Business data push method, device and computer-readable storage medium

By receiving push configuration parameters, pushing business data to the target object group and using quantile inspection model to determine the optimal configuration parameters, the problem of low data processing efficiency in AB test is solved, and efficient business data push and optimization solutions are realized. Quickly confirm.

CN114912023BActive Publication Date: 2024-12-06TENCENT TECHNOLOGY (SHENZHEN) CO LTD
View PDF 2 Cites 0 Cited by

Patent Information

Application Number
CN202210509709.9
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2022-05-11
Publication Date
2024-12-06
Estimated Expiration
2042-05-11

AI Technical Summary

Technical Problem

In AB testing, it is difficult for the prior art to efficiently process and analyze large amounts of business data, which affects the efficiency and accuracy of the test data processing, resulting in slow determination of optimization solutions.

Method used

By receiving push configuration parameters, pushing business data to the target object group, and calculating cumulative business indicators based on feedback interactive data, using quantile inspection model to determine the optimal push configuration parameters, improve the determination efficiency of push configuration parameters and business data push experience effect.

Benefits of technology

It realizes faster and more accurate determination of target push configuration parameters, improves the efficiency and effectiveness of business data push, simplifies the data processing process, and has strong applicability.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN114912023B_ABST
    Figure CN114912023B_ABST
Patent Text Reader

Abstract

The embodiment of the present application discloses a method, device and computer-readable storage medium for pushing business data, wherein the method includes: when receiving a first push configuration parameter, pushing business data to a first target object group based on the first push configuration parameter, and pushing business data to a second target object group based on a second push configuration parameter. Obtaining business interaction data of the first target object group and the second target object group, obtaining a target object data set and a target business indicator data set based on the business interaction data, obtaining cumulative business indicator data of each target object based on the target object data set and the business indicator data set, and determining the cumulative business indicator data as input data of a quantile test model, and obtaining the first push configuration parameter as a confidence score of the target push configuration parameter through the quantile test model. By adopting the present application, the efficiency of determining the target push configuration parameter can be improved, and the experience effect of business data push can be enhanced.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present application relates to the field of computer technology, and in particular to a method and device for pushing business data, and a computer-readable storage medium. Background Art

[0002] With the rapid development of Internet technology, the optimization of various Internet products can be achieved through continuous algorithm iteration and parameter iteration. For example, in the process of optimizing various Internet products, AB testing is usually used to test the effect of product optimization, where AB testing refers to the candidate algorithms (or parameters) A and B in the optimization of Internet products, and each candidate algorithm (or parameter) can correspond to an optimization plan for a product. In AB testing, a series of business indicators (such as click-through rate) can be defined, and then algorithm A and algorithm B can be used to test on different test groups respectively, and the business indicator values ​​in the test group using a specific algorithm can be evaluated to determine which algorithm A and algorithm B are better, so as to determine the final optimization plan to update the Internet product.

[0003] In order to ensure the effectiveness of AB testing, in product optimization testing, it is necessary to process a large amount of test data (such as hundreds of billions) based on the feedback from the test group (such as test information, business indicator data, etc.) and obtain the test results through the AB testing model. In the process of processing a large amount of test data, how to improve the test data processing efficiency to ensure the accuracy of the test and the efficiency of AB testing has become one of the technical problems that need to be solved urgently. Summary of the invention

[0004] The embodiments of the present application provide a method, device and computer-readable storage medium for pushing business data, which can improve the efficiency of determining push configuration parameters for business data push, enhance the experience effect of business data push, and have strong applicability.

[0005] In a first aspect, an embodiment of the present application provides a method for pushing business data, the method comprising:

[0006] When the first push configuration parameter is received, the service data is pushed to the first target object group based on the first push configuration parameter, and the service data is pushed to the second target object group based on the determined second push configuration parameter;

[0007] Acquire the business interaction data fed back by the first target object group and the second target object group;

[0008] Obtaining a target object data set based on the business interaction data, wherein the target object data set includes target object data associated with each target object in each target object group;

[0009] A target business indicator data set is obtained based on the business interaction data, wherein the target business indicator data set includes multiple business indicator data of each target object, wherein the nth business indicator data of the multiple business indicator data of any target object is the sum of the business indicator data of the first statistical period to the nth statistical period in multiple consecutive statistical periods of the any target object, and n is less than or equal to the number of statistical periods in the multiple statistical periods;

[0010] Based on the above target object data set and the above business indicator data set, the cumulative business indicator data of each target object is obtained, and the above cumulative business indicator data is determined as the input data of the quantile test model, so as to obtain the above first push configuration parameter as the confidence score of the target push configuration parameter through the above quantile test model.

[0011] In a possible implementation, obtaining the target object data set based on the business interaction data includes:

[0012] Acquire object data in a current statistical period based on the business interaction data, and acquire a set of object data in a previous statistical period of the current statistical period based on the business interaction data;

[0013] Based on the object data in the current statistical period and the object data set, the object data set of the current statistical period is obtained as the target object data set;

[0014] The object data set in any statistical period includes object data of one or more target objects in each target object group, and the object data includes object attribute information or at least one item in the statistical period in which the object appears.

[0015] In a possible implementation, obtaining the target business indicator data set based on the business interaction data includes:

[0016] Acquire the first business indicator data of any of the target objects in the current statistical period based on the business interaction data, and acquire multiple second business indicator data of any of the target objects in a statistical period before the current statistical period based on the business interaction data;

[0017] Based on the first business indicator data and the plurality of second business indicator data, obtaining a plurality of business indicator data of any one of the target objects in the current statistical period;

[0018] The target business indicator data set is obtained based on the multiple business indicator data of the target objects in the current statistical period.

[0019] The business interaction data include at least one of the number of clicks or browsing time, and the business indicator data include at least one of the click rate or browsing time.

[0020] In a possible implementation, obtaining the object data set of the current statistical period based on the object data in the current statistical period and the object data set includes:

[0021] Select object data that is not repeated in the object data set from the object data in the current statistical period as newly added object data, and add the newly added object data to the object data set to obtain the object data set for the current statistical period.

[0022] In a possible implementation, the target business indicator data set includes M business indicator data of any one of the target objects, and the multiple business indicator data of any one of the target objects in the current statistical period obtained based on the first business indicator data and the multiple second business indicator data include:

[0023] Subtract the first second business indicator data from each second business indicator data in the plurality of second business indicator data to obtain M-1 business indicator data of any target object in the current statistical period;

[0024] Based on the above-mentioned first business indicator data and the M-1th business indicator data among the above-mentioned M-1 business indicator data, the Mth business indicator data of the above-mentioned any target object in the above-mentioned current statistical period is generated to obtain the M business indicator data of the above-mentioned any target object in the above-mentioned current statistical period.

[0025] In a possible implementation, the obtaining of the accumulated business indicator data of each target object based on the target object data set and the business indicator data set includes:

[0026] Determine a statistical period during which any of the target objects appear based on the target object data associated with any of the target objects in the target object data set;

[0027] When the statistical period in which any of the target objects first appears is one of the multiple consecutive statistical periods, the sum of the business indicator data of any of the target objects in the business indicator data set in each statistical period between the statistical period in which any of the target objects first appears and the current statistical period is determined as the cumulative business indicator data of any of the target objects;

[0028] When the statistical period in which any of the above-mentioned target objects first appears is not one of the above-mentioned multiple consecutive statistical periods, the sum of the business indicator data of any of the above-mentioned target objects in each statistical period in the above-mentioned multiple consecutive statistical periods in the above-mentioned business indicator data set shall be determined as the cumulative business indicator data of any of the above-mentioned target objects to obtain the cumulative business indicator data of each of the above-mentioned target objects.

[0029] In a possible implementation, the method further includes:

[0030] Input the accumulated business indicator data into the quantile test model, and obtain the confidence score of the first push configuration parameter as the target push configuration parameter through the quantile test model;

[0031] If the confidence score belongs to the preset rejection domain, the second push configuration parameter is used as the target push configuration parameter; if the confidence score does not belong to the preset rejection domain, the first push configuration parameter is used as the target push configuration parameter.

[0032] In a possible implementation, the business data includes multimedia data or interactive interface data, the first target object group includes multiple first target objects among all target objects of the target application, and the second target object group includes multiple second target objects other than the multiple first target objects among all target objects; the pushing of the business data to the first target object group based on the first push configuration parameter, and the pushing of the business data to the second target object group based on the determined second push configuration parameter include:

[0033] Pushing the multimedia data or the interactive interface data to the multiple first target objects based on the first push configuration parameter, and pushing the multimedia data or the interactive interface data to the multiple second target objects based on the second push configuration parameter, so as to display the multimedia data or the interactive interface corresponding to the interactive interface data through the application client of the target application associated with each of the target objects;

[0034] The obtaining of the business interaction data fed back by the first target object group and the second target object group includes:

[0035] The interactive operations performed by each target object in the first target object group and the second target object group based on the multimedia data displayed by the application client or the interactive interface are obtained, and the business interaction data fed back by each target object group is obtained based on the interactive operations.

[0036] In a second aspect, an embodiment of the present application provides a device for pushing service data, the device comprising:

[0037] A business data push module, configured to push business data to a first target object group based on the first push configuration parameter when receiving the first push configuration parameter, and to push the business data to a second target object group based on the determined second push configuration parameter;

[0038] An acquisition module, used to acquire the business interaction data fed back by the first target object group and the second target object group;

[0039] A cumulative business indicator data generation module, used to obtain a target object data set based on the business interaction data, obtain a target business indicator data set based on the business interaction data, and obtain the cumulative business indicator data of each target object based on the target object data set and the business indicator data set;

[0040] The target push configuration parameter determination module is used to determine the above-mentioned accumulated business indicator data as input data of the quantile test model, so as to obtain the confidence score of the above-mentioned first push configuration parameter as the target push configuration parameter through the above-mentioned quantile test model.

[0041] In a third aspect, an embodiment of the present application provides a computer device, the computer device comprising: a processor, a memory, and a network interface;

[0042] The above-mentioned processor is connected to a memory and a network interface, wherein the network interface is used to provide a data communication function, the above-mentioned memory is used to store program codes, and the above-mentioned processor is used to call the above-mentioned program codes to execute the method in the first aspect of the embodiment of the present application.

[0043] In a fourth aspect, an embodiment of the present application provides a computer-readable storage medium, wherein the computer-readable storage medium stores a computer program, wherein the computer program includes program instructions, and when the processor executes the program instructions, the method in the first aspect of the embodiment of the present application is executed.

[0044] In the present application, when receiving a first push configuration parameter, business data is pushed to a first target object group based on the first push configuration parameter, and business data is pushed to a second target object group based on a determined second push configuration parameter. Next, business interaction data fed back by the first target object group and the second target object group is obtained, and a target object data set and a target business indicator data set are obtained based on the business interaction data. The target object data set includes target object data associated with each target object in each target object group, and the target business indicator data set includes multiple business indicator data of each target object. Here, the nth business indicator data in the multiple business indicator data of any target object is the sum of the business indicator data of any target object from the first statistical period to the nth statistical period in multiple consecutive statistical periods. Based on the target object data set and the business indicator data set, the cumulative business indicator data of each target object is obtained, and the above-mentioned cumulative business indicator data is determined as the input data of the quantile test model, so as to obtain the first push configuration parameter as the confidence score of the target push configuration parameter through the quantile test model. Based on the present application, the determination efficiency of the target push configuration parameter can be improved, the experience effect of business data push can be enhanced, and the operation is simple and the applicability is strong. BRIEF DESCRIPTION OF THE DRAWINGS

[0045] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings required for use in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0046] Figure 1 It is a schematic diagram of the system architecture provided by the embodiment of the present application;

[0047] Figure 2 It is a scenario diagram of the method for pushing business data provided in an embodiment of the present application;

[0048] Figure 3 It is a flowchart of a method for pushing business data provided by an embodiment of the present application;

[0049] Figure 4 This is a schematic diagram of generating a target business indicator data set provided by an embodiment of the present application;

[0050] Figure 5 This is a schematic diagram of generating cumulative business indicator data provided by an embodiment of the present application;

[0051] Figure 6 It is a structural diagram of a device for pushing business data provided in an embodiment of the present application;

[0052] Figure 7It is a schematic diagram of the structure of a computer device provided in an embodiment of the present application. DETAILED DESCRIPTION

[0053] The following will be combined with the drawings in the embodiments of the present application to clearly and completely describe the technical solutions in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of this application.

[0054] The method for pushing business data provided by the embodiment of the present application (or simply the method provided by the embodiment of the present application) is applicable to push various types of business data (such as multimedia data and interactive interface data, etc.) in applications (such as video applications, news applications, short video applications, and social applications, etc.) with video and graphic content viewing and / or uploading functions. Specifically, the method provided by the embodiment of the present application is applicable to selecting the optimal push configuration parameter (which can be called the target push configuration parameter) from multiple push configuration parameters to push business data. Here, the above configuration parameters can correspond to an application strategy (such as a business data push strategy), an algorithm model (such as a business data recommendation model), and the type of application interactive interface, which can be determined according to the actual application scenario and is not limited here. The above multiple push configuration parameters may include new configuration parameters developed for the above application (which can be called the target application) (which can be developed through the business background or algorithm background), and the configuration parameter (which can be the first push configuration parameter) can be compared with other existing configuration parameters (which can be the second push configuration parameter) to determine the target push configuration parameter, so that the business data can be pushed based on the target push configuration parameter to enhance the business data push effect. The target push configuration parameter can be determined by AB testing. AB testing refers to selecting one of multiple candidate parameters (or algorithms) A and B (such as the first push configuration parameter and the second push configuration parameter) to optimize the algorithm or model. Each candidate algorithm or parameter can correspond to an optimization solution. In AB testing, a series of business indicators (such as click-through rate, browsing time, etc.) can be defined, and then parameters (or algorithms) A and parameters (or algorithms) B can be used to test different test groups. The business indicator values ​​in the test group and control group using different parameters and algorithms are evaluated to determine which parameter A or parameter B is better, so as to determine the final optimization solution (which can correspond to an application strategy, algorithm model or application interaction interface, etc.), and apply it to all objects associated with the above target application (i.e., all are switched to the optimal solution or strategy).

[0055] In the method provided in the embodiment of the present application, in the process of determining the target push configuration parameters through AB testing, the mean is usually used as a test indicator for AB testing (or called mean test). However, using the mean as a test indicator sometimes cannot accurately obtain the differences between different strategies and schemes. For example, some business indicators have no obvious mean differences under different strategies or schemes due to distribution problems, but the distribution curves are different (such as multimedia data loading time, etc.). This kind of business indicator often uses quantiles as test indicators to evaluate strategies. Or, some business indicators such as the number of rewards, if the number of target objects in the group is small, it is very likely that the large amount of rewards from a very small number of target objects makes the mean of the strategy higher than that of other groups (different groups correspond to different strategies), and if the quantile is used as a test indicator, there is no obvious difference between the strategies. For another example, there are some business indicators such as the number of publications. A large number of target objects may have 0 related business indicators every day. If the sampling mean is used as a test indicator, the average business indicator of each group is almost 0. The judgment of this result is sensitive to outlier data anomalies, and it is easy to make wrong judgments on the pros and cons of strategies or plans. In this case, using quantiles as test indicators for AB testing can well detect the difference in strategies. For example, 90% of the target objects in strategy A have 0 publications, while 95% of the target objects in strategy B have 0 publications, which greatly reduces the sensitivity to outlier data anomalies. It can be seen that using quantiles as test indicators for AB testing (or quantile testing) is more applicable, and the selection effect of target push configuration parameters is better.

[0056] In the method provided in the embodiment of the present application, all or part of the objects associated with the above-mentioned target application (which may be all objects using the target application) can be selected as target objects to determine the target push configuration parameters. In other words, the object that is hit and participates in the determination process of the above-mentioned target push configuration parameters can be used as the target object. Further, different target object groups (such as a first target object group and a second target object group, the first target object group includes a plurality of first target objects, and the second target object group includes a plurality of second target objects) can be divided from the multiple target objects associated with the above-mentioned target application based on the above-mentioned first push configuration parameters and the above-mentioned second push configuration parameters, and the business data generated based on different push configuration parameters can be pushed to different target object groups, so that feedback data (such as business interaction data) from different target object groups can be obtained, so as to determine the target push configuration parameters based on the business interaction data of different target object groups (for example, determining the target push configuration parameters through AB testing). Optionally, in the process of determining the target push configuration parameters from the first push configuration parameters and the second push configuration parameters through AB testing, multiple business indicators (including click-through rate, browsing time, etc.) can be defined, that is, in the process of selecting the target push configuration parameters through AB testing, multiple experiments can be included, and each experiment can correspond to a business indicator. In addition, each experiment can include multiple groups, and each group corresponds to a strategy (for example, a push strategy based on the first push configuration parameter or a push strategy based on the second push configuration parameter). In other words, in the process of selecting a target object from the objects associated with the above-mentioned target application to participate in the AB test to determine the target push configuration parameters, the target object can participate in an experiment in the AB test (corresponding to a business indicator, which can be a click-through rate), and the target object can participate in (or be hit by) a group in the above-mentioned experiment to participate in the experiment, and the target objects in different groups correspond to different strategies. In the experiment, the business indicators corresponding to the target objects in each group participating in the experiment are obtained, so as to select the target push configuration parameters based on the business indicators fed back by each target object. Optionally, each of the above-mentioned target objects can participate in multiple experiments at the same time, but can only enter one group of the experiment in each participating experiment. If the process of determining the target push configuration parameters from the first push configuration parameters and the second push configuration parameters through AB testing includes multiple experiments, each experiment may include a group corresponding to the push strategy based on the first push configuration parameters and a group corresponding to the push strategy based on the second push configuration parameters. The object group composed of target objects in the group corresponding to the push strategy based on the first push configuration parameters in each experiment can correspond to the above-mentioned first target object group, and the object group composed of target objects in the group corresponding to the push strategy based on the second push configuration parameters in each experiment can correspond to the above-mentioned second target object group.

[0057] In the method provided in the embodiment of the present application, in order to ensure the effect of AB testing, it is necessary to test based on a large number of target objects, that is, a large amount of test data (such as hundreds of billions of orders of magnitude) fed back by the target object (which can be business interaction data) is processed and then the test results (such as determined target push configuration parameters) are obtained through the relevant AB test model (such as a quantile test model). In the process of performing AB testing through a quantile test model (such as a quantile test model based on the ConQ algorithm), it is necessary to obtain the sum of the business indicators of each target object (also referred to as the cumulative business indicator data of each target object) since each target object has been hit (that is, participating in the AB test as a target object) in each statistical period (such as every day). The cumulative business indicator data of each target object can be expressed in the form of a five-tuple field table, and the fields in the five-tuple field table may include: date, object identity, experiment identifier, group identifier and cumulative business indicator. The date in the above five-tuple field table is used to indicate the date corresponding to the current statistical period (such as the date of the day). The object identity in the above five-tuple field table is used to indicate the target object participating in the above AB test in the current statistical period. The above-mentioned AB test may include multiple experiments, and the experiment identifier in the above-mentioned five-tuple field table is used to indicate the type of experiment in which each target object participates, for example, it may be an experiment whose corresponding business indicator is click-through rate, or an experiment whose corresponding business indicator is browsing time, etc. Each experiment in the above-mentioned AB test may include multiple groups, and each group corresponds to a strategy (for example, a push strategy based on the first push configuration parameter or a push strategy based on the second push configuration parameter). The group identifier in the above-mentioned five-tuple field table is used to indicate the strategy corresponding to each target object, wherein each target user may be hit by a large number of different experiments in different statistical periods, but it is impossible to be assigned to multiple groups in the same experiment. The above-mentioned five-tuple field table can be obtained by associating the hit table and the business table, so as to obtain the cumulative business indicator data of each target object included in the above-mentioned five-tuple field table, and the above-mentioned hit table can record the fields including: date, object identity, experiment identifier and group identifier, and the above-mentioned business table can record the fields including: date, object identity and business indicator. Unlike the mean test, which requires a five-tuple field table to calculate the difference in the mean values ​​of the business indicators of each group under each experiment, the quantile test requires obtaining the cumulative business indicator data of each target object for AB testing. The number of statistical cycles for each experiment is different (it can be a few days, a few weeks, etc.). Therefore, when the business table and the hit table are associated, the business table cannot only read the results of the current statistical cycle. In order to avoid data loss, all business table data since the beginning of the earliest ongoing experiment must be rolled back and read. Therefore, the calculation becomes a problem of associating a table of hundreds of billions of people with a table of tens of billions of people. This is very expensive in terms of both machine cost and calculation time.In addition, since each target object is hit at different times in different experiments during association, the associated key conditions cannot be directly implemented through simple SQL statements, and very detailed processing logic needs to be developed. Since it takes too long to obtain the cumulative business indicator data of each target object, the AB test results are generated too slowly and the optimal solution (such as the business data push solution corresponding to the target push configuration parameters) cannot be determined in time, which delays the testing of other new solutions and results in poor business data push results.

[0058] In the method provided in the embodiment of the present application, by obtaining the business interaction data fed back by each target object (which may be the target object in the first target object group and the second target object group), and obtaining the target object data set and the target business indicator data set based on the above business interaction data, the above target object data set includes the target object data associated with each target object, and the above target business indicator data set includes multiple business indicator data of each target object, where the nth business indicator data in the multiple business indicator data of any target object is the sum of the business indicator data of any target object from the first statistical period to the nth statistical period in multiple consecutive statistical periods. Then, based on the above target object data set and the above target business indicator data set, the cumulative business indicator data of each target object is obtained, and the cumulative business indicator data is input into the quantile test model, so as to determine the target push configuration parameter for business data push based on the test result of the quantile test model (which may be the confidence score corresponding to the first push configuration parameter as the target push configuration parameter). In other words, by processing the business table (obtained from business interaction data) into a score table (i.e., corresponding to the above target business indicator data set), and then associating it with the cumulative hit table obtained after processing the hit table (i.e., corresponding to the above target object data set), the entire operation can be converted into an association between billions and billions of tables, greatly reducing the consumption of computing resources, and the acquisition of the cumulative business indicator data of each target object can be completed in a shorter time, without the need to roll back a large number of historical business tables. The calculation principle can be implemented through the simplest SQL development, making it easy to achieve automatic deployment / batch deployment of different indicators. Therefore, determining the target push configuration parameters through AB testing can consume less machine computing resources, more efficiently generate AB test results to determine the target push configuration parameters, thereby quickly completing the scheme comparison evaluation, and more quickly determining the optimal scheme (such as the business data push scheme corresponding to the target push configuration parameters), and pushing business data based on the target push configuration parameters, with good business data push effect and strong applicability.

[0059] See also Figure 1 , Figure 1 Schematic diagram of the system architecture provided by the embodiment of the present application. Figure 1As shown, the system architecture may include a business server 100 and a terminal cluster, and the terminal cluster may include: terminal devices 200a, terminal device 200b, terminal device 200c, ..., terminal device 200n and the like. Among them, the above-mentioned business server 100 may be an independent physical server, or a server cluster or distributed system composed of multiple physical servers, or a cloud server that provides basic cloud computing services such as cloud database, cloud service, cloud computing, cloud function, cloud storage, network service, cloud communication, middleware service, domain name service, security service, CDN, and big data and artificial intelligence platform. The terminal device (including terminal device 200a, terminal device 200b, terminal device 200c, ..., terminal device 200n) may be a smart phone, tablet computer, laptop computer, desktop computer, PDA, mobile internet device (mobile internet device, MID), wearable device (such as smart watch, smart bracelet, etc.), smart computer, smart car and other smart terminals. Among them, the business server 100 can establish a communication connection with each terminal device in the terminal cluster, and a communication connection can also be established between each terminal device in the terminal cluster. In other words, the business server 100 can establish a communication connection with each terminal device in the terminal device 200a, the terminal device 200b, the terminal device 200c, ..., the terminal device 200n. For example, a communication connection can be established between the terminal device 200a and the business server 100. A communication connection can be established between the terminal device 200a and the terminal device 200b, and a communication connection can also be established between the terminal device 200a and the terminal device 200c. Among them, the above-mentioned communication connection does not limit the connection method, and can be directly or indirectly connected through a wired communication method, or directly or indirectly connected through a wireless communication method, etc., which can be determined according to the actual application scenario, and this application does not limit it here.

[0060] It should be understood that Figure 1 Each terminal device in the terminal cluster shown in the figure may be installed with an application client. When the application client runs in each terminal device, it can be respectively connected to the above Figure 1The business servers 100 shown in the figure interact with each other so that the business server 100 can receive business data from each terminal device, or the business server 100 pushes business data (which can be multimedia data, interactive interface data, etc.) to each terminal device. Among them, the above-mentioned application client can be an application client with the function of displaying data information such as text, images and videos, such as social applications, instant messaging applications, live broadcast applications, news applications, short video applications, video applications, music applications, shopping applications, novel applications, payment applications, etc., which can be determined according to the actual application scenario requirements and are not limited here. Among them, the application client can be an independent client or an embedded sub-client integrated in a client (such as an instant messaging client, a social client, etc.), which can be determined according to the actual application scenario and are not limited here. Taking the short video application as an example, when the user uses the short video application through the terminal device, the user can upload the original content of the relevant user (such as the video shot and uploaded by the user himself, the video made by himself, etc.) through the terminal device. As a server of the short video application, the business server 100 can be a collection of multiple servers including the background server corresponding to the application client and the data processing server. The business server 100 can receive multimedia data from the terminal device (for example, user-generated content uploaded by the user through the terminal device). In addition, the business server 100 can also push multimedia data to each terminal device. The business server 100 can push multimedia data to different terminal devices based on different push configuration parameters, that is, the business server 100 can divide different push groups composed of multiple terminal devices from multiple terminal devices associated with the short video application (each terminal device can correspond to an object) based on different push configuration parameters, so as to push business data generated based on different push configuration parameters to the terminal devices in different push groups. For example, the business server 100 can push multimedia data to the terminal devices in the first push group based on the first push configuration parameter, and can push the above multimedia data to the terminal devices in the second push group based on the second push configuration parameter. Each of the above terminal devices receives and displays the multimedia data pushed by the business server 100 to the corresponding object. Here, the business server 100 pushes multimedia data to the terminal devices in different push groups based on different push configuration parameters, and the terminal devices in different push groups display different multimedia data on their associated short video applications (for example, the order of pushing the displayed multimedia data is different). Please refer to Figure 2 , Figure 2 Schematic diagram of a scenario of a method for pushing business data provided by an embodiment of the present application. Figure 2 As shown, the above short video application can be Figure 2As shown in interface 1, the multimedia data displayed in the interface 1 may be multimedia data pushed by the business server 100 to the terminal devices in the first push group based on the first push configuration parameters. The multimedia data displayed in interface 1 may include multimedia data 10, multimedia data 20 and multimedia data 30. The push order of each multimedia data is: multimedia data 10, multimedia data 20 and multimedia data 30. Figure 2 The interface 2 also includes an interface 2, and the multimedia data displayed in the interface 2 may be multimedia data pushed by the business server 100 to the terminal devices in the second push group based on the second push configuration parameter. The multimedia data displayed in the interface 2 may include multimedia data 10, multimedia data 20 and multimedia data 30, and the order of pushing each multimedia data is: multimedia data 20, multimedia data 10 and multimedia data 30. That is, the interface 1 and the interface 2 include the same multimedia data, and because the business server 100 pushes multimedia data to the terminal devices corresponding to the interface 1 and the interface 2 based on different push configuration parameters, the order of pushing the multimedia data displayed in each interface is different. The object corresponding to the terminal device can browse the displayed media data, and the business server 100 can obtain different interactive operations (such as clicking to browse, triggering the praise function, etc.) of each object based on the multimedia data displayed by the short video application, and obtain the business interaction data fed back by each terminal device based on the above-mentioned interactive operations, so as to determine (for example, it can be selected through AB testing) the target push configuration parameters based on the business interaction data and push the subsequent multimedia data to all terminal devices associated with the above-mentioned short video application based on the target push configuration parameters.

[0061] The method provided in the embodiment of the present application can be performed by Figure 1 The service server 100 shown in the figure may also be executed by a terminal device (such as Figure 1 The method may be executed by any one of the terminal devices 200a, 200b, ..., 200n shown in the figure, or may be executed jointly by the terminal device and the service server. The specific method may be determined according to the actual application scenario and is not limited here.

[0062] In some feasible implementations, the business server can push business data to multiple terminal devices installed with the application client based on the first push configuration parameter, and the multiple terminal devices can be associated with multiple target objects (for the convenience of description, the first target object group can be used as an example for illustration, and the first target object group can include multiple first target objects). The above business data may include multimedia data or interactive interface data. The business server can also push business data to other terminal devices other than the multiple terminal devices associated with the first target object group based on the second push configuration parameter, and the other terminal devices can be associated with multiple target objects (for the convenience of description, the second target object group can be used as an example for illustration, and the second target object group can include multiple second target objects). For the convenience of description, the data interaction between the business server and the above terminal device can also be equivalent to the data interaction between the business server and the first target object group or the second target object group, which will not be repeated below. The business server can obtain (through the corresponding terminal device) the interactive operation performed by the first target object group and the second target object group based on the multimedia data or interactive interface displayed by the application client, and obtain the business interaction data fed back by each target object group (through the corresponding terminal device) based on the interactive operation. Since different target object groups display different business data on the application client of the target application associated with them (for example, the order of multimedia data push displayed is different, or different user interaction interfaces are displayed), different interactive operations performed by each target object in different target object groups based on the multimedia data or interaction interface displayed by the above application client can be obtained, and the business interaction data fed back by each target object group can be obtained based on the above interactive operations to determine the target push configuration parameters based on the business interaction data. The business server can obtain the object data in the current statistical period based on the above business interaction data, and obtain the object data set of the previous statistical period of the current statistical period based on the above business interaction data. The business server can obtain the object data set of the current statistical period as the target object data set based on the object data in the current statistical period and the above object data set, and the target object data set includes the target object data corresponding to each target object in the current statistical period, that is, all target object data participating in the AB test (to determine the target push configuration parameters) in the current statistical period, so as to generate the cumulative business indicator data of each target object for determining the target push configuration parameters based on the target object data set and the target business indicator data set.The business server can also obtain a target business indicator data set based on the above business interaction data, and the above business indicator data set can include multiple business indicator data of each target object. Here, the nth business indicator data in the multiple business indicator data of any target object is the sum of the business indicator data of any target object from the first statistical period to the nth statistical period in multiple consecutive statistical periods, and n is not greater than the number of statistical periods in the above multiple statistical periods. By obtaining multiple business indicator data of each target object, the business server can quickly obtain the cumulative business indicator data of each target object based on the multiple business indicator data of each target object to select the target push configuration parameters. The business server can determine the statistical period in which any target object first appears based on the target object data associated with any target object in the above target object data set. When the above-mentioned statistical period is one of the above-mentioned continuous multiple statistical periods, that is, any of the above-mentioned target objects is hit in the above-mentioned continuous multiple statistical periods and participates in the AB test (that is, participates in the determination process of the above-mentioned target push configuration parameters), the business server can obtain the sum of the business indicator data of each statistical period from the above-mentioned first statistical period to the above-mentioned current statistical period based on the above-mentioned target business indicator data set as the cumulative business indicator data of any of the above-mentioned target objects. Otherwise, the sum of the business indicator data of each statistical period in the above-mentioned continuous multiple statistical periods of any of the above-mentioned target objects is obtained as the cumulative business indicator data of any of the above-mentioned target objects, thereby obtaining the cumulative business indicator data of each target object. By obtaining the cumulative business indicator data of each target object based on the target object data set and the target business indicator data set, the acquisition of the cumulative business indicator data of each target object can be completed in a shorter time. The business server can also input the above-mentioned accumulated business indicator data into a quantile test model (such as a quantile test model based on the ConQ algorithm). The quantile test model can be based on the assumption that the above-mentioned first push configuration parameter is the target push configuration parameter (or, the above-mentioned first push configuration parameter is better than the above-mentioned second push configuration parameter). The confidence score of the above-mentioned assumption is obtained through the quantile test model to determine the target push configuration parameter based on the confidence score, so that the push strategy for business data push based on the above-mentioned target push configuration parameter can be applied to all terminal devices associated with the above-mentioned application clients, thereby further enhancing the business data push effect and having strong applicability.

[0063] In some feasible implementations, the terminal device 200a may push business data to multiple other terminal devices installed with the application client based on the first push configuration parameter, and the multiple terminal devices may be associated with multiple target objects (for the convenience of description, the first target object group may be used as an example for illustration, and the first target object group may include multiple first target objects). The business data may include multimedia data or interactive interface data. The terminal device 200a may also push business data to other terminal devices other than the multiple terminal devices associated with the first target object group based on the second push configuration parameter, and the other terminal devices may be associated with multiple target objects (for the convenience of description, the second target object group may be used as an example for illustration, and the second target object group may include multiple second target objects). For the convenience of description, the data interaction between the terminal device 200a and the other terminal devices may also be equivalent to the data interaction between the terminal device 200a and the first target object group or the second target object group, which will not be repeated below. The terminal device 200a can obtain the interactive operations performed by the first target object group and the second target object group based on the multimedia data or interactive interface displayed by the application client, and obtain the business interaction data fed back by each target object group based on the interactive operations. Since different target object groups display different business data on the application client of their associated target application (for example, the displayed multimedia data push order is different, or different user interaction interfaces are displayed), different interactive operations performed by each target object in the different target object groups based on the multimedia data or interactive interface displayed by the application client can be obtained, and the business interaction data fed back by each target object group can be obtained based on the interactive operations to determine the target push configuration parameters based on the business interaction data. The terminal device 200a can obtain the object data in the current statistical period based on the business interaction data, and obtain the object data set of the previous statistical period of the current statistical period based on the business interaction data. The terminal device 200a can obtain the object data set of the current statistical period as the target object data set based on the object data in the current statistical period and the object data set. The target object data set includes the target object data corresponding to each target object in the current statistical period, that is, all target object data participating in the AB test (to determine the target push configuration parameters) in the current statistical period, so as to generate the cumulative business indicator data of each target object for determining the target push configuration parameters based on the target object data set and the target business indicator data set.The terminal device 200a can also obtain a target business indicator data set based on the above-mentioned business interaction data, and the above-mentioned business indicator data set can include multiple business indicator data of each target object. Here, the nth business indicator data in the multiple business indicator data of any target object is the sum of the business indicator data of any target object from the 1st statistical period to the nth statistical period in multiple consecutive statistical periods, and n is not greater than the number of statistical periods in the above-mentioned multiple statistical periods. By obtaining multiple business indicator data of each target object, the terminal device 200a can quickly obtain the cumulative business indicator data of each target object based on the multiple business indicator data of each target object to select the target push configuration parameters. The terminal device 200a can determine the statistical period in which any target object first appears based on the target object data associated with any target object in the above-mentioned target object data set. When the above-mentioned statistical period is one of the above-mentioned continuous multiple statistical periods, that is, any of the above-mentioned target objects is hit in the above-mentioned continuous multiple statistical periods and participates in the AB test (that is, participates in the determination process of the above-mentioned target push configuration parameters), the terminal device 200a can obtain the sum of the business indicator data of each statistical period from the above-mentioned first statistical period to the above-mentioned current statistical period based on the above-mentioned target business indicator data set as the cumulative business indicator data of any of the above-mentioned target objects. Otherwise, the sum of the business indicator data of each statistical period in the above-mentioned continuous multiple statistical periods of any of the above-mentioned target objects is obtained as the cumulative business indicator data of any of the above-mentioned target objects, thereby obtaining the cumulative business indicator data of each target object. By obtaining the cumulative business indicator data of each target object based on the target object data set and the target business indicator data set, the acquisition of the cumulative business indicator data of each target object can be completed in a shorter time. The terminal device 200a can also input the above-mentioned accumulated business indicator data into a quantile test model (such as a quantile test model based on the ConQ algorithm), which can be based on the assumption that the above-mentioned first push configuration parameter is the target push configuration parameter (or, the above-mentioned first push configuration parameter is better than the above-mentioned second push configuration parameter), and obtain the confidence score of the above-mentioned assumption through the quantile test model to determine the target push configuration parameter based on the confidence score. Therefore, the push strategy for business data push based on the above-mentioned target push configuration parameter can be applied to all terminal devices associated with the above-mentioned application client, thereby further enhancing the business data push effect, and the applicability is strong.

[0064] For the convenience of description, the terminal device will be used as the execution subject of the method provided in the embodiment of the present application, and the method for pushing business data through the terminal device will be specifically explained through the following embodiment.

[0065] See also Figure 3 , Figure 3FIG. 1 is a flow chart of a method for pushing service data provided in an embodiment of the present application. Figure 3 As shown, the method comprises the following steps:

[0066] S101: When a first push configuration parameter is received, business data is pushed to a first target object group based on the first push configuration parameter, and business data is pushed to a second target object group based on a determined second push configuration parameter.

[0067] In some feasible implementations, a terminal device (such as terminal device 200a) may receive a new configuration parameter developed for a target application (such as a short video application) (which may be developed through a business background or an algorithm background), and the configuration parameter (which may be a first push configuration parameter) may correspond to an application strategy (such as a short video push strategy), an algorithm model (such as a short video recommendation model), and an application interaction interface, which may be determined according to the actual application scenario and is not limited here. After receiving the first push configuration parameter, the terminal device may push business data to multiple target objects (for the convenience of description, the first target object group may be used as an example, and the first target object group may include multiple first target objects) using the above-mentioned target application based on the first push configuration parameter. Specifically, the above-mentioned business data may include multimedia data or interaction interface data, and the above-mentioned multimedia data may include various types of video and graphic content, and the above-mentioned various types of video and graphic content may be professionally produced content (Professional Generated Content, PGC) released by self-media and institutions, or may be user-generated content (User Generated Content, UGC) created by users, which may be determined according to the actual application scenario and is not limited here by this application. Taking the example that the business data includes multimedia data, the terminal device can push multiple target multimedia data to the first target object group through the algorithm model (such as the short video recommendation model) configured with the first push configuration parameter. In addition, the terminal device can also push multiple target multimedia data to the second target object group other than the first target object group among the multiple target objects of the target application through the algorithm model configured with the second push configuration parameter, so as to push the business data to different target object groups (such as the first target object group and the second target object group) to determine the configuration parameters (which can be called target push configuration parameters) with the best business data push effect. Here, the second push configuration parameter can be a push configuration parameter obtained before receiving the first push configuration parameter (such as obtained through the business background or algorithm background development). In other words, the terminal device can divide different target object groups from the multiple target objects associated with the target application based on the first push configuration parameter and the second push configuration parameter to push business data generated based on different push configuration parameters to different target object groups.For example, when the business data is multimedia data (which may include multimedia data 10, multimedia data 20, and multimedia data 30), the terminal device may push the business data to the first target object group based on the first push configuration parameter (which may be an algorithm model configured with the first push configuration parameter), and the order of pushing the multimedia data displayed by the application client of the target application associated with each first target object in the first target object group is multimedia data 10, multimedia data 20, and multimedia data 30. Similarly, the terminal device may push the business data to the second target object group based on the second push configuration parameter (which may be an algorithm model configured with the second push configuration parameter), and the order of pushing the multimedia data displayed by the application client of the target application associated with each second target object in the second target object group is multimedia data 20, multimedia data 10, and multimedia data 30. It can be seen that since business data is pushed to different target object groups based on different push configuration parameters, different target object groups display different business data on the application client of their associated target applications (for example, the order of multimedia data push displayed is different, or different user interaction interfaces are displayed), so that feedback data (such as business interaction data) from different target object groups can be obtained, and the target push configuration parameters can be determined based on the business interaction data of different target object groups to obtain the target push configuration parameters to update the business data push plan of the above-mentioned target application, thereby achieving better business data push effect.

[0068] S102: Acquire business interaction data fed back by the first target object group and the second target object group.

[0069] In some feasible implementations, the terminal device can obtain the business interaction data fed back from the above-mentioned first target object group and the above-mentioned second target object group. Specifically, the terminal device can obtain the interactive operations performed by the above-mentioned first target object group and the above-mentioned second target object group based on the multimedia data or interactive interface displayed by the above-mentioned application client, and obtain the business interaction data fed back by the above-mentioned target object groups based on the interactive operations. Since different target object groups display different business data on the application client of their associated target applications (for example, the displayed multimedia data push order is different, or different user interaction interfaces are displayed), different interactive operations performed by each target object in the different target object groups based on the multimedia data or interactive interface displayed by the above-mentioned application client can be obtained. Based on the above-mentioned interactive operations, the business interaction data fed back by each target object group can be obtained to determine the target push configuration parameters based on the business interaction data. For example, taking the example that the business data includes multimedia data, the push order of each multimedia data displayed by the application client of the target application associated with each first target object in the above-mentioned first target object group can be multimedia data 10, multimedia data 20 and multimedia data 30 (see Figure 2 In the middle interface 1), the order of pushing each multimedia data displayed by the application client of the target application associated with each second target object in the second target object group may be multimedia data 20, multimedia data 10 and multimedia data 30 (see Figure 2 Middle interface 2). The terminal device can obtain different interactive operations (such as clicking to browse, triggering a praise function, etc.) of each target object in different target object groups based on the multimedia data displayed by the application client, and based on the above interactive operations, the business interaction data fed back by each target object group can be obtained to determine (for example, through AB testing) the target push configuration parameters based on the business interaction data.

[0070] S103, obtaining a target object data set based on the business interaction data, wherein the target object data set includes target object data associated with each target object in each target object group.

[0071] In some feasible implementations, the terminal device may obtain the object data in the current statistical period based on the above-mentioned business interaction data, and obtain the object data set of the previous statistical period of the current statistical period based on the above-mentioned business interaction data. Specifically, the above-mentioned statistical periods (including the current statistical period and the previous statistical period of the current statistical period) may be one day, one week, one month, etc., which may be determined according to the actual application scenario, and the embodiment of the present application is not limited here. The following will take the statistical period of one day as an example for illustration, and will not be repeated below. The object data in the above-mentioned current statistical period may be the object data corresponding to the newly added target object (or the target object hit on this day) within the day corresponding to the current statistical period, and the object data set of the previous statistical period of the above-mentioned current statistical period may be the object data set corresponding to all target objects before the day corresponding to the current statistical period. The object data of any target object (which may be the above-mentioned first target object or the above-mentioned second target object) may include the object attribute information of the target object and at least one of the statistical periods (which may include the first statistical period), where the above-mentioned object attribute information may include the object identity, experiment identification and group identification of any target object. The above-mentioned experiment identifier is used to indicate the type of experiment associated with any target object (for example, it may be an experiment whose corresponding business indicator is click-through rate, or an experiment whose corresponding business indicator is browsing time, etc.), and the above-mentioned group identifier is used to indicate the strategy corresponding to any target object (for example, a business data push strategy based on the first push configuration parameter or based on the second push configuration parameter). The terminal device can obtain the object data set of the current statistical period as the target object data set based on the object data in the current statistical period and the above-mentioned object data set, and the target object data set includes the target object data corresponding to each target object in the current statistical period (which may include the object attribute information of the target object and at least one item in the statistical period in which it occurs), that is, all target object data participating in the AB test in the current statistical period, so as to generate the cumulative business indicator data of each target object for determining the target push configuration parameters based on the target object data set and the target business indicator data set. Specifically, the terminal device may compare the object data in the current statistical period with the above-mentioned object data set (of the previous statistical period of the current statistical period) to select object data that is not repeated with the object data included in the above-mentioned object data set from the object data in the current statistical period as the newly added object data. Here, the newly added object data may be represented as the object data corresponding to the target object that is participating in the AB test (may be an experiment in the AB test) in the current statistical period and whose first statistical period is the current statistical period. The newly added object data is added to the above-mentioned object data set to obtain the above-mentioned object data set for the current statistical period.The terminal device can also detect whether the experiment corresponding to the experimental identifier associated with the object data corresponding to any target object in the above object data set is ongoing in the current statistical period. If the experiment corresponding to the experimental identifier associated with the object data has ended in the current statistical period (or the object data is invalid), the object data is removed. By adding non-repetitive object data and removing invalid object data, a target object data set that more accurately reflects the participation of the target object in the AB test in the current statistical period can be obtained. The target object data set is highly efficient in acquisition, and the cumulative business indicator data of each target object for determining the target push configuration parameters can be generated based on the above target object data set and the target business indicator data set.

[0072] In some feasible implementations, the process in which the terminal device obtains the object data set of the current statistical period as the target object data set based on the object data in the current statistical period and the object data set of the previous statistical period of the current statistical period can also be manifested as generating a cumulative hit table in the current statistical period based on the hit table in the current statistical period and the cumulative hit table of the previous statistical period, and the above-mentioned cumulative hit table in the current statistical period corresponds to the above-mentioned target object data set. Specifically, the above-mentioned hit table fields can be: date, object identity, experiment identifier, and group identifier; the cumulative hit table fields can be: date, object identity, experiment identifier, group identifier, and the statistical period of first occurrence. Any object data in the hit table within the current statistical period that does not appear in the cumulative hit table of the previous statistical period (which may include the date, object identity, experiment identifier and group identifier of the target object) can be added to the cumulative hit table of the previous day, and the statistical period corresponding to the first appearance of any of the above object data in the cumulative hit table can be set as the current statistical period, and the object data in the cumulative hit table of the previous day (experiments associated with the experiment identifier corresponding to any object data) that does not belong to the current statistical period and is still in progress is deleted to obtain the cumulative hit table within the above-mentioned current statistical period (that is, the above-mentioned target object data set). The operation is simple and the target object data set acquisition efficiency is high.

[0073] S104, obtaining a target business indicator data set based on the business interaction data, wherein the target business indicator data set includes a plurality of business indicator data of each target object.

[0074] In some feasible implementations, the terminal device can obtain a target business indicator data set based on the above-mentioned business interaction data, and the above-mentioned business indicator data set can include multiple business indicator data of each target object. Here, the nth business indicator data in the multiple business indicator data of any target object is the sum of the business indicator data of any target object from the first statistical period to the nth statistical period in multiple consecutive statistical periods, and n is not greater than the number of statistical periods in the above-mentioned multiple statistical periods. By obtaining multiple business indicator data of each target object, the terminal device can quickly obtain the cumulative business indicator data of each target object based on the multiple business indicator data of each target object to select the target push configuration parameters. Since the statistical period corresponding to the first appearance of each target object and the duration of participating in the AB test may be different, the cumulative business indicator data of each target object in the above-mentioned multiple consecutive statistical periods can be uniformly obtained by determining multiple consecutive statistical periods to improve the efficiency of obtaining the cumulative business indicator data, so as to determine the target push configuration parameters based on the above-mentioned cumulative business indicator data through AB testing (such as AB testing based on a quantile test model) to optimize business data push. Specifically, by analyzing the duration of each experiment (the experiments corresponding to each experimental identifier), it can be ensured that the duration of 99% of the experiments will not exceed the number of statistical periods corresponding to the above-mentioned multiple consecutive statistical periods. Since the above-mentioned multiple consecutive statistical periods contain more statistical periods, only obtaining the cumulative business indicator data of each target object in the above-mentioned multiple consecutive statistical periods will not lead to erroneous judgment on the result analysis of the final quantile test.

[0075] Further, the terminal device may obtain the business indicator data (which may be referred to as the first business indicator data) of any target object in the current statistical period based on the above-mentioned business interaction data (which may include at least one of the number of clicks and browsing time of the target object based on the above-mentioned pushed business data in the current statistical period), and the above-mentioned first business indicator data may be the business indicator data of each target object in the current statistical period (for example, the click rate and browsing time of the target object based on the above-mentioned pushed business data in the current statistical period). The terminal device obtains multiple business indicator data (which may be referred to as the second business indicator data) obtained by any target object in the previous statistical period of the current statistical period based on the above-mentioned business interaction data, thereby obtaining multiple business indicator data of any target object in the current statistical period based on the above-mentioned first business indicator data and the above-mentioned multiple second business indicator data, so as to obtain the above-mentioned target business indicator data set through the multiple business indicator data of each target object in the above-mentioned current statistical period. Since the nth business indicator data among the multiple business indicator data of any target object is the sum of the business indicator data of the target object from the 1st statistical period to the nth statistical period in multiple consecutive statistical periods, the cumulative business indicator data of any target object (in multiple consecutive statistical periods) can be quickly obtained based on the above-mentioned target business indicator data set combined with the above-mentioned target object data set, and the cumulative business indicator data is obtained accurately.

[0076] In some feasible implementations, the above-mentioned multiple continuous statistical periods may include M statistical periods, and the terminal device may subtract the first second business indicator data in the above-mentioned multiple second business indicator data from each second business indicator data in the above-mentioned multiple second business indicator data to obtain the M-1 business indicator data of the above-mentioned any target object in the current statistical period (i.e., only retain the last M-1 business indicator data of any target object). Based on the above-mentioned first business indicator data and the M-1th business indicator data in the above-mentioned M-1 business indicator data, the Mth business indicator data of the above-mentioned any target object in the above-mentioned current statistical period (i.e., the last business indicator data of any target object in the current statistical period) is generated to obtain the M business indicator data of the above-mentioned any target object in the above-mentioned current statistical period. It can be understood that the process of the terminal device obtaining multiple business indicator data of any target object in the above-mentioned current statistical period based on the first business indicator data and multiple second business indicator data to obtain the target business indicator data set can also be expressed as generating a score table in the current statistical period based on the business table in the current statistical period and the score table of the previous statistical period, and the score table in the above-mentioned current statistical period corresponds to the target business indicator data set. Taking the statistical period as one day as an example, assuming that the above multiple consecutive statistical periods include T max Days, that is, the duration of the above multiple consecutive statistical periods is T maxDay, among the multiple second business indicator data acquired by the terminal device, each target object has T max The previous statistical period of the current statistical period (that is, yesterday) is represented by T, and the current statistical period (that is, today) is represented by T+1. Indicates the mth business indicator data of any target object in the statistical period T (including T max , which can be expressed as ), here, since statistical period T is the previous statistical period of the current statistical period T+1, the business indicator data of any target object in statistical period T is called second business indicator data to distinguish it from the business indicator data of any target object in the current statistical period. A score table in statistical period T (i.e., the previous statistical period) can be obtained, and the score table may include T max +2 columns, the points table fields can be: date (that is, T), object identity, here, Indicates that any target object in the statistical period T is max -1) Business indicator data for the day, Indicates that any target object in the statistical period T is max -1) to T-(T max -1) The sum of the business indicator data for two days, Indicates that any target object in the statistical period T is max -1) to T-(T max =The sum of the business indicator data of m days in a statistical period T, that is, in the statistical period T, the mth second business indicator data among the multiple second business indicator data of any target object is the sum of the business indicator data of the target object from the first statistical period to the mth statistical period in multiple consecutive statistical periods. In the statistical period T, the mth second business indicator data among the multiple second business indicator data of any target object is expressed as The business indicator data of any of the above target objects on date d can be expressed as K d , then The expression is:

[0077]

[0078] Subtract the points table field from each column of the points table corresponding to the statistical period T The corresponding column of data (i.e., the corresponding second business indicator data in the plurality of second business indicator data minus the first second business indicator data in the plurality of second business indicator data), and then each column is shifted to the left by one column and the last column is copied, then the original score table becomes From the business table corresponding to the statistical period T+1, the first business indicator data of any target object in the statistical period T+1 can be obtained, which can be expressed as K T+1 , based on the above statistical period T corresponding to the integral table and K T+1 Generate a score table corresponding to the statistical period T+1 (i.e., the score table in the current statistical period, which may correspond to the above-mentioned corresponding target business indicator data set), which may include T max +2 columns, the points table fields can be: date (that is, T+1), object identity, The points table for each statistical period (such as every day) can be obtained by calculating the points table of the previous statistical period and the business table of the current statistical period. Each row of the points table of the current statistical period contains multiple business indicator data of a target object. Since the calculation process does not involve associated operations, the calculation efficiency is very high, and the target business indicator data set of the current statistical period is acquired efficiently.

[0079] It can be understood that the mth column in the above-mentioned score table in the current statistical period can be expressed as:

[0080]

[0081] Substituting this formula into the above The expression can be obtained The expression is:

[0082]

[0083] as well as The expression is:

[0084]

[0085] Here, when m is T max hour, It can be expressed as:

[0086]

[0087] This can be used to obtain the current statistical period The expression is:

[0088]

[0089] Will Replace T with T+1 in the expression of This expression also corresponds to the business indicator data of the score table corresponding to the current statistical period. Please also refer to Figure 4 , Figure 4This is a schematic diagram of generating a target business indicator data set provided by an embodiment of the present application. Figure 4 As shown, Figure 4 The score table 1 in the previous statistical period may include a plurality of corresponding business indicator data contents (i.e., second business indicator data) in the score table, and tm (which may include t1, t2, ..., tmax in total) is used. max ) represents the mth second business indicator data of any target object in the previous statistical period. Subtract the first second business indicator data of the plurality of second business indicator data from each second business indicator data in the score table 1 to obtain the score table 2, and the last T of any target object in the score table 2 is max -1 second service indicator data (may include t2-t1, t3-t1, ..., tmax-t1, a total of T max -1) as any of the above target objects T max -1 business indicator data. From the business table of the current statistical period ( Figure 4 (not shown) obtain the first business indicator data K of any target object in the current statistical period, based on the first business indicator data and the T max -1 business indicator data in the Tth max -1 business indicator data tmax-t1 generates the Tth max business indicator data tmax-t1+K, thus obtaining T of any of the above target objects max business indicator data, T of any of the above target objects max Each business indicator data may correspond to the score table 3, so as to obtain the target business indicator data set based on the business indicator data in the score table 3. Therefore, the cumulative business indicator data of any target object (in multiple consecutive statistical periods) can be quickly obtained based on the above target business indicator data set combined with the above target object data set, and the cumulative business indicator data is obtained accurately. In addition, if all the multiple business indicator data corresponding to a target object are equal to 0, then the target object can be deleted to further reduce the data magnitude.

[0090] In some feasible implementations, after receiving the target object data set and the target object data set from the first target object group and the second target object group, the terminal device can also regularly store the target object data set and the target object data set in a designated storage space of the terminal device, such as a database of the terminal device, to facilitate data management and maintenance.

[0091] S105, obtaining cumulative business indicator data of each target object based on the target object data set and the business indicator data set, and determining the cumulative business indicator data as input data of the quantile test model.

[0092] In some feasible implementations, the terminal device can determine the statistical period in which any target object first appears based on the target object data associated with any target object in the target object data set. When the statistical period is one of the above-mentioned multiple continuous statistical periods, that is, any of the above-mentioned target objects is hit and participates in the AB test (that is, participates in the determination process of the above-mentioned target push configuration parameters) in the above-mentioned multiple continuous statistical periods, the terminal device can obtain the sum of the business indicator data of each statistical period from the above-mentioned first appearance statistical period to the above-mentioned current statistical period based on the above-mentioned target business indicator data set as the cumulative business indicator data of any of the above-mentioned target objects. Otherwise, if the statistical period in which any of the above-mentioned target objects first appears is not one of the above-mentioned multiple continuous statistical periods, that is, any of the above-mentioned target objects is hit and participates in the AB test before the above-mentioned multiple continuous statistical periods, the terminal device can obtain the sum of the business indicator data of each statistical period in the above-mentioned multiple continuous statistical periods of any of the above-mentioned target objects based on the above-mentioned target business indicator data set as the cumulative business indicator data of any of the above-mentioned target objects, thereby obtaining the cumulative business indicator data of each target object. By obtaining the cumulative business indicator data of each target object based on the target object data set and the target business indicator data set, the cumulative business indicator data of each target object can be obtained in a shorter time, and there is no need to roll back a large number of historical business tables. In this way, the target push configuration parameters can be determined more quickly through AB testing and business data can be pushed based on the target push configuration parameters. The business data push effect is good and the applicability is strong.

[0093] In some feasible implementations, it is understandable that the process of the terminal device obtaining the cumulative business indicator data of each target object based on the target object data set and the business indicator data set can also be expressed as generating the cumulative business indicator data of each target object by associating the business table in the current statistical period and the cumulative hit table (corresponding to the above-mentioned five-tuple field table, which can also be called an association table). Taking the statistical period as one day as an example, assuming that the current statistical period (that day) is represented by T, the above-mentioned multiple consecutive statistical periods include T max God, use D f K represents the statistical period in which any target object first appears in the statistical period T. The business indicator data of any target object on date d can be expressed as K d , then when the statistical period of the first appearance of any of the above target objects is one of the above multiple consecutive statistical periods, it can be expressed as:

[0094] D f ≥T-(T max -1)

[0095] Then the sum of the business indicator data of any of the above target objects in each statistical period from the above first statistical period to the above current statistical period can be expressed as:

[0096]

[0097] Here, based on the above The expressions are:

[0098]

[0099] Simplifying this expression yields:

[0100]

[0101] That is, we can get:

[0102]

[0103] When the statistical period in which any of the above target objects first appears does not belong to one of the above multiple consecutive statistical periods, that is:

[0104] D f <T-(T max -1)

[0105] Then the sum of the business indicator data of any of the above target objects in each of the above multiple consecutive statistical periods can be expressed as:

[0106]

[0107] In summary, we can obtain an expression for associating the cumulative hit table (i.e., corresponding to the above target object data set) and the score table (i.e., corresponding to the target business indicator data set) in the current statistical period to obtain the cumulative business indicator data of each target object:

[0108]

[0109] Here, we can define that when m is not greater than 0, If the value of is 0, the above formula can be simplified to:

[0110]

[0111] Among them, TD fIndicates the difference (i.e., time difference) between the statistical period of the first occurrence and the current statistical period. By associating the cumulative hit table (i.e., corresponding to the above target object data set) and the points table (i.e., corresponding to the target business indicator data set) in the current statistical period, the cumulative business indicator data of each target object is obtained, that is, an operation that requires cumulative rolling association of a large number of data tables is converted into a fast operation of accumulation of the two tables and a table lookup operation for the points table. The acquisition of the cumulative business indicator data of each target object can be completed in a shorter time, and there is no need to roll back a large number of historical business tables, so that the target push configuration parameters can be determined more quickly through AB testing and business data push is performed based on the target push configuration parameters. The business data push effect is good and the applicability is strong.

[0112] In some feasible implementations, the terminal device may input the accumulated business indicator data into a quantile test model (such as a quantile test model based on the ConQ algorithm), and the quantile test model may be based on the assumption that the first push configuration parameter is the target push configuration parameter (or, the first push configuration parameter is better than the second push configuration parameter), and the confidence score of the assumption is obtained through the quantile test model. If the confidence score belongs to the preset rejection domain, the original hypothesis is rejected and the second push configuration parameter is used as the target push configuration parameter. Otherwise, if the confidence score does not belong to the preset rejection domain, the original hypothesis is accepted and the first push configuration parameter is used as the target push configuration parameter. Here, if it is determined that the first push configuration parameter is the target push configuration parameter, business data can be pushed to the object associated with the target application based on the first push configuration parameter. In other words, the push strategy for pushing business data based on the first push configuration parameter is applied to all objects associated with the target application, thereby further enhancing the business data push effect.

[0113] See also Figure 5 , Figure 5 is a schematic diagram of generating cumulative business indicator data provided by an embodiment of the present application, such as Figure 5As shown, based on the above-mentioned cumulative hit table 1 and hit table, cumulative hit table 2 can be obtained. Cumulative hit table 1 can be a cumulative hit table of the previous statistical period, and cumulative hit table 1 can include: date, object identity, experiment identifier, group identifier and the statistical period of first appearance, and hit table can be a hit table in the current statistical period, and hit table can include: date, object identity, experiment identifier and group identifier. Cumulative hit table 2 can be a cumulative hit table in the current statistical period, and cumulative hit table 2 can include: date, object identity, experiment identifier, group identifier and the statistical period of first appearance. Points table 2 can be obtained based on points table 1 and business table. The above-mentioned business table can be a business table in the current statistical period (which can be expressed as T), and points table 1 can be a points table of the previous statistical period. Taking the statistical period of one day as an example, assuming that the above-mentioned multiple consecutive statistical periods include P max days, the score table 1 includes: date, object identity, t1, t2, ..., t max , where t1 to t max Any one of t m (which can be called the second business indicator data) indicates that any target object in the statistical period T-1 is within T-1-(P max -1) to T-1-(P max -m) the sum of the business indicator data for a total of m days, where m is not greater than the number of statistical periods included in the above-mentioned multiple consecutive statistical periods. The business table may be a business table within the current statistical period, and the business table includes: date, object identity, and the first business indicator. The score table 2 may be a score table within the current statistical period, and the score table 2 includes: date, object identity, T1, T2, ..., T max , where T1 to T max Any one of T m Indicates that any target object in the statistical period T is max -1) to T-(P max-m) the sum of the business indicator data for a total of m days. The object identity, experiment identifier and group identifier corresponding to each target object in the cumulative hit table 2 are extracted to obtain a triple (object identity, experiment identifier, statistical period of first appearance) to generate a cumulative hit table 3. Based on the above-mentioned score table 2, the above-mentioned cumulative hit table 3 is associated to obtain an association table, and the association table may include: date, object identity, experiment identifier, group identifier and cumulative business indicators. The cumulative business indicator data of each target object can be obtained from the above-mentioned association table and the above-mentioned cumulative business indicator data can be input into the quantile test model for AB testing. The above-mentioned cumulative hit table 2 (which can correspond to the above-mentioned target object data set) is based on the cumulative hit table 1 (i.e., the cumulative hit table of the previous day) and the hit table generation. The order of magnitude processed by the terminal device can be hundreds of billions and hundreds of billions of orders of magnitude splicing. Since no association operation is involved, the calculation speed can also be kept within a reasonable range (for example, the calculation is completed within 1 hour), and the result is kept at hundreds of billions. The target object data set corresponding to the cumulative hit table 2 is highly efficient. The above-mentioned score table 2 (which can correspond to the above-mentioned target business indicator data set) is generated based on score table 1 (i.e., the score table of the previous day) and the business table. The order of magnitude processed by the terminal device can be a splicing of billions and billions, and the result can also be maintained at the billions, which can ensure the calculation speed in the big data scenario, and the target business indicator data set corresponding to score table 2 is acquired efficiently. Since the cumulative hit table 2 of the order of hundreds of billions and the score table 2 of the order of billions cannot be directly associated, the cumulative hit table 2 is processed into the cumulative hit table 3, that is, the target object in the cumulative hit table 2 is used as the key value (key), and the triples (object identity identifier, experimental identifier, and statistical period of first appearance) of each target object are collected, and the cumulative hit table 2 is converted into the cumulative hit table 3, and then associated with the score table 2 of the order of billions. Since the same target object in the cumulative hit table 3, no matter which experiment it corresponds to, is the same piece of data (including multiple business indicator data) in the associated score table 2, and since the cumulative hit table 3 and the score table 2 are keyed by the target object, there is no data skew, so as long as the terminal device has sufficient memory resources, it can be easily calculated. After the cumulative hit table 3 is associated based on the score table 2, the triples of each target object are expanded, and each triple can be quickly queried in the score table 2 to get the required results (the accumulated business indicator data of each target object). The operation is simple, the accumulated business indicator data of each target object is acquired efficiently, and the applicability is strong.

[0114] In the method provided in the embodiment of the present application, the terminal device can push business data to the first target object group based on the first push configuration parameter, and the terminal device can also push multiple target multimedia data to the second target object group other than the first target object group among the multiple target objects of the target application through the algorithm model configured with the second push configuration parameter, so as to determine the target push configuration parameter with the best business data push effect by pushing business data to different target object groups. The terminal device can obtain the interactive operation performed by the first target object group and the second target object group based on the multimedia data or interactive interface displayed by the application client, and obtain the business interaction data fed back by the target object groups based on the interactive operation. Since the business data displayed by different target object groups on the application client of the target application associated with them is different (for example, the multimedia data push order displayed is different, or different user interaction interfaces are displayed), the different interactive operations performed by each target object in the different target object groups based on the multimedia data or interactive interface displayed by the application client can be obtained, and the business interaction data fed back by each target object group can be obtained based on the interactive operation to determine the target push configuration parameter based on the business interaction data. The terminal device can obtain the object data in the current statistical period based on the above-mentioned business interaction data, and obtain the object data set of the previous statistical period of the current statistical period based on the above-mentioned business interaction data. The terminal device can obtain the object data set of the current statistical period as the target object data set based on the above-mentioned object data in the current statistical period and the above-mentioned object data set, and the target object data set includes the target object data corresponding to each target object in the current statistical period, that is, all target object data participating in the AB test (to determine the target push configuration parameters) in the current statistical period, so as to generate the cumulative business indicator data of each target object for determining the target push configuration parameters based on the target object data set and the target business indicator data set. The terminal device can also obtain the target business indicator data set based on the above-mentioned business interaction data, and the above-mentioned business indicator data set may include multiple business indicator data of each target object, where the nth business indicator data in the multiple business indicator data of any target object is the sum of the business indicator data of any target object from the first statistical period to the nth statistical period in multiple consecutive statistical periods, and n is not greater than the number of statistical periods in the above-mentioned multiple statistical periods. By obtaining multiple business indicator data of each target object, the terminal device can quickly obtain the cumulative business indicator data of each target object based on the multiple business indicator data of each target object to select the target push configuration parameters. The terminal device can determine the statistical period in which any target object first appears based on the target object data associated with any target object in the above target object data set.When the above-mentioned statistical period is one of the above-mentioned multiple continuous statistical periods, that is, any of the above-mentioned target objects is hit in the above-mentioned multiple continuous statistical periods and participates in the AB test (that is, participates in the determination process of the above-mentioned target push configuration parameters), the terminal device can obtain the sum of the business indicator data of each statistical period from the above-mentioned first appearance statistical period to the above-mentioned current statistical period based on the above-mentioned target business indicator data set as the cumulative business indicator data of any of the above-mentioned target objects. Otherwise, the terminal device obtains the sum of the business indicator data of each statistical period of any of the above-mentioned target objects in the above-mentioned multiple continuous statistical periods as the cumulative business indicator data of any of the above-mentioned target objects, thereby obtaining the cumulative business indicator data of each target object. By obtaining the cumulative business indicator data of each target object based on the target object data set and the target business indicator data set, the acquisition of the cumulative business indicator data of each target object can be completed in a shorter time. The terminal device can also input the above-mentioned accumulated business indicator data into a quantile test model (such as a quantile test model based on the ConQ algorithm), which can be based on the assumption that the above-mentioned first push configuration parameter is the target push configuration parameter (or, the above-mentioned first push configuration parameter is better than the above-mentioned second push configuration parameter), and obtain the confidence score of the above-mentioned assumption through the quantile test model to determine the target push configuration parameter based on the confidence score. Therefore, the push strategy for business data push based on the above-mentioned first push configuration parameter can be applied to all objects associated with the above-mentioned target application, and the target push configuration parameter confirmation efficiency is high, the business data push effect is good, and the applicability is strong.

[0115] Based on the description of the above-mentioned business data push method embodiment, the present application embodiment also discloses a business data push device. The business data push device can be applied to Figures 3 to 5 In the method for pushing service data of the embodiment shown, the steps in the method for pushing service data are performed. Here, the device for pushing service data may be the above-mentioned Figures 3 to 5 The service server or terminal device in the embodiment shown, that is, the push device of the service data can be the above-mentioned Figures 3 to 5 The execution subject of the method for pushing business data in the embodiment shown. Figure 6 , Figure 6 : is a schematic diagram of the structure of the device for pushing business data provided in the embodiment of the present application. In the embodiment of the present application, the device can run the following modules:

[0116] The business data push module 41 is used to push the business data to the first target object group based on the first push configuration parameter when receiving the first push configuration parameter, and push the business data to the second target object group based on the determined second push configuration parameter;

[0117] An acquisition module 42 is used to acquire the business interaction data fed back by the first target object group and the second target object group;

[0118] The cumulative business indicator data generating module 43 is used to obtain a target object data set based on the business interaction data, obtain a target business indicator data set based on the business interaction data, and obtain the cumulative business indicator data of each target object based on the target object data set and the business indicator data set;

[0119] The target push configuration parameter determination module 44 is used to determine the above-mentioned accumulated business indicator data as input data of the quantile test model, so as to obtain the confidence score of the above-mentioned first push configuration parameter as the target push configuration parameter through the above-mentioned quantile test model.

[0120] In some feasible implementations, the above-mentioned cumulative business indicator data generation module 43 is also used to:

[0121] Acquire object data in a current statistical period based on the business interaction data, and acquire a set of object data in a previous statistical period of the current statistical period based on the business interaction data;

[0122] Based on the object data in the current statistical period and the object data set, the object data set of the current statistical period is obtained as the target object data set;

[0123] The object data set in any statistical period includes object data of one or more target objects in each target object group, and the object data includes object attribute information or at least one item in the statistical period in which the object appears.

[0124] In some feasible implementations, the above-mentioned cumulative business indicator data generation module 43 is also used to:

[0125] Acquire the first business indicator data of any of the target objects in the current statistical period based on the business interaction data, and acquire multiple second business indicator data of any of the target objects in a statistical period before the current statistical period based on the business interaction data;

[0126] Based on the first business indicator data and the plurality of second business indicator data, obtaining a plurality of business indicator data of any one of the target objects in the current statistical period;

[0127] Obtaining the target business indicator data set based on the multiple business indicator data of the target objects in the current statistical period;

[0128] The business interaction data include at least one of the number of clicks or browsing time, and the business indicator data include at least one of the click rate or browsing time.

[0129] In some feasible implementations, the above-mentioned cumulative business indicator data generation module 43 is also used to:

[0130] Select object data that is not repeated in the object data set from the object data in the current statistical period as newly added object data, and add the newly added object data to the object data set to obtain the object data set for the current statistical period.

[0131] In some feasible implementations, the target business indicator data set includes M business indicator data of any target object, and the accumulated business indicator data generating module 43 is further used to:

[0132] Subtract the first second business indicator data from each second business indicator data in the plurality of second business indicator data to obtain M-1 business indicator data of any target object in the current statistical period;

[0133] Based on the above-mentioned first business indicator data and the M-1th business indicator data among the above-mentioned M-1 business indicator data, the Mth business indicator data of the above-mentioned any target object in the above-mentioned current statistical period is generated to obtain the M business indicator data of the above-mentioned any target object in the above-mentioned current statistical period.

[0134] In some feasible implementations, the above-mentioned cumulative business indicator data generation module 43 is also used to:

[0135] Determine the statistical period in which any of the target objects first appears based on the target object data associated with any of the target objects in the target object data set;

[0136] When the statistical period in which any of the target objects first appears is one of the multiple consecutive statistical periods, the sum of the business indicator data of any of the target objects in the business indicator data set in each statistical period between the statistical period in which any of the target objects first appears and the current statistical period is determined as the cumulative business indicator data of any of the target objects;

[0137] When the statistical period in which any of the above-mentioned target objects first appears is not one of the above-mentioned multiple consecutive statistical periods, the sum of the business indicator data of any of the above-mentioned target objects in each statistical period in the above-mentioned multiple consecutive statistical periods in the above-mentioned business indicator data set shall be determined as the cumulative business indicator data of any of the above-mentioned target objects to obtain the cumulative business indicator data of each of the above-mentioned target objects.

[0138] In some feasible implementations, the target push configuration parameter determination module 44 is further used to:

[0139] Input the accumulated business indicator data into the quantile test model, and obtain the confidence score of the first push configuration parameter as the target push configuration parameter through the quantile test model;

[0140] If the confidence score belongs to the preset rejection domain, the second push configuration parameter is used as the target push configuration parameter; if the confidence score does not belong to the preset rejection domain, the first push configuration parameter is used as the target push configuration parameter.

[0141] In some feasible implementations, the business data includes multimedia data or interactive interface data, the first target object group includes multiple first target objects among all target objects of the target application, the second target object group includes multiple second target objects among all target objects except the multiple first target objects, and the business data push module 41 is further used to:

[0142] Based on the above-mentioned first push configuration parameters, multimedia data or interactive interface data are pushed to the above-mentioned multiple first target objects, and based on the above-mentioned second push configuration parameters, the above-mentioned multimedia data or interactive interface data are pushed to the above-mentioned multiple second target objects, so that the above-mentioned multimedia data or the interactive interface corresponding to the above-mentioned interactive interface data can be displayed through the application client of the above-mentioned target application associated with the above-mentioned target objects.

[0143] In some feasible implementations, the acquisition module 42 is further used to:

[0144] The interactive operations performed by each target object in the first target object group and the second target object group based on the multimedia data displayed by the application client or the interactive interface are obtained, and the business interaction data fed back by each target object group is obtained based on the interactive operations.

[0145] According to the above Figure 3 The corresponding embodiment, Figure 3 The implementation method described in steps S101 to S105 in the method for pushing service data shown in FIG. Figure 5 Each module of the device shown in the figure is executed. For example, the above Figure 3 The implementation method described in step S101 of the service data push method shown in FIG. Figure 5The implementation method described in step S102 can be performed by the business data push module 41 in the device shown in the figure, the implementation method described in step S103 and step S104 can be performed by the cumulative business indicator data generation module 43, and the implementation method described in step S105 can be performed by the cumulative business indicator data generation module 43 and the target push configuration parameter determination module 44. The implementation methods performed by the above-mentioned business data push module 41, the acquisition module 42, the cumulative business indicator data generation module 43 and the target push configuration parameter determination module 44 can refer to the above-mentioned Figure 3 The implementation methods provided in each step of the corresponding embodiment will not be repeated here.

[0146] In an embodiment of the present application, the push of business data can push business data to the first target object group based on the first push configuration parameter, and the push of business data can also push multiple target multimedia data to the second target object group other than the first target object group among the multiple target objects of the target application through an algorithm model configured with a second push configuration parameter, so as to determine the target push configuration parameter with the best business data push effect by pushing business data to different target object groups. The push of business data can obtain the interactive operation performed by the first target object group and the second target object group based on the multimedia data or interactive interface displayed by the application client, and obtain the business interaction data fed back by the target object groups based on the interactive operation. Since different target object groups display different business data on the application client of their associated target application (for example, the displayed multimedia data push order is different, or different user interaction interfaces are displayed), different interactive operations performed by each target object in different target object groups based on the multimedia data or interactive interface displayed by the application client can be obtained, and the business interaction data fed back by each target object group can be obtained based on the interactive operation to determine the target push configuration parameter based on the business interaction data. The push of business data can obtain the object data in the current statistical period based on the above business interaction data, and obtain the object data set of the previous statistical period of the current statistical period based on the above business interaction data. The push of business data can obtain the object data set of the current statistical period as the target object data set based on the object data in the current statistical period and the above object data set, and the target object data set includes the target object data corresponding to each target object in the current statistical period, that is, all target object data participating in the AB test (to determine the target push configuration parameters) in the current statistical period, so as to generate the cumulative business indicator data of each target object for determining the target push configuration parameters based on the target object data set and the target business indicator data set. The push of business data can also obtain the target business indicator data set based on the above business interaction data, and the above business indicator data set can include multiple business indicator data of each target object, where the nth business indicator data in the multiple business indicator data of any target object is the sum of the business indicator data of any target object from the first statistical period to the nth statistical period in multiple consecutive statistical periods, and n is not greater than the number of statistical periods in the above multiple statistical periods. By obtaining multiple business indicator data of each target object, the push of business data can quickly obtain the cumulative business indicator data of each target object based on the multiple business indicator data of each target object to select the target push configuration parameters. The push of business data can determine the statistical period of the first appearance of any target object based on the target object data associated with any target object in the above target object data set.When the above-mentioned statistical period is one of the above-mentioned multiple continuous statistical periods, that is, any of the above-mentioned target objects is hit in the above-mentioned multiple continuous statistical periods and participates in the AB test (that is, participates in the determination process of the above-mentioned target push configuration parameters), the push of business data can obtain the sum of the business indicator data of each statistical period from the above-mentioned first appearance statistical period to the above-mentioned current statistical period based on the above-mentioned target business indicator data set, as the cumulative business indicator data of any of the above-mentioned target objects. Otherwise, the push of business data obtains the sum of the business indicator data of each statistical period of any of the above-mentioned target objects in the above-mentioned multiple continuous statistical periods, as the cumulative business indicator data of any of the above-mentioned target objects, thereby obtaining the cumulative business indicator data of each target object. By obtaining the cumulative business indicator data of each target object based on the target object data set and the target business indicator data set, the acquisition of the cumulative business indicator data of each target object can be completed in a shorter time. The push of business data can also input the above-mentioned accumulated business indicator data into a quantile test model (such as a quantile test model based on the ConQ algorithm). The quantile test model can be based on the assumption that the above-mentioned first push configuration parameter is the target push configuration parameter (or, the above-mentioned first push configuration parameter is better than the above-mentioned second push configuration parameter). The confidence score of the above-mentioned assumption is obtained through the quantile test model to determine the target push configuration parameter based on the confidence score. Therefore, the push strategy for business data push based on the above-mentioned first push configuration parameter can be applied to all objects associated with the above-mentioned target application, and the target push configuration parameter confirmation efficiency is high, the business data push effect is good, and the applicability is strong.

[0147] In the embodiment of the present application, the above Figure 6 Each module in the device shown can be separately or all combined into one or several other modules to constitute, or one (some) of the modules can also be split into multiple smaller modules in function to constitute, which can achieve the same operation without affecting the realization of the technical effects of the embodiments of the present application. The above modules are divided based on logical functions. In practical applications, the function of a module can also be implemented by multiple modules, or the function of multiple modules can be implemented by one module. In other feasible implementations of the present application, the above device can also include other modules. In practical applications, these functions can also be implemented with the assistance of other modules, and can be implemented by the collaboration of multiple modules, which is not limited here.

[0148] See also Figure 7 , Figure 7 Schematic diagram of the structure of the computer device provided in the embodiment of the present application. Figure 7 As shown, the computer device 1000 may be the above Figures 3 to 6The terminal device in the corresponding embodiment. The computer device 1000 may include: a processor 1001, a network interface 1004 and a memory 1005. In addition, the computer device 1000 may also include: a user interface 1003, and at least one communication bus 1002. The communication bus 1002 is used to realize the connection and communication between these components. The user interface 1003 may include a display screen (Display), a keyboard (Keyboard), and the user interface 1003 may optionally include a standard wired interface and a wireless interface. The network interface 1004 may optionally include a standard wired interface and a wireless interface (such as a WI-FI interface). The memory 1005 may be a high-speed RAM memory, or a non-volatile memory (non-volatile memory), such as at least one disk storage. The memory 1005 may optionally be at least one storage device located away from the aforementioned processor 1001. Figure 7 As shown, the memory 1005 as a computer-readable storage medium may include an operating system, a network communication module, a user interface module, and a device control application program.

[0149] The network interface 1004 in the computer device 1000 can also be connected to the above Figure 3 The terminal device in the corresponding embodiment is connected to the network, and the optional user interface 1003 may also include a display screen (Display) and a keyboard (Keyboard). Figure 7 In the computer device 1000 shown in the figure, the network interface 1004 can provide a network communication function; the user interface 1003 is mainly used to provide an input interface for the user (or developer); and the processor 1001 can be used to call the device control application stored in the memory 1005 to implement the above Figure 3 The method for pushing business data in the corresponding embodiment.

[0150] It should be understood that the computer device 1000 described in the embodiment of the present application can execute the above Figure 3 The description of the method for pushing the service data in the corresponding embodiment will not be repeated here. In addition, the description of the beneficial effects of adopting the same method will not be repeated here.

[0151] In addition, it should be pointed out that the embodiment of the present application further provides a computer-readable storage medium, and the computer-readable storage medium stores a computer program executed by the device for pushing the business data mentioned above, and the computer program includes program instructions. When the processor executes the program instructions, the above-mentioned Figure 3The description of the method for pushing business data in the corresponding embodiment will not be repeated here. In addition, the description of the beneficial effects of adopting the same method will not be repeated. For technical details not disclosed in the computer-readable storage medium embodiment involved in this application, please refer to the description of the method embodiment of this application.

[0152] Those skilled in the art can understand that all or part of the processes in the above-mentioned embodiments can be implemented by instructing the relevant hardware through a computer program, and the above-mentioned program can be stored in a computer-readable storage medium, and when the program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, the above-mentioned storage medium can be a disk, an optical disk, a read-only memory (ROM) or a random access memory (RAM), etc.

[0153] The above disclosure is only the preferred embodiment of the present application, which certainly cannot be used to limit the scope of rights of the present application. Therefore, equivalent changes made according to the claims of the present application are still within the scope covered by the present application.

Claims

1. A method for pushing business data, characterized in that: The method comprises: When the first push configuration parameter is received, the service data is pushed to the first target object group based on the first push configuration parameter, and the service data is pushed to the second target object group based on the determined second push configuration parameter; Acquire business interaction data fed back by the first target object group and the second target object group; Obtaining a target object data set based on the business interaction data, wherein the target object data set includes target object data associated with each target object in each target object group; A target business indicator data set is obtained based on the business interaction data, wherein the target business indicator data set includes a plurality of business indicator data obtained by each target object in a current statistical period, wherein the plurality of business indicator data obtained by any target object in the current statistical period includes M business indicator data sorted in chronological order of the statistical periods in which the business indicator data are obtained, and the nth business indicator data among the M business indicator data is the sum of all business indicator data of n statistical periods from the first statistical period to the nth statistical period in the M consecutive statistical periods of any target object, n is less than or equal to M, and the nth business indicator data is obtained in the nth statistical period; Based on the target object data set and the business indicator data set, the cumulative business indicator data of each target object in the M consecutive statistical periods are obtained, and the cumulative business indicator data are determined as the input data of the quantile test model, so as to obtain the first push configuration parameter as the confidence score of the target push configuration parameter through the quantile test model.

2. The method according to claim 1, characterized in that The obtaining of the target object data set based on the business interaction data comprises: Acquire object data in a current statistical period based on the business interaction data, and acquire a set of object data in a previous statistical period of the current statistical period based on the business interaction data; Based on the object data in the current statistical period and the object data set, obtaining the object data set of the current statistical period as a target object data set; The object data set in any statistical period includes object data of one or more target objects in each target object group, and the object data includes object attribute information or at least one item in the statistical period in which the object appears.

3. The method according to claim 1, characterized in that The obtaining of a target business indicator data set based on the business interaction data includes: Acquire first business indicator data of any target object in the current statistical period based on the business interaction data, and acquire multiple second business indicator data of any target object obtained in a statistical period before the current statistical period based on the business interaction data; Obtaining a plurality of business indicator data of any target object in the current statistical period based on the first business indicator data and the plurality of second business indicator data; Obtaining the target business indicator data set based on the multiple business indicator data of each target object in the current statistical period; The business interaction data includes at least one of the number of clicks or the browsing time, and the business indicator data includes at least one of the click rate or the browsing time.

4. The method according to claim 2, characterized in that: The obtaining of the object data set of the current statistical period based on the object data in the current statistical period and the object data set comprises: Select object data that is not repeated with the object data included in the object data set from the object data in the current statistical period as newly added object data, and add the newly added object data to the object data set to obtain the object data set of the current statistical period.

5. The method according to claim 3, characterized in that: The target business indicator data set includes M business indicator data of any target object, and the multiple business indicator data of any target object in the current statistical period obtained based on the first business indicator data and the multiple second business indicator data include: Subtract the first second business indicator data from each second business indicator data in the plurality of second business indicator data to obtain M-1 business indicator data of any target object in the current statistical period; Based on the first business indicator data and the M-1th business indicator data among the M-1 business indicator data, the Mth business indicator data of any target object in the current statistical period is generated to obtain the M business indicator data of any target object in the current statistical period.

6. The method according to claim 1, characterized in that The obtaining of the accumulated business indicator data of each target object based on the target object data set and the business indicator data set includes: Determine the statistical period in which any target object first appears based on the target object data associated with any target object in the target object data set; When the statistical period in which any target object first appears is one of the multiple consecutive statistical periods, the sum of the business indicator data of any target object in the business indicator data set in each statistical period between the statistical period in which any target object first appears and the current statistical period is determined as the accumulated business indicator data of any target object; When the statistical period in which any target object first appears is not one of the multiple consecutive statistical periods, the sum of the business indicator data of any target object in the business indicator data set in each statistical period in the multiple consecutive statistical periods is determined as the cumulative business indicator data of any target object to obtain the cumulative business indicator data of each target object.

7. The method according to any one of claims 1 to 6, characterized in that: The method further comprises: Inputting the accumulated business indicator data into the quantile test model, and obtaining the confidence score of the first push configuration parameter as the target push configuration parameter through the quantile test model; If the confidence score belongs to the preset rejection domain, the second push configuration parameter is used as the target push configuration parameter; if the confidence score does not belong to the preset rejection domain, the first push configuration parameter is used as the target push configuration parameter.

8. The method according to claim 7, characterized in that The business data includes multimedia data or interactive interface data, the first target object group includes multiple first target objects among all target objects of the target application, and the second target object group includes multiple second target objects among all target objects except the multiple first target objects; pushing the business data to the first target object group based on the first push configuration parameter, and pushing the business data to the second target object group based on the determined second push configuration parameter includes: Pushing multimedia data or interactive interface data to the multiple first target objects based on the first push configuration parameter, and pushing the multimedia data or interactive interface data to the multiple second target objects based on the second push configuration parameter, so as to display the multimedia data or the interactive interface corresponding to the interactive interface data through the application client of the target application associated with each target object; The acquiring of the business interaction data fed back by the first target object group and the second target object group includes: The interactive operation performed by each of the multiple first target objects and the multiple second target objects based on the multimedia data or the interactive interface displayed by the application client is obtained, and the business interaction data fed back by each target object group is obtained based on the interactive operation.

9. A device for pushing business data, characterized in that: include: A business data push module, configured to push business data to a first target object group based on the first push configuration parameter when receiving the first push configuration parameter, and to push the business data to a second target object group based on the determined second push configuration parameter; An acquisition module, configured to acquire business interaction data fed back by the first target object group and the second target object group; a cumulative business indicator data generating module, configured to obtain a target object data set based on the business interaction data, obtain a target business indicator data set based on the business interaction data, wherein the target business indicator data set includes a plurality of business indicator data obtained by each target object in a current statistical period, wherein the plurality of business indicator data obtained by any target object in the current statistical period includes M business indicator data sorted in chronological order of the statistical periods in which the business indicator data are obtained, and the nth business indicator data among the M business indicator data is the sum of all business indicator data of n statistical periods from the first statistical period to the nth statistical period in the consecutive M statistical periods of any target object, n is less than or equal to M, the nth business indicator data is obtained in the nth statistical period, and the cumulative business indicator data of each target object in the consecutive M statistical periods is obtained based on the target object data set and the business indicator data set; The target push configuration parameter determination module is used to determine the accumulated business indicator data as input data of a quantile verification model, so as to obtain the confidence score of the first push configuration parameter as the target push configuration parameter through the quantile verification model.

10. A computer device, characterized in that: include: Processor, memory, and network interface; The processor is connected to the memory and the network interface, wherein the network interface is used to provide a data communication function, the memory is used to store program codes, and the processor is used to call the program codes to execute the method described in any one of claims 1-8.

11. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a computer program, and the computer program is suitable for being loaded by a processor and executing the method according to any one of claims 1 to 8.

Citation Information

Patent Citations

  • Project test selecting method and device

    CN107402881A

  • Configuration data testing method and device of recommendation system and electronic equipment

    CN110995529A