continuous variable, and there are only two groups, the experimental sample set and the control sample set, the aforementioned execution entity can perform the Shapiro-Wilk normality test and the Levene homogeneity of variance test on the experimental sample set and the control sample set. Based on the number of data points and the results of the Shapiro-Wilk normality test and the Levene homogeneity of variance test, the entity can choose the t-test for homoscedasticity or heteroscedasticity, or the Mann-Whitney U test for two independent samples. The passing criterion for this test is that if the p-value is less than the preset α value, then the difference between the two groups, the experimental sample set and the control sample set, is significant, and the test passes the significance test.

[0077] If the data type is a discrete variable, the above-mentioned execution entity can choose the ratio Z-test or Fisher's exact test based on the amount of data in the experimental sample set and the control sample set. The criterion for passing this test is that if the p-value is less than the preset α value, then the difference between the experimental sample set and the control sample set is considered significant, and the test passes the significance test.

[0078] In this implementation, the number of data corresponding to the test feature data and historical feature data is determined to meet the quantity condition. On the basis of meeting the quantity condition, the significance test of the experimental sample set and the control sample set is further performed using the significance test method corresponding to the data type. Significance test is only performed when data that meets the quantity condition is obtained, which makes the significance test more accurate and more targeted. Appropriate test methods can be used for significance test, which improves the targeting and accuracy of significance test.

[0079] The time-slice rotation-based data testing method provided in this disclosure first determines the initial test parameters of the test object based on its historical feature data. Then, based on the initial test parameters and the attribute adjustment strategy of the test object, it determines the time-slice rotation rules of the test object. Subsequently, based on the time-slice rotation rules, it obtains the corresponding test feature data of the test object. Finally, based on the test feature data and historical feature data, it generates test results for evaluating the attribute adjustment strategy. This method can predetermine the initial test parameters using the historical feature data of the test object, determine the time-slice rotation rules using the initial test parameters and attribute adjustment strategy, and obtain the test feature data. Compared to existing methods for obtaining test feature data, this method does not require traversing and simulating the entire application scenario of the test object. It can obtain test feature data using the determined time-slice rotation rules, saving computational resources and improving data testing efficiency. Furthermore, by using historical feature data to calculate the initial test parameters and referencing the data attributes of the historical feature data, the test feature data is more closely aligned with the test scenario of the test object, improving data testing efficiency and accuracy. This method is applicable to various application scenarios, enhancing the accuracy and applicability of data testing.

[0080] As an application example of a time-slice rotation-based data testing method, the terminal device can pre-acquire historical sales data for the items to be sold, as shown in the table below:

[0081] Sample number Historical test dates Historical tag value S0001 2020-01-01 20 S0001 2020-02-01 30 S0001 2020-03-01 40 S0001 2020-04-01 50

[0082] The terminal device can perform data analysis on historical sales data, analyze the periodicity of historical sales data, obtain the initial period parameters corresponding to historical sales data, and analyze the initial residual duration of historical sales data affected by strategies during time slice rotation. Then, based on the demand content provided by the demand side users, it can determine the target random strategy and target data quantity required for the items to be sold, and obtain initial test parameters including initial period parameters, initial residual duration, target random strategy, and target data quantity.

[0083] Afterwards, the terminal device can determine the time slice rotation rule of the items to be sold based on the initial test parameters and the appearance adjustment strategy corresponding to the items to be sold, and conduct comparative tests on the items to be sold according to the time slice rotation rule. The items to be sold are compared and tested according to the appearance adjustment strategy. During the test, the items to be sold can be compared and tested according to the set time slice and initial test parameters, and the appearance adjustment strategy is used to generate test sales data corresponding to the items to be sold.

[0084] Finally, the terminal device can determine the experimental sample set and control sample set corresponding to the items to be sold from the test sales data and historical sales data, and perform significance calculation on the experimental sample set and control sample set to test whether there is a significant difference between the label values ​​of the experimental sample set and control sample set. Based on the significance calculation results, test results are generated to evaluate the appearance adjustment strategy. The test results can characterize the feasibility of the appearance adjustment strategy.

[0085] refer to Figure 3 , Figure 3 A flowchart 300 illustrating an embodiment of determining initial test parameters, namely step 210 above, shows the determination of initial test parameters for the object under test based on historical feature data of the object under test, which may include the following steps:

[0086] The initial test parameters mentioned above may include an initial period parameter and an initial residual duration. The initial period parameter can characterize the periodicity of the time series data corresponding to the test object, and the initial residual duration can characterize the duration of the influence of the previous attribute adjustment strategy when performing time slice rotation.

[0087] Step 310: Perform periodic calculations based on the historical feature data of the object under test to generate the initial periodic parameters corresponding to the object under test.

[0088] In this step, the aforementioned execution entity can use the periodogram method to perform periodic calculations on the historical feature data of the test object. That is, it can directly perform Fourier transform on the historical feature data, then take the square of the amplitude-frequency feature and divide it by the data length of the historical feature data to obtain the initial period parameters corresponding to the test object.

[0089] Step 320: Determine the initial residual duration of the test object based on the test type and historical feature data corresponding to the test object.

[0090] In this step, the executing entity can obtain the test type of the object to be tested. This test type can include the data type of the historical feature data of the object to be tested, which can include continuous and discrete variables. The test type can also include the application type of the object to be tested, which can be pre-defined based on the object type of the object to be tested. Each application type can correspond to multiple different types of objects to be tested. Application types can include those with a greater impact from attribute adjustments and those with a smaller impact from attribute adjustments, determined based on the object type. Different test types can correspond to different residual duration calculation methods; that is, different data types can correspond to different residual duration calculation methods, or different application types can correspond to different residual duration calculation methods. Therefore, the executing entity can determine the residual duration calculation method corresponding to the object to be tested based on the test type, and use the determined residual duration calculation method to calculate the initial residual duration corresponding to the object to be tested.

[0091] In this implementation, the initial cycle parameters and initial residual duration are calculated using historical feature data. Before testing, the feature parameters corresponding to the test object can be analyzed in advance based on the historical feature data. By referencing the data attributes of the historical feature data, the test feature data obtained based on the initial test parameters is more in line with the test scenario of the test object, thereby improving the efficiency and accuracy of data testing.

[0092] Continue to refer to Figure 3 The initial test parameters mentioned above may also include the target randomization method and the target data quantity, thus the method may further include the following steps:

[0093] Step 330: Based on the test type corresponding to the object to be tested, determine the target random strategy corresponding to the object to be tested from multiple candidate random strategies.

[0094] In this step, the aforementioned execution entity can determine the test type corresponding to the object to be tested. Different test types can correspond to different candidate random strategies. If the test type corresponds to a data type, different data types can correspond to different candidate random strategies. That is, continuous variables can correspond to one candidate random strategy, and discrete variables can also correspond to one candidate random strategy. If the test type corresponds to an application type, different application types can correspond to different candidate random strategies. That is, application types with a greater impact from attribute adjustments can correspond to one candidate random strategy, and application types with a smaller impact from attribute adjustments can correspond to one candidate random strategy.

[0095] Multiple candidate random strategies can include single-target random strategies and overall random strategies. Operators can pre-set the correspondence between test types and candidate random strategies based on experience. After determining the test type of the object to be tested, the execution entity can select the target random strategy corresponding to the test type from multiple candidate random strategies according to the above correspondence.

[0096] Step 340: Calculate the number of target data corresponding to the test object based on the test type corresponding to the test object.

[0097] In this step, the aforementioned execution entity can determine the test type corresponding to the object to be tested, and then use the data quantity calculation method corresponding to the test type to calculate the target data quantity corresponding to the object to be tested.

[0098] If the test type corresponds to a data type, the target data quantity can be calculated using the following formula for continuous variables:

[0099]

[0100] Where N is the number of target data points. yes The corresponding Z-score, α is usually taken as 5%, Z 1-β X is the Z-score corresponding to 1-β, where β is usually taken as 20%, Δ is the label mean of the experimental sample data minus the label mean of the control sample data, n1 is the experimental sample data in the historical feature data, n2 is the control sample data in the historical feature data, and X... 1i X 2i These are the values ​​of individual data points in each group.

[0101] If the test type corresponds to a data type, then the target data quantity can be calculated using the following formula for discrete variables:

[0102]

[0103] Δ=p2-p1

[0104] σ pooled 2 =p1(1-p1)+p2(1-p2)

[0105] Where N is the number of target data points. yes The corresponding Z-score, α is usually taken as 5%, Z 1-β 1-β is the Z score, where β is usually taken as 20%. p2 and p1 are the conversion rates of experimental sample data and control sample data in historical feature data, respectively.

[0106] The aforementioned target data quantity can be obtained through calculation methods for estimating the minimum sample size in the prior art, and this disclosure does not impose specific limitations on it.

[0107] In this implementation, by determining the target random strategy and target data quantity of the test object based on the test type, the random strategy and data quantity can be made more targeted, and the target random strategy and target data quantity can be obtained in a suitable way.

[0108] refer to Figure 4 , Figure 4 A flowchart 400 illustrating an embodiment for determining the initial residual duration, namely step 320 above, shows the determination of the initial residual duration corresponding to the test object based on the test type and historical feature data of the test object. This may include the following steps:

[0109] Step 410: In response to determining that the test type is the first preset type, determine whether the historical feature data meets the data fitting conditions.

[0110] In this step, if the executing entity determines that the test type is the first preset type, it can judge the data content in the historical feature data to determine whether there are historical attribute adjustment strategies and date annotations, in order to determine whether the historical feature data meets the data fitting conditions. The data fitting conditions may include historical attribute adjustment strategies and date annotations in the historical feature data. Furthermore, the first preset type may be a data type that is a continuous variable or a discrete variable, or an application type that has a significant impact from attribute adjustments or an application type that has a relatively small impact from attribute adjustments. This first preset type can be a type pre-set by the operator based on experience.

[0111] Step 420: In response to determining that the historical feature data meets the data fitting conditions, the mean of the first label and the first fitting curve corresponding to the test object are determined based on the historical feature data.

[0112] In this step, the aforementioned executing entity determines that the historical feature data meets the data fitting conditions. Then, it can adjust the data alignment strategy according to the historical attributes in the historical feature data to obtain aligned historical feature data. Following the time progression rule, the aligned historical feature data is fitted with time as the X-axis and historical label value as the Y-axis. That is, the aligned historical feature data is fitted with a gamma distribution to obtain the first fitting curve. This first fitting curve can be a curve obtained by fitting data with time as the X-axis and historical label value as the Y-axis.

[0113] Then, the aforementioned execution entity obtains each historical label value from the historical feature data, sums up multiple historical label values, and divides the resulting label sum by the total number of historical feature data to obtain the first label mean.

[0114] As an example, if the historical feature data is shown in the table below:

[0115] Sample number Historical test dates Historical tag value S0001 2020-01-01 20 S0001 2020-02-01 30 S0001 2020-03-01 40 S0001 2020-04-01 30

[0116] The aforementioned executing entity can first sum the historical label values ​​20, 30, 40, and 30, resulting in 20 + 30 + 40 + 30 = 120. Then, if the total number of historical feature data is determined to be 4, the executing entity can divide the sum of the historical label values ​​by the total number, i.e., 120 / 4 = 30, to obtain the average value of the first label.

[0117] Step 430: Based on the first fitted curve and the mean of the first label, determine the initial residual duration corresponding to the test object.

[0118] In this step, the execution entity can compare the first label mean and the first fitted curve, draw a straight line with the first label mean and time, intersect the straight line with the first fitted curve to obtain the intersection point between the first label mean and the first fitted curve, and determine the time point corresponding to the intersection point as the initial residual duration of the test object.

[0119] In this implementation, by judging the data fitting conditions of historical feature data, and on the basis that the historical feature data meets the data fitting conditions, the initial residual duration corresponding to the test object is determined by using the first fitting curve and the mean of the first label. Multiple judgments can be made on the historical feature data, making the historical feature data more accurate, thereby obtaining a more accurate initial residual duration.

[0120] Continue to refer to Figure 4 Step 320 above, which determines the initial residual duration of the test object based on the test type and historical feature data of the test object, may also include the following steps:

[0121] Step 440: In response to determining that the historical feature data does not meet the data fitting conditions, obtain the similar feature data corresponding to the historical feature data.

[0122] In this step, if the execution entity determines that the historical feature data does not meet the data fitting conditions, it can obtain feature data similar to the historical feature data based on the data content of the historical feature data or the object type of the object to be tested, and determine the feature data as the similar feature data corresponding to the historical feature data. The similar feature data may include historical attribute adjustment strategies and date annotations, etc.

[0123] Step 450: Update the historical feature data based on the similar feature data to obtain the updated historical feature data.

[0124] In this step, the aforementioned executing entity can update the historical feature data based on the similar feature data, supplementing the historical feature data with content such as historical attribute adjustment strategies, date annotations, and historical label values ​​from the similar feature data, thus obtaining the updated historical feature data.

[0125] Step 460: Based on the updated historical feature data, determine the mean of the second label and the second fitted curve corresponding to the test object.

[0126] In this step, the aforementioned execution entity can perform data alignment based on the historical attribute adjustment strategy in the updated historical feature data to obtain aligned historical feature data. Following the time progression rule, the aligned historical feature data is fitted with time as the X-axis and historical label value as the Y-axis. That is, the aligned historical feature data is fitted with a gamma distribution to obtain a second fitting curve. This second fitting curve can be a curve obtained by fitting data with time as the X-axis and historical label value as the Y-axis.

[0127] Then, the aforementioned executing entity obtains each historical label value from the updated historical feature data, sums up multiple historical label values, and divides the resulting label sum by the total number of historical feature data to obtain the second label mean.

[0128] Step 470: Based on the second fitted curve and the mean of the second label, determine the initial residual duration corresponding to the test object.

[0129] In this step, the execution entity can compare the second label mean and the second fitted curve, draw a straight line with the second label mean and the time, and intersect the straight line and the second fitted curve to obtain the intersection point between the second label mean and the second fitted curve. The time point corresponding to the intersection point is determined as the initial residual duration of the test object.

[0130] In this implementation, by judging the data fitting conditions of historical feature data, if the historical feature data does not meet the data fitting conditions, it is necessary to update the historical feature data using similar feature data to make the historical feature data meet the data fitting conditions. Then, the initial residual duration corresponding to the test object is determined by using the second fitting curve and the mean of the second label. Multiple judgments and data updates can be performed on the historical feature data to make the historical feature data more accurate, thereby making the obtained initial residual duration more accurate.

[0131] Continue to refer to Figure 4 The above method may also include the following steps:

[0132] Step 480: In response to determining that the test type is the second preset type, select the initial residual duration corresponding to the test object from multiple preset residual durations.

[0133] In this step, if the execution entity determines that the test type is the second preset type, it can select the initial residual duration corresponding to the test object from multiple preset residual durations based on the historical characteristic data of the test object. This preset residual duration can be a pre-set fixed residual duration, for example, the overall time window is 100%, the preset residual duration is 5%, and only the middle 90% of the time window is used. The second preset type can be a data type that is a continuous variable or a discrete variable, or an application type that is significantly affected by attribute adjustments or has a relatively small impact on attribute adjustments. This second preset type can be a type pre-set by the operator based on experience.

[0134] Specifically, the aforementioned execution entity can present multiple preset residual durations to the user, who can then select an initial residual duration from these preset durations. Alternatively, the aforementioned execution entity can select the initial residual duration corresponding to the test object from multiple preset residual durations based on the correspondence between the test type and the preset residual durations. This correspondence can be a correspondence table pre-set by the operator based on experience, and this disclosure does not impose any specific limitations on it.

[0135] In this implementation, by selecting the initial residual duration corresponding to the test object from multiple preset residual durations, the initial residual duration corresponding to the test object can be determined simply and conveniently, thus improving the efficiency of determining the initial residual duration.

[0136] refer to Figure 5 , Figure 5 A flowchart 500 illustrates an embodiment of acquiring test feature data. Step 230 above, based on the time slice rotation rule, acquires the test feature data corresponding to the object under test, and may include the following steps:

[0137] The test feature data may include first test feature data and second test feature data. The first test feature data may be data obtained based on the time slice rotation rule, and the second test feature data may be data obtained based on the updated time slice rotation rule, wherein the time slice rotation rule can be updated in real time according to the obtained test feature data.

[0138] Step 510: Based on the time slice rotation rule, obtain the first test feature data corresponding to the object to be tested.

[0139] In this step, the aforementioned execution entity can conduct comparative testing on the test object according to the time slice rotation rule. This comparative testing is a comparative test conducted on the test object according to the attribute adjustment strategy. During the testing process, the test object can be compared and tested according to the set time slice and initial test parameters, using the attribute adjustment strategy to generate the first test feature data corresponding to the test object. The data format of the first test feature data is the same as that of the historical feature data. The first test feature data may include the sample number of the test object, the test date, the test label value, and feature information of multiple different dimensions.

[0140] Step 520: Based on the first test feature data and historical feature data, determine the new test parameters for the object to be tested.

[0141] In this step, after the aforementioned execution entity obtains the first test feature data, it can simultaneously perform data analysis on the first test feature data and historical feature data, analyze the periodicity of the first test feature data and historical feature data, obtain new period parameters corresponding to the first test feature data and historical feature data, and analyze the new residual duration of the first test feature data and historical feature data that is affected by the strategy during time slice rotation. Then, new test parameters including new period parameters, new residual duration, target random strategy, and target data quantity can be obtained.

[0142] Step 530: Based on the new test parameters, update the time slice rotation rules to obtain new time slice rotation rules.

[0143] In this step, after the aforementioned execution entity obtains the new test parameters, it can adjust the strategy according to the new test parameters and the attributes corresponding to the test object to update the time slice rotation rule and determine the new time slice rotation rule. The new time slice rotation rule is to conduct comparative testing according to the new test parameters and attribute adjustment strategy based on the time slice. The time slice rotation rule can be to divide the time into multiple time periods, each time period as a time slice, and conduct comparative testing on the data in each time slice according to the new test parameters and attribute adjustment strategy.

[0144] Step 540: Based on the new time slice rotation rule, obtain the second test feature data corresponding to the object to be tested.

[0145] In this step, after the aforementioned execution entity obtains the new time-slice rotation rules, it can continue to conduct comparative testing on the test object according to the new time-slice rotation rules. This comparative testing is conducted on the test object according to the attribute adjustment strategy. During the testing process, the test object can be compared and tested according to the set time slices and new test parameters, using the attribute adjustment strategy to generate the second test feature data corresponding to the test object. The data format of the second test feature data is the same as that of the historical feature data. The second test feature data may include the sample number of the test object, the test date, the test label value, and feature information of multiple different dimensions.

[0146] This disclosure does not specifically limit the first test feature data and the second test feature data. The aforementioned execution entity can determine new test parameters in real time based on the acquired test feature data, generate new time slice rotation rules based on the new test parameters, and continue to conduct comparative tests according to the new time slice rotation rules to acquire new test feature data until the total amount of acquired test feature data and historical feature data is not less than the target amount of data.

[0147] In this implementation, by dynamically updating the test parameters of the time slice rotation rule in real time during the acquisition of test feature data, the accuracy of the time slice rotation rule can be guaranteed, thereby improving the accuracy of the time slice rotation rule and making the acquired test feature data more accurate.

[0148] refer to Figure 6 , Figure 6 A flowchart 600 illustrates an embodiment of determining new period parameters, wherein the new test parameters include new period parameters; and step 520, which determines new test parameters for the object under test based on first test feature data and historical feature data, may include the following steps:

[0149] Step 610: Determine whether the test period corresponding to the first test feature data meets the period update conditions.

[0150] In this step, after obtaining the first test feature data, the execution entity can determine the test period corresponding to the first test feature data. This test period can be the execution period when performing comparative testing based on time slice rotation rules. Then, the execution entity can determine whether the test period corresponding to the first test feature data meets the period update condition, which may include being greater than or equal to the second test period.

[0151] Step 620: In response to determining that the test period corresponding to the first test feature data meets the period update condition, periodic calculations are performed based on the first test feature data and historical feature data to generate new period parameters for the object to be tested.

[0152] In this step, if the execution entity determines that the test period corresponding to the first test feature data meets the period update conditions, it can use the periodic graph method to perform periodic calculations on the first test feature data and historical feature data to obtain the new period parameters of the object to be tested.

[0153] In this implementation, by entering the second or subsequent test cycle, the periodic detection algorithm is used to re-detect the periodic parameters, and the latest periodic parameters are used in the time slice rotation. This ensures the accuracy of the time slice rotation rules, improves the accuracy of the time slice rotation rules, and makes the obtained test feature data more accurate.

[0154] refer to Figure 7 , Figure 7 A flowchart 700 illustrates an embodiment for determining a new residual duration, where the new test parameters include the new residual duration; and step 520 above, which determines new test parameters for the object under test based on first test feature data and historical feature data, may include the following steps:

[0155] Step 710: Based on the first test feature data and historical feature data, determine the mean of the third label and the third fitted curve corresponding to the test object.

[0156] In this step, the aforementioned execution entity can perform data alignment based on the attribute adjustment strategy in the first test feature data and historical feature data to obtain aligned feature data. Following the time progression rule, the aligned feature data is fitted with time as the X-axis and label value as the Y-axis, that is, the aligned feature data is fitted with a gamma distribution to obtain a third fitting curve. This third fitting curve can be a curve obtained by fitting data with time as the X-axis and label value as the Y-axis.

[0157] Then, the aforementioned execution entity obtains each label value from the first test feature data and the historical feature data, sums up the multiple label values, and divides the resulting label sum by the total number of the first test feature data and the historical feature data to obtain the third label mean.

[0158] Step 720: Based on the third label mean and the third fitted curve, determine the new residual duration of the test object.

[0159] In this step, the aforementioned execution entity can compare the third label mean and the third fitted curve, draw a straight line with the third label mean and time, and intersect the straight line with the third fitted curve to obtain the intersection point between the third label mean and the third fitted curve. The time point corresponding to the intersection point is determined as the new residual duration corresponding to the test object.

[0160] Alternatively, the aforementioned executing entity can obtain the shape parameters and scale parameters corresponding to the third fitted curve, multiply both shape parameters and scale parameters by a preset value, which can be (1-5%), to obtain new shape parameters and new scale parameters. A new fitted curve is obtained using the new shape parameters and new scale parameters. A straight line is drawn between the mean of the third label and the time, and the intersection of the straight line and the new fitted curve is obtained to obtain the intersection point between the mean of the third label and the new fitted curve. The time point corresponding to the intersection point is determined as the new residual duration corresponding to the test object.

[0161] In this implementation, by re-detecting the residual duration using the first test feature data and historical feature data, and adopting the latest residual duration in the time slice rotation, the accuracy of the time slice rotation rule can be guaranteed and improved, making the obtained test feature data more accurate.

[0162] refer to Figure 8 As an implementation of the methods shown in the above figures, this disclosure provides an embodiment of a data testing device based on time-slice rotation. This device embodiment is similar to... Figure 2 The method embodiments shown correspond to those described.

[0163] like Figure 8 As shown, the data testing device 800 based on time slice rotation in this embodiment may include: a parameter determination module 810, a rule determination module 820, an acquisition module 830, and a generation module 840.

[0164] Among them, the parameter determination module 810 is configured to determine the initial test parameters of the object under test based on the historical feature data of the object under test;

[0165] The rule determination module 820 is configured to determine the time slice rotation rules of the test object based on the initial test parameters and the attribute adjustment strategy of the test object.

[0166] The acquisition module 830 is configured to acquire test feature data corresponding to the object under test based on the time slice rotation rule.

[0167] The generation module 840 is configured to generate test results for evaluating attribute adjustment strategies based on test feature data and historical feature data.

[0168] In some optional implementations of this embodiment, the initial test parameters include initial period parameters and initial residual duration; and the parameter determination module 810 is further configured to: perform periodic calculations based on the historical feature data of the object under test to generate the initial period parameters corresponding to the object under test; and determine the initial residual duration corresponding to the object under test based on the test type and historical feature data corresponding to the object under test.

[0169] In some optional implementations of this embodiment, the initial test parameters also include the target randomization method and the target data quantity; and the parameter determination module 810 is further configured to: determine the target randomization strategy corresponding to the test object from multiple candidate randomization strategies based on the test type corresponding to the test object; and calculate the target data quantity corresponding to the test object based on the test type corresponding to the test object.

[0170] In some optional implementations of this embodiment, the parameter determination module 810 is further configured to: in response to determining that the test type is a first preset type, determine whether the historical feature data meets the data fitting conditions; in response to determining that the historical feature data meets the data fitting conditions, determine the first label mean and the first fitting curve corresponding to the test object based on the historical feature data; and determine the initial residual duration corresponding to the test object based on the first fitting curve and the first label mean.

[0171] In some optional implementations of this embodiment, the parameter determination module 810 is further configured to: in response to determining that the historical feature data does not meet the data fitting conditions, obtain similar feature data corresponding to the historical feature data; update the historical feature data based on the similar feature data to obtain updated historical feature data; determine the second label mean and the second fitting curve corresponding to the test object based on the updated historical feature data; and determine the initial residual duration corresponding to the test object based on the second fitting curve and the second label mean.

[0172] In some optional implementations of this embodiment, the parameter determination module 810 is further configured to: in response to determining that the test type is a second preset type, select the initial residual duration corresponding to the test object from a plurality of preset residual durations.

[0173] In some optional implementations of this embodiment, the test feature data includes first test feature data and second test feature data; and the acquisition module 830 is further configured to: acquire the first test feature data corresponding to the object under test based on the time slice rotation rule; determine new test parameters for the object under test based on the first test feature data and historical feature data; update the time slice rotation rule based on the new test parameters to obtain a new time slice rotation rule; and acquire the second test feature data corresponding to the object under test based on the new time slice rotation rule.

[0174] In some optional implementations of this embodiment, the new test parameters include new period parameters; and the acquisition module 830 is further configured to: determine whether the test period corresponding to the first test feature data meets the period update condition; in response to determining that the test period corresponding to the first test feature data meets the period update condition, perform periodic calculations based on the first test feature data and historical feature data to generate new period parameters for the object to be tested.

[0175] In some optional implementations of this embodiment, the new test parameters include a new residual duration; and the acquisition module 830 is further configured to: determine the third label mean and the third fitting curve corresponding to the test object based on the first test feature data and historical feature data; and determine the new residual duration of the test object based on the third label mean and the third fitting curve.

[0176] In some optional implementations of this embodiment, the generation module 840 is further configured to: determine whether the number of data corresponding to the test feature data and the historical feature data meets the quantity condition; in response to determining that the number of data corresponding to the test feature data and the historical feature data meets the quantity condition, determine the experimental sample set and the control sample set from the test feature data and the historical feature data; perform a significance test on the experimental sample set and the control sample set, and generate test results for evaluating the attribute adjustment strategy.

[0177] In some optional implementations of this embodiment, the historical characteristic data of the object under test satisfies at least one of the following: For warehouse storage scenarios, the historical characteristic data of the object under test includes storage data of the warehouse under adjusted warehouse attributes and storage data under unadjusted warehouse attributes; for item delivery scenarios, the historical characteristic data of the object under test includes delivery data of the items to be delivered under adjusted delivery attributes and delivery data under unadjusted delivery attributes; for item sales scenarios, the historical characteristic data of the object under test includes sales data of the items to be sold under adjusted item attribute information and sales data under unadjusted item attribute information.

[0178] The time-slice rotation-based data testing apparatus provided in the above embodiments of this disclosure first determines the initial test parameters of the test object based on its historical feature data. Then, based on the initial test parameters and the attribute adjustment strategy of the test object, it determines the time-slice rotation rules of the test object. Subsequently, based on the time-slice rotation rules, it obtains the corresponding test feature data of the test object. Finally, based on the test feature data and historical feature data, it generates test results for evaluating the attribute adjustment strategy. This apparatus can predetermine the initial test parameters using the historical feature data of the test object, determine the time-slice rotation rules using the initial test parameters and attribute adjustment strategy, and obtain the test feature data. Compared with existing methods for obtaining test feature data, it does not require traversing and simulating the entire application scenario of the test object. It can obtain test feature data using the determined time-slice rotation rules, saving computational resources and improving data testing efficiency. Furthermore, by using historical feature data to calculate the initial test parameters and referencing the data attributes of the historical feature data, the test feature data is more closely aligned with the test scenario of the test object, improving data testing efficiency and accuracy. It can be applied to various different application scenarios, improving the accuracy and applicability of data testing.

[0179] Those skilled in the art will understand that the above-described apparatus also includes other well-known structures, such as processors and memories. To avoid unnecessarily obscuring the embodiments of this disclosure, these well-known structures are... Figure 8 Not shown in the image.

[0180] It should be noted that the collection, gathering, updating, analysis, processing, use, transmission, and storage of user personal information involved in this disclosed technical solution all comply with relevant laws and regulations, are used for legitimate purposes, and do not violate public order and good morals. Necessary measures are taken to prevent unauthorized access to user personal information data and to safeguard user personal information security, network security, and national security.

[0181] According to embodiments of this disclosure, this disclosure also provides an electronic device, a readable storage medium, and a computer program product.

[0182] Figure 9 A schematic diagram of the structure of an electronic device 900 suitable for implementing embodiments of the present disclosure is shown. The terminal devices in the embodiments of the present disclosure may include, but are not limited to, mobile terminals such as smart screens, laptops, PADs (tablet computers), PMPs (portable multimedia players), in-vehicle terminals (e.g., in-vehicle navigation terminals), and fixed terminals such as digital TVs and desktop computers. Figure 9 The terminal device shown is merely an example and should not be construed as limiting the functionality and scope of the embodiments of this disclosure.

[0183] like Figure 9 As shown, electronic device 900 may include a processing device (e.g., a central processing unit, a graphics processing unit, etc.) 901, which can perform various appropriate actions and processes according to a program stored in read-only memory (ROM) 902 or a program loaded from storage device 908 into random access memory (RAM) 903. RAM 903 also stores various programs and data required for the operation of electronic device 900. Processing device 901, ROM 902, and RAM 903 are interconnected via bus 904. Input / output (I / O) interface 905 is also connected to bus 904.

[0184] Typically, the following devices can be connected to I / O interface 905: input devices 906 including, for example, touchscreens, touchpads, keyboards, mice, cameras, microphones, accelerometers, gyroscopes, etc.; output devices 907 including, for example, liquid crystal displays (LCDs), speakers, vibrators, etc.; storage devices 908 including, for example, magnetic tapes, hard disks, etc.; and communication devices 909. Communication device 909 allows electronic device 900 to communicate wirelessly or wiredly with other devices to exchange data. Although Figure 9 An electronic device 900 with various devices is shown; however, it should be understood that it is not required to implement or possess all of the devices shown. More or fewer devices may be implemented or possessed alternatively. Figure 9 Each box shown can represent a device or multiple devices as needed.

[0185] Specifically, according to embodiments of this disclosure, the processes described above with reference to the flowcharts can be implemented as computer software programs. For example, embodiments of this disclosure include a computer program product comprising a computer program carried on a computer-readable medium, the computer program containing program code for performing the methods shown in the flowcharts. In such embodiments, the computer program can be downloaded and installed from a network via a communication device 909, or installed from a storage device 908, or installed from a ROM 902. When the computer program is executed by a processing device 901, it performs the functions defined in the methods of embodiments of this disclosure. It should be noted that the computer-readable medium of embodiments of this disclosure can be a computer-readable signal medium or a computer-readable storage medium, or any combination of the two. The computer-readable storage medium can be, for example, but not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any combination thereof. More specific examples of computer-readable storage media may include, but are not limited to: electrical connections having one or more wires, portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fiber, portable compact disk read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination thereof. In embodiments of this disclosure, a computer-readable storage medium may be any tangible medium containing or storing a program that can be used by or in connection with an instruction execution system, apparatus, or device. In embodiments of this disclosure, a computer-readable signal medium may include a data signal propagated in baseband or as part of a carrier wave, carrying computer-readable program code. Such propagated data signals may take various forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination thereof. A computer-readable signal medium may also be any computer-readable medium other than a computer-readable storage medium, which can send, propagate, or transmit a program for use by or in connection with an instruction execution system, apparatus, or device. Program code contained on a computer-readable medium may be transmitted using any suitable medium, including but not limited to: wires, optical fibers, RF (radio frequency), etc., or any suitable combination thereof.

[0186] Computer program code for performing the operations of embodiments of this disclosure can be written in one or more programming languages ​​or a combination thereof. Programming languages ​​include object-oriented programming languages ​​such as Java, Smalltalk, and C++, as well as conventional procedural programming languages ​​such as the "C" language or similar programming languages. The program code can be executed entirely on the user's computer, partially on the user's computer, as a standalone software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In cases involving remote computers, the remote computer can be connected to the user's computer via any type of network, including a local area network (LAN) or a wide area network (WAN), or it can be connected to an external computer (e.g., via the Internet using an Internet service provider).

[0187] The flowcharts and block diagrams in the accompanying drawings illustrate the architecture, functionality, and operation of possible implementations of systems, methods, and computer program products according to various embodiments of this application. In this regard, each block in a flowchart or block diagram may represent a module, segment, or portion of code containing one or more executable instructions for implementing a specified logical function. It should also be noted that in some alternative implementations, the functions indicated in the blocks may occur in a different order than those indicated in the drawings. For example, two consecutively indicated blocks may actually be executed substantially in parallel, and they may sometimes be executed in reverse order, depending on the functions involved. It should also be noted that each block in the block diagrams and / or flowcharts, and combinations of blocks in the block diagrams and / or flowcharts, can be implemented using a dedicated hardware-based system that performs the specified function or operation, or using a combination of dedicated hardware and computer instructions.

[0188] The units described in the embodiments of this application can be implemented in software or hardware. The described units can also be housed in a processor; for example, a processor can be described as including a parameter determination module, a rule determination module, an acquisition module, and a generation module, wherein the names of these modules do not necessarily limit the module itself.

[0189] In another aspect, this application also provides a computer-readable medium, which may be included in the aforementioned electronic device or may exist independently without being assembled into the electronic device. The computer-readable medium carries one or more programs that, when executed by the electronic device, cause the electronic device to: determine initial test parameters for the object under test based on historical feature data of the object under test; determine time-slice rotation rules for the object under test based on the initial test parameters and the attribute adjustment strategy of the object under test; acquire test feature data corresponding to the object under test based on the time-slice rotation rules; and generate test results for evaluating the attribute adjustment strategy based on the test feature data and historical feature data.

[0190] The systems and technologies described herein can be implemented in computing systems that include backend components (e.g., as a data server), or computing systems that include middleware components (e.g., an application server), or computing systems that include frontend components (e.g., a user computer with a graphical user interface or web browser through which a user can interact with embodiments of the systems and technologies described herein), or any combination of such backend, middleware, or frontend components. The components of the system can be interconnected via digital data communication of any form or medium (e.g., a communication network). Examples of communication networks include local area networks (LANs), wide area networks (WANs), and the Internet.

[0191] Computer systems can include clients and servers. Clients and servers are generally located far apart and typically interact through communication networks. Client-server relationships are created by computer programs running on the respective computers and having a client-server relationship with each other.

[0192] The above description is merely a preferred embodiment of this disclosure and an explanation of the technical principles employed. Those skilled in the art should understand that the scope of the invention involved in the embodiments of this disclosure is not limited to technical solutions formed by specific combinations of the above-described technical features, but should also cover other technical solutions formed by arbitrary combinations of the above-described technical features or their equivalents without departing from the above-described inventive concept. For example, technical solutions formed by substituting the above-described features with (but not limited to) technical features with similar functions disclosed in the embodiments of this disclosure.

Claims

1. A time slice rotation based data testing method, the method comprising: determining initial testing parameters of a to-be-tested object based on historical feature data of the to-be-tested object; determining a time slice rotation rule of the to-be-tested object based on the initial testing parameters and an attribute adjustment strategy of the to-be-tested object; obtaining testing feature data corresponding to the to-be-tested object based on the time slice rotation rule; generating a testing result for evaluating the attribute adjustment strategy based on the testing feature data and the historical feature data.

2. The method of claim 1, wherein, the initial testing parameters comprise an initial period parameter and an initial residual duration; and the determining the initial testing parameters of the to-be-tested object based on the historical feature data of the to-be-tested object comprises: generating the initial period parameter corresponding to the to-be-tested object based on period calculation of the historical feature data of the to-be-tested object; determining the initial residual duration corresponding to the to-be-tested object based on a testing type corresponding to the to-be-tested object and the historical feature data.

3. The method of claim 2, wherein, the initial testing parameters further comprise a target random method and a target data quantity; and the method further comprises: determining the target random strategy corresponding to the to-be-tested object from a plurality of candidate random strategies based on the testing type corresponding to the to-be-tested object; calculating the target data quantity corresponding to the to-be-tested object based on the testing type corresponding to the to-be-tested object.

4. The method of claim 2, wherein, the determining the initial residual duration corresponding to the to-be-tested object based on the testing type corresponding to the to-be-tested object and the historical feature data comprises: in response to determining that the testing type is a first preset type, judging whether the historical feature data satisfies a data fitting condition; in response to determining that the historical feature data satisfies the data fitting condition, determining a first label mean value and a first fitting curve corresponding to the to-be-tested object based on the historical feature data; determining the initial residual duration corresponding to the to-be-tested object based on the first fitting curve and the first label mean value.

5. The method of claim 4, wherein, the determining the initial residual duration corresponding to the to-be-tested object based on the testing type corresponding to the to-be-tested object and the historical feature data further comprises: in response to determining that the historical feature data does not satisfy the data fitting condition, obtaining similar feature data corresponding to the historical feature data; updating the historical feature data based on the similar feature data to obtain updated historical feature data; determining a second label mean value and a second fitting curve corresponding to the to-be-tested object based on the updated historical feature data; determining the initial residual duration corresponding to the to-be-tested object based on the second fitting curve and the second label mean value.

6. The method of claim 2, further comprising: in response to determining that the testing type is a second preset type, selecting the initial residual duration corresponding to the to-be-tested object from a plurality of preset residual durations.

7. The method of claim 1, wherein, the testing feature data comprises first testing feature data and second testing feature data; and the obtaining the testing feature data corresponding to the to-be-tested object based on the time slice rotation rule comprises: acquire first test feature data corresponding to the to-be-tested object based on the time slice rotation rule; determine a new test parameter of the to-be-tested object based on the first test feature data and the historical feature data; update the time slice rotation rule based on the new test parameter, to obtain a new time slice rotation rule; acquire second test feature data corresponding to the to-be-tested object based on the new time slice rotation rule.

8. The method of claim 7, wherein, The new test parameter includes a new period parameter, and determining the new test parameter of the to-be-tested object based on the first test feature data and the historical feature data includes: determining whether a test period corresponding to the first test feature data meets a period update condition; in response to determining that the test period corresponding to the first test feature data meets the period update condition, performing periodic calculation based on the first test feature data and the historical feature data to generate a new period parameter of the to-be-tested object.

9. The method of claim 7 or 8, wherein, The new test parameter includes a new residual duration, and determining the new test parameter of the to-be-tested object based on the first test feature data and the historical feature data includes: determining a third label mean and a third fitting curve corresponding to the to-be-tested object based on the first test feature data and the historical feature data; determining a new residual duration of the to-be-tested object based on the third label mean and the third fitting curve.

10. The method of claim 1, wherein, The generating of the test result for evaluating the attribute adjustment strategy based on the test feature data and the historical feature data includes: determining whether a data quantity corresponding to the test feature data and the historical feature data meets a quantity condition; in response to determining that the data quantity corresponding to the test feature data and the historical feature data meets the quantity condition, determining an experimental sample set and a control sample set from the test feature data and the historical feature data; performing significance test on the experimental sample set and the control sample set to generate a test result for evaluating the attribute adjustment strategy.

11. The method of claim 1, wherein, The historical feature data of the to-be-tested object meets at least one of the following conditions: for a warehouse storage scenario, the historical feature data of the to-be-tested object includes storage data of a to-be-stored warehouse under an adjusted warehouse attribute and storage data of the to-be-stored warehouse under an unadjusted warehouse attribute; for an article distribution scenario, the historical feature data of the to-be-tested object includes distribution data of a to-be-distributed article under an adjusted distribution attribute and distribution data of the to-be-distributed article under an unadjusted distribution attribute; for an article sale scenario, the historical feature data of the to-be-tested object includes sales data of a to-be-sold article under an adjusted article attribute information and sales data of the to-be-sold article under an unadjusted article attribute information.

12. A time slice rotation-based data testing device, the device comprising: a parameter determination module configured to determine an initial test parameter of a to-be-tested object based on historical feature data of the to-be-tested object; a rule determination module configured to determine a time slice rotation rule of the to-be-tested object based on the initial test parameter and an attribute adjustment strategy of the to-be-tested object; An obtaining module configured to obtain, based on the time slice rotation rule, test feature data corresponding to the to-be-tested object; A generating module configured to generate, based on the test feature data and the historical feature data, a test result for evaluating the attribute adjustment strategy.

13. The apparatus of claim 12, wherein, The test feature data includes first test feature data and second test feature data; and the obtaining module is further configured to: obtain, based on the time slice rotation rule, first test feature data corresponding to the to-be-tested object; determine, based on the first test feature data and the historical feature data, a new test parameter of the to-be-tested object; update, based on the new test parameter, the time slice rotation rule to obtain a new time slice rotation rule; obtain, based on the new time slice rotation rule, second test feature data corresponding to the to-be-tested object.

14. An electronic device, comprising: one or more processors; storage configured to store one or more programs, when the one or more programs are executed by the one or more processors, the one or more processors are caused to implement the method of any one of claims 1-11.

15. A computer readable medium having stored thereon a computer program, characterized in that, The computer program is executed by the processor to implement the method of any one of claims 1-11.

16. A computer program product, comprising a computer program which, when executed by a processor, implements the method according to any one of claims 1-11.

Images