CN116976979A

CN116976979A - Popularization information delivery effect evaluation method and device and computer equipment

Info

Publication number: CN116976979A
Application number: CN202310001498.2A
Authority: CN
Inventors: 钱宇秋; 林总总
Original assignee: Tencent Technology Shenzhen Co Ltd
Current assignee: Tencent Technology Shenzhen Co Ltd
Priority date: 2023-01-03
Filing date: 2023-01-03
Publication date: 2023-10-31

Abstract

The application relates to a popularization information release effect evaluation method, a popularization information release effect evaluation device and computer equipment, which can realize release optimization of popularization information by combining an artificial intelligence technology. The method comprises the following steps: determining an implemented object corresponding to the popularization information after being put in; acquiring actual operation data obtained by operating a first object in the implemented objects based on the popularization information; the operation data obtained by operating the second object based on the promotion information in the implemented objects is unknown; fitting according to actual operation data to obtain a target distribution function corresponding to a target evaluation index; and sampling the target distribution function according to the sampling times, and determining index information of the target evaluation index based on the sampling result. By adopting the method, the accuracy of popularization information delivery effect evaluation can be improved.

Description

Popularization information delivery effect evaluation method and device and computer equipment

Technical Field

The present application relates to the field of computer technologies, and in particular, to a method and apparatus for evaluating a popularization information delivery effect, and a computer device.

Background

With the rapid development of the internet, the number of internet users is increasing, and the internet is an important way for advertisers to deliver advertisements to promote products or services. Advertisers will typically place advertisements on each platform and by evaluating the effectiveness of each large platform advertisement placement, serve as one of the considerations in formulating advertisement placement policies and optimizing traffic. Therefore, it is important for advertisers how to accurately and comprehensively count advertisement effects of advertisements.

Currently, for advertisement effect delivery, click rate is generally counted by whether a user clicks a platform media to play an advertisement, conversion rate is counted by installing APP (Application) corresponding to the advertisement by the user, and the like. However, this approach requires acquisition of a large amount of post-link information for advertising. In the case where a large amount of post-link information for advertisement delivery cannot be obtained, the effect of advertisement delivery cannot be accurately estimated.

Disclosure of Invention

Based on this, it is necessary to provide a popularization information delivery effect evaluation method, apparatus, computer device, computer readable storage medium, and computer program product capable of improving the accuracy of the delivery effect evaluation in view of the above-described technical problems.

On one hand, the application provides a popularization information delivery effect evaluation method. The method comprises the following steps:

determining an implemented object corresponding to the popularization information after being put in; the implemented object includes a first object and a second object;

acquiring actual operation data obtained by the operation of the first object based on the popularization information; wherein operation data obtained by the operation of the second object based on the popularization information is unknown;

Fitting according to the actual operation data to obtain a target distribution function corresponding to a target evaluation index;

acquiring a sampling number, the sampling number being determined based on at least one of the number of implemented objects or the number of second objects;

sampling the target distribution function according to the sampling times, and determining index information of the target evaluation index based on a sampling result; the index information is used for evaluating the release effect of the popularization information.

On the other hand, the application also provides a popularization information delivery effect evaluation device. The device comprises:

the determining module is used for determining an implemented object corresponding to the popularization information after being put in; the implemented object includes a first object and a second object;

the acquisition module is used for acquiring actual operation data obtained by the operation of the first object based on the popularization information; wherein operation data obtained by the operation of the second object based on the popularization information is unknown;

the fitting module is used for fitting according to the actual operation data to obtain a target distribution function corresponding to the target evaluation index;

a sampling module for obtaining a sampling number, the sampling number being determined based on at least one of the number of implemented objects or the number of second objects;

The sampling module is further used for sampling the target distribution function according to the sampling times and determining index information of the target evaluation index based on a sampling result; the index information is used for evaluating the release effect of the popularization information.

On the other hand, the application also provides computer equipment. The computer equipment comprises a memory and a processor, wherein the memory stores a computer program, and the processor realizes the steps of the popularization information throwing effect evaluation method when executing the computer program.

In another aspect, the present application also provides a computer-readable storage medium. The computer readable storage medium has stored thereon a computer program which, when executed by a processor, implements the steps of the promotional information delivery effect assessment method described above.

In another aspect, the present application also provides a computer program product. The computer program product comprises a computer program which, when executed by a processor, implements the steps of the promotional information delivery effect assessment method described above.

According to the popularization information throwing effect evaluation method, device, computer equipment, storage medium and computer program product, through obtaining actual operation data obtained by operating a first object based on popularization information for an implemented object corresponding to the popularization information after throwing, fitting the actual operation data according to the data, fitting different target distribution functions according to different target evaluation indexes, a large amount of data under corresponding evaluation indexes can be estimated based on limited small amount of data, sampling is conducted on the target distribution functions, index information of the target evaluation indexes is determined based on sampling results, and then the situation of all objects under the corresponding indexes is estimated, so that the throwing effect evaluation of the popularization information is achieved. The accurate evaluation of the popularization information throwing effect can be realized under the condition of fully protecting the privacy safety of the data through part of actual operation data.

Drawings

FIG. 1 is an application environment diagram of a promotional information delivery effect evaluation method in one embodiment;

FIG. 2 is a flow chart of a method for evaluating a promotional information delivery effect in one embodiment;

FIG. 3 is a flow chart illustrating steps for fitting an objective distribution function in one embodiment;

FIG. 4 is a flow chart illustrating steps for determining a sampling number in one embodiment;

FIG. 5 is a schematic diagram of a principle of delivery optimization for different popularization platforms in one embodiment;

FIG. 6 is a schematic diagram of a principle of evaluating a delivery effect of promotional information in one embodiment;

FIG. 7 is a diagram of an application environment for multiparty interactions in one embodiment;

FIG. 8 is a flow diagram of an estimation calculation flow in one embodiment;

FIG. 9 is a block diagram of a promotional information delivery effect evaluation apparatus in one embodiment;

fig. 10 is an internal structural view of a computer device in one embodiment.

Detailed Description

The present application will be described in further detail with reference to the drawings and examples, in order to make the objects, technical solutions and advantages of the present application more apparent. It should be understood that the specific embodiments described herein are for purposes of illustration only and are not intended to limit the scope of the application.

The popularization information delivery effect evaluation method provided by the embodiment of the application can be applied to an application environment shown in fig. 1. Wherein the terminal 102 is connected to the server 104 for communication. The terminal 102 and the server 104 may be directly or indirectly connected through wired or wireless communication, and the present application is not limited herein. The data storage system may store data that the server 104 needs to process. The data storage system may be integrated on the server 104 or may be located on the cloud or other servers.

The evaluation task of the popularization information throwing effect can be executed by the terminal or the server. The terminal or the server determines an implemented object corresponding to the popularization information after the popularization information is put in, and acquires actual operation data obtained by operating the first object based on the popularization information. It should be noted that, in the embodiment of the present application, the operation data obtained by the operation of the second object based on the promotion information is unknown. In other words, the terminal or the server needs to obtain an accurate delivery effect evaluation result based on the actual operation data of the first object. The terminal or the server fits according to the actual operation data to obtain a target distribution function corresponding to the target evaluation index, samples the target distribution function according to the sampling times, and determines index information of the target evaluation index based on the sampling result. Therefore, the terminal or the server can evaluate the release effect of the release promotion information on a certain promotion platform according to the index information.

In some embodiments, promotion information is used to promote an application, and the implemented object is an object in which the application is installed. The actual operation data is acquired and reported by a data reporting interface built in the application program. Taking the terminal as an example based on the iOS operating system, a plurality of SDKs (Software Development Kit, software development kits) are illustratively built in the application program promoted by the promotion information, namely a SKAdNetwork SDK, a three-party attribution SDK, and an SDK used by the application program itself to store operation data. The SKAdNetwork framework has two access modes, one is to directly access the related suite provided by the iOS-based operating system by itself, and the other is to update the three-party attribution SDK to a version with the SKAdNetwork framework. As the most basic data feedback, the SKAdNetwork SDK can provide the installation quantity of the application programs corresponding to each promotion information in each evaluation period after being attributed. Wherein the evaluation period is for example one day. The three-party attribution platform acquires attribution information reported by the three-party attribution SDK through the server, wherein the attribution information is used for indicating from which popularization platform the installed operation is specifically sourced. The SDK used by the application itself to store the operation data reports the operation and the corresponding operation data generated in the application to the background server of the application, so as to record the operations such as registration, activity, or payment generated in the application. Therefore, the data are collected and reported through the built-in data reporting interface, and the effect of the popularization information can be evaluated.

The evaluation index of the popularization information throwing effect can be set according to actual requirements, and can be preset according to different popularization application programs. In some embodiments, the pre-set evaluation metrics include, but are not limited to, one or more of a registration metric, a retention metric, or a consumption metric, among others. The target evaluation index is any one of preset evaluation indexes.

In some embodiments, in order to obtain the effect of the same promotion information on different promotion platforms, the server may evaluate the effect of each promotion platform, so as to perform statistics or comparison according to the evaluation result. Based on the above, the popularization information throwing effect evaluation method provided by the embodiment of the application can be combined with an artificial intelligence technology to realize intelligent throwing of popularization information on different popularization platforms. For example, the server intelligently selects popularization platforms to be continuously launched after the popularization information or intelligently adjusts popularization flow on each popularization platform according to the launching effect of the same popularization information on each popularization platform. For another example, the server intelligently selects popularization platforms to be put in by the popularization information to be put in according to the putting effect of the popularization information of the same type on each popularization platform, or intelligently adjusts the popularization flow on each popularization platform, and the like.

In some embodiments, the server may automatically adjust the delivery specific gravity in each popularization platform according to the delivery effect of the popularization information in each popularization platform by using a deep learning model or other neural network model based on a machine learning technology, so as to intelligently optimize the popularization information delivery channel, etc.

The terminal may be, but not limited to, one or more of various desktop computers, notebook computers, smart phones, tablet computers, internet of things devices, portable wearable devices, etc., and the internet of things devices may be one or more of smart speakers, smart televisions, smart air conditioners, or smart vehicle devices, etc. The portable wearable device may be one or more of a smart watch, a smart bracelet, or a headset device, etc.

The server may be an independent physical server, a server cluster or a distributed system formed by a plurality of physical servers, or a cloud server providing cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communication, middleware services, domain name services, security services, CDNs (Content Delivery Network, content delivery networks), basic cloud computing services such as big data and artificial intelligent platforms, and the like.

Big data (Big data) refers to a data set which cannot be captured, managed and processed by a conventional software tool within a certain time range, and is a massive, high-growth-rate and diversified information asset which needs a new processing mode to have stronger decision-making ability, insight discovery ability and flow optimization ability. With the advent of the cloud age, big data has attracted more and more attention, and special techniques are required for big data to effectively process a large amount of data within a tolerant elapsed time. Technologies applicable to big data include massively parallel processing databases, data mining, distributed file systems, distributed databases, cloud computing platforms, the internet, and scalable storage systems.

According to the embodiment of the application, by combining a big data technology and a big data calculation tool, the behavior distribution under natural conditions is fitted to the obtained authorized small data, so that the release effect of the popularization information is rapidly estimated on the premise of protecting the privacy security of the data, the difference between the estimated value and the actual value can be greatly eliminated, and the evaluation accuracy of the release effect of the popularization information is greatly improved. Meanwhile, the data support can be provided for the follow-up intelligent optimization of popularization information delivery based on the big data technology.

In some embodiments, the terminal may have APP (Application) applications loaded thereon, including applications that traditionally require separate installation, and applet applications that can be used without downloading an installation, such as one or more of a browser client, a web page client, an audio client, a video client, or a game client. The application program can be used as a software carrier of the promotion platform so as to provide promotion information put on the promotion platform to the user.

In some embodiments, various applications may be provided in the terminal, and an application service, such as a search service, a communication service, a payment service, or a game service, may be provided to the user through the application. In the process of providing application services for users, the terminal can initiate service call to the servers of all popularization platforms, and the servers push popularization information to the terminal for playing.

In one embodiment, as shown in fig. 2, a popularization information delivery effect evaluation method is provided, and the method can be applied to a terminal or a server, or can be cooperatively executed by the terminal and the server. The following description will take an example in which the method is applied to a computer device, which may be a terminal or a server. The method comprises the following steps:

Step S202, determining implemented objects corresponding to the popularization information after being put in; the implemented objects include a first object and a second object.

The server can put the promotion information on each promotion platform, and a user can browse and watch the promotion information on the promotion platform through the terminal. Wherein the promotion platform comprises one or more of a social platform, an audio/video playing platform, a game platform and the like. Promotion information is commonly used to promote and disseminate application products, such as game applications, as a way of promoting.

In some embodiments, the terminal obtains the promoted application program by triggering promotion information displayed on the terminal by the user, and the promoted application program may be, for example, a game APP or the like. Based on the promotion information triggered by the user, the terminal jumps to an application program product downloading interface corresponding to the promotion information, downloads and installs the application program product, and can be called as the terminal of the user at this time being an implemented object of the promotion information. The user's terminal can send the data such as download or installation to the server for the server to count.

Among the implemented objects, a part of the objects are authorized objects for which the user authorization application can acquire operation data, and are called first objects. While another part of the objects are unauthorized objects, called second objects, where the operation data cannot be obtained.

It should be noted that the above terms first and second, etc. are used in the present application to describe implemented objects in which application programs corresponding to promotion information are installed, but these objects should not be limited by these terms. These terms are only used to distinguish one object from another. For example, a first object may be referred to as a second object, and similarly, a second object may be referred to as a first object without departing from the scope of the various described embodiments, but they are not the same object unless the context clearly indicates otherwise. Similar situations also include a first distribution parameter, a second distribution parameter, a third distribution parameter, and so on.

The popularization information delivery effect evaluation method provided by the embodiment of the application can be applied to application program popularization information delivery scenes installed in terminal equipment based on an iOS operating system. The user may choose whether to allow the application to obtain the data based on his own will. The user may be free to choose to allow the application to obtain its operational data within the application, or may choose to reject, i.e., not allow the application to obtain its operational data within the application.

Under the condition that the user is allowed to be authorized, the application program can collect operation data such as whether the user registers, is active, or pays in the application program, and specific operation data such as the number of active days, or the consumption value of payment. Selecting the part of users allowing the application program to acquire the operation data, namely the first object; and the other part selects the part of users which do not allow the application program to acquire the operation data, namely the second object.

In some embodiments, the computer device determining an implemented object to which the promotional information corresponds after delivery includes: the computer equipment determines the number of implemented objects corresponding to the promotion information after delivery.

Taking the terminal based on the iOS operating system as an example, the SKAdNetwork based on the iOS operating system provides an installation-based attribution framework for terminal equipment manufacturers, and the attribution framework can provide the total number of installation of the application program, namely the number of implemented objects, on the premise of protecting the privacy of users. Taking the terminal of the Android operating system as an example, the computer device can directly obtain the number of implemented objects.

In the case that a part of the users select to allow the application program to acquire the operation data, no matter what privacy protection mode or the operation system, the computer device can count based on the acquired operation data to obtain the number of the first objects. In an embodiment of the present application, the computer device uses the difference between the number of implemented objects and the number of first objects as the number of second objects.

Step S204, obtaining actual operation data obtained by the operation of the first object based on the popularization information; wherein operation data obtained by the operation of the second object based on the promotion information is unknown.

The operation data obtained by the operation of the object based on the popularization information refers to data for operating the application program after the user triggers the popularization information to download the corresponding application program.

In some embodiments, the operations performed by the object based on the promotional information include, but are not limited to, one or more of downloading an application, installing an application, registering an account within an application, logging in an account, purchasing virtual items provided within an application, or paying for a certain consumption value, etc.

Accordingly, the operation data obtained by the operation of the object based on the promotion information includes, but is not limited to, one or more of a time stamp corresponding to various operations, whether an account is registered, the number of accounts registered, login behavior data, or specific consumption values paid, etc.

It should be noted that, since the first object is a user who selects to allow the application to acquire the operation data, the computer device may acquire the actual operation data obtained by the operation of the first object based on the promotion information. The second object is an unauthorized object, and the computer device does not acquire actual operation data of the second object, so that the operation data obtained by the operation of the second object based on the popularization information is unknown.

In some embodiments, the actual operation data of the first object further includes an object identifier of the first object, such as an IDFA (Identifier For Advertising, advertisement identifier) or the like, for distinguishing different first objects and for counting the number related to the first object later.

In some embodiments, the computer device obtaining actual operation data of the first object based on the promotion information, including: and the computer equipment acquires actual operation data reported by the data reporting interface according to the data reporting interface built in the application program promoted by the promotion information.

The application program promoted by the promotion information is internally provided with the SDK, and the computer equipment can report the operation and corresponding operation data generated in the application program to a background server of the application program according to the SDK interface, so that the operations such as registration, activity, payment and the like generated in the application program are recorded.

Step S206, fitting according to the actual operation data to obtain a target distribution function corresponding to the target evaluation index.

Specifically, the computer device fits a distribution function to the actual operational data of the first object to estimate the operational data distribution of all implemented objects. The target evaluation index is any one of a plurality of preset evaluation indexes. The preset evaluation index includes, but is not limited to, one or more of a registration index, a retention index, a consumption index, or the like.

And for the target evaluation index, the computer equipment fits according to the distribution function corresponding to the target evaluation index according to the actual operation data obtained by the operation of the first object based on the popularization information to obtain a target distribution function.

It should be noted that, according to different evaluation indexes, the computer device also has different types of corresponding actual operation data, so that the fitted distribution functions are also different. For example, for evaluation indexes such as registration, retention, etc., whether the first object is registered with an account, the number of registered accounts, whether it is in a retention state, and the number of days of retention are discrete, and the fitting may be generally performed based on a distribution function such as poisson distribution (Poisson Distribution). Whereas for an evaluation index such as consumption, the data of how much consumption value the first object pays is continuous, fitting is performed by a distribution function that may be based on a lognormal distribution (Log-Normal Distribution).

It is to be understood that the above distribution function is merely exemplary, and may be appropriately adjusted according to practical situations in a specific application scenario, for example, an appropriate distribution function may be selected according to data characteristics under different evaluation indexes.

In some embodiments, the computer device fits a target distribution function corresponding to the target evaluation index according to the actual operation data, including: the computer equipment estimates and obtains the distribution parameters of the target distribution function according to the actual operation data, and fits according to the distribution parameters, so as to obtain the target distribution function.

In step S208, the number of samples is acquired, and the number of samples is determined based on at least one of the number of implemented objects or the number of second objects.

After the target distribution function is obtained through fitting, the sampling times for sampling the target distribution function are required to be determined, so that the distribution condition of the whole objects is estimated according to the sampling result. Specifically, the computer device determines the number of samplings based on at least one of the number of implemented objects or the number of second objects.

In some embodiments, the computer device takes the number of implemented objects as the number of samples to sample the target distribution function a corresponding number of times. In other embodiments, the computer device takes the number of second objects as the number of samples to sample the target distribution function a corresponding number of times.

In still other embodiments, the computer device estimates operation data obtained by operating the second object based on the promotion information according to the number of the first objects among the number of the implemented objects, and determines the sampling number according to the estimation result. By taking the target evaluation index as a registration index example, the computer equipment calculates the registration rate in the first object according to the number of the first objects in the number of the implemented objects and the number of the registration accounts corresponding to the first objects; based on the registration rate and the number of the second objects, the computer equipment estimates and obtains the number of the registration accounts corresponding to the second objects, and takes the number of the registration accounts as the sampling times to sample the target distribution function corresponding to the registration index for corresponding times.

Step S210, sampling the target distribution function according to the sampling times, and determining index information of a target evaluation index based on a sampling result; the index information is used for evaluating the release effect of the popularization information.

Specifically, for any target evaluation index, for the fitted target distribution function corresponding to the target evaluation index, the computer device samples the target distribution function for a plurality of times based on the sampling times, and obtains a sampling result of each sampling. And the computer equipment performs statistics based on sampling results of each sampling to obtain index information under the target evaluation index. Therefore, the computer equipment can evaluate the throwing effect of the popularization information according to the index information. For a plurality of target evaluation indexes, the computer equipment respectively samples and counts the index information of each target evaluation index, so as to obtain the index information corresponding to each target evaluation index.

In some embodiments, the computer device may comprehensively evaluate the delivery effect of the promotional information in multiple dimensions based on the index information of a single target evaluation index, or a combination of the index information of multiple target evaluation indexes. For example, taking a registration index as an example, the computer device determines that the registration rate is lower than (or exceeds) the standard registration rate according to the difference between the evaluated registration rate and the standard registration rate, which indicates that the promotion information has a good effect in registration.

In some embodiments, the computer device may further obtain, after obtaining the index information corresponding to each target evaluation index, an overall delivery effect for the popularization information by combining the index information evaluations of all the target evaluations. For this purpose, the target evaluation index is any one of a plurality of preset evaluation indexes, and the method further includes: acquiring index information of each evaluation index, wherein standard index information is preset in each evaluation index; based on the difference between the index information of each evaluation index and the corresponding standard index information, the overall throwing effect aiming at the popularization information is evaluated.

Wherein, each evaluation index is preset with standard index information. For example, the standard indicator information may be a threshold as a standard for measuring indicator conditions.

Specifically, the computer device calculates, for a plurality of preset evaluation indexes, index information of each evaluation index, for example, a registration rate under a registration index, an average retention day under a retention index, an average consumption value under a consumption index, and the like. Accordingly, the standard index information of each index is, for example, a standard registration rate, a standard average retention day, or a standard average consumption value.

Thereby, the computer device calculates differences between the index information of each index and the corresponding standard index information, respectively. According to the difference corresponding to each index, the computer equipment combines all the indexes to comprehensively evaluate the overall throwing effect aiming at the popularization information.

In some embodiments, the computer device may further determine an error condition corresponding to each index according to a difference between the index information of each index and the corresponding standard index information. For example, taking the registration index as an example, the computer device determines that the registration rate error is reduced by 30% or the like based on the difference between the evaluated registration rate and the standard registration rate.

In the embodiment, the comprehensive evaluation is performed on the release effect of the popularization information by combining a plurality of indexes, so that the evaluation result is more accurate, the release effect of the popularization information can be inspected from various dimensions, the method is more comprehensive, and data support can be provided for subsequent optimization of the popularization information.

In some embodiments, the estimated release effect of the popularization information can be displayed through a release data report interface. In the data report interface, indexes such as the number of real installation users, the number of authorized users, the estimated registration rate, the estimated seven-day average retention days, the estimated average payment amount and the like for the promoted application program can be shown in the users using the terminal equipment based on the iOS operating system. Compared with a statistical report provided by a third party attribution platform, the method and the device provided by the embodiment of the application have the advantages that various index data are more accurate, and diversified indexes can be supported.

According to the popularization information delivery effect evaluation method, the actual operation data obtained by operating the first object based on the popularization information is obtained for the implemented object corresponding to the popularization information after delivery, fitting is performed according to the data, different target distribution functions are fitted according to different target evaluation indexes, a large amount of data under the corresponding evaluation indexes can be evaluated based on limited small amount of data, the target distribution functions are sampled, the index information of the target evaluation indexes is determined based on the sampling result, and then the situation of all objects under the corresponding indexes is evaluated, so that the evaluation of the delivery effect of the popularization information is achieved, and accurate effect evaluation can be achieved under the condition of fully protecting the privacy and safety of the data.

In one embodiment, as shown in fig. 3, fitting according to actual operation data to obtain a target distribution function corresponding to a target evaluation index includes:

step S302, acquiring the number of samples based on a first object executing a target operation; the target operation has a corresponding relation with the target evaluation index.

When the computer equipment performs fitting on the target score function according to the actual operation data of the first object, the adopted target score function is different according to different evaluation indexes, so that the distribution parameters required to be estimated are also different.

Specifically, the computer device estimates the distribution parameters using a maximum likelihood estimation method. Taking poisson distribution as an example, for n sample samples x of a random variable ₁ ,…,x _n For the distribution parameters thereofThe maximum likelihood estimate of (2) may be expressed by the following formula:

taking a lognormal distribution as an example, for n sample samples x of a random variable ₁ ,…,x _n For the distribution parameters thereofAnd->The maximum likelihood estimate of (2) may be expressed by the following formula:

in a similar manner for other distribution functions. The computer device obtains the number of samples based on the first object performing the target operation, which is n in each of the above formulas.

Wherein the target operation corresponds to a target evaluation index. For example, when the target evaluation index is a registration index, the target operation is a registration operation; accordingly, the number of samples of the first object performing the target operation may be the number of first objects registered with the account.

For another example, when the target evaluation index is a retention index, the target operation is a retention operation, such as a login operation. Accordingly, the sample number of the first object performing the target operation may be the account number of the first object in the persisted state.

For another example, when the target evaluation index is a consumption index, the target operation is a payment operation. Accordingly, the sample number of the first object performing the target operation may be the account number of the first object having the payment operation.

Step S304, determining target operation data corresponding to the target operation from the actual operation data.

For each sample, it is necessary to determine the actual target operation data under the target index, where the target operation data is the specific statistical data of the target operation, that is, x in the above formulas _i . Specifically, the computer device collects or counts actual operands of the bottom layer of the first object, thereby determining target operation data corresponding to each sample based on the actual operation data.

For example, when the target evaluation index is a registration index, each first object in which an account is registered may be registered with one or more accounts, and the target operation data is, for example, the registered account number.

For another example, when the target evaluation index is a retention index, the target operation data is an average retention day corresponding to the account in the retention state. In some embodiments, the retention indicator evaluates an average number of retention days over a preset period, e.g., an average number of retention days over seven days; the target operational data is, for example, an average number of retention days over seven days.

For another example, when the target evaluation index is a consumption index, the target operation data is, for example, an average consumption value corresponding to an account where a payment operation exists.

Step S306, estimating and obtaining distribution parameters according to the number of samples and target operation data, and fitting according to the distribution parameters to obtain a target distribution function; the target distribution function corresponds to a target evaluation index.

Specifically, the computer device performs maximum likelihood estimation based on the number of samples and the target operation data, thereby estimating the distribution parameters. The computer equipment can estimate by using the formula so as to obtain the distribution parameters of the Poisson distributionOr distribution parameters of lognormal distribution +.>And->Etc., and are not described in detail herein. Thus, after the distribution parameters are estimated, the computer device can fit the corresponding distribution functions based on the estimated distribution parameters, thereby obtaining the target distribution functions.

In the above embodiment, by estimating the distribution parameter fitting distribution function, the natural rule conforming to the natural human behavior conforming to the specific distribution under the natural condition, and sampling based on the distribution function obtained by utilizing the natural rule, the more accurate distribution condition of all users can be obtained.

Taking the target evaluation index as a registration index for example, in some embodiments, the target operation includes an account registration operation, the sample number includes a number of first objects performing the account registration operation, and the target operation data includes an account registration number corresponding to each of the first objects performing the account registration operation.

Correspondingly, according to the number of samples and the target operation data, estimating to obtain a distribution parameter, and according to the distribution parameter fitting to obtain a target distribution function, including: performing maximum likelihood estimation according to the number of the first objects for executing the account registration operation and the account registration number corresponding to each first object for executing the account registration operation, so as to obtain a first distribution parameter; and carrying out poisson distribution fitting according to the first distribution parameters to obtain a target distribution function.

Specifically, the computer device uses the number of first objects performing the account registration operation as a sample number n, and uses the account registration number corresponding to each first object performing the account registration operation as target operation data x _i Distribution parameters based on poisson distributionAnd (3) calculating to obtain a first distribution parameter.

Illustratively, for the registration indicator, the computer device is based on the number n of first objects performing the account registration operation and the account registration number x respectively corresponding to each first object performing the account registration operation _i Performing maximum likelihood estimation to obtain a first distribution parameterThe calculation formula of (a) is, for example:

in the above embodiment, the fitting of the number of registered accounts is performed on the first object with registration through poisson distribution, so that global estimation can be performed on the overall registration condition of the popularization information based on the related data of account registration by using the very limited first object, and accurate estimation of the registration index can be realized.

Taking the target evaluation index as a retention index for example, in some embodiments, the target operation includes a retention operation, the sample number includes the number of accounts in a retention state, and the target operation data includes retention days corresponding to each account in the retention state.

Correspondingly, according to the number of samples and the target operation data, estimating to obtain a distribution parameter, and according to the distribution parameter fitting to obtain a target distribution function, including: carrying out maximum likelihood estimation according to the number of accounts in the retention state and the retention days corresponding to the accounts in the retention state respectively to obtain second distribution parameters; and carrying out poisson distribution fitting according to the second distribution parameters to obtain a target distribution function.

Specifically, the computer device uses the number of first objects for executing the retention operation as the sample number n, and uses the retention days corresponding to the accounts in the retention state as the target operation data x _i Distribution parameters based on poisson distributionAnd (3) calculating to obtain a second distribution parameter.

For the retention index, the computer device is based on the number n of the first objects for executing the retention operation and the retention days x corresponding to the accounts in the retention state _i Performing maximum likelihood estimation to obtain second distribution parametersThe calculation formula of (a) is, for example:

in the above embodiment, the number of retention days is fitted to the account with the registered first object through the poisson distribution, so that global estimation can be performed on the overall retention condition of the popularization information based on the related data of the retention account with the very limited first object, and accurate evaluation of the retention index can be realized.

Taking the target evaluation index as a consumption index example, in some embodiments, the target operation comprises a payment operation, the sample number comprises the number of accounts with the payment operation, and the target operation data comprises consumption values respectively corresponding to the accounts with the payment operation.

Correspondingly, according to the number of samples and the target operation data, estimating to obtain a distribution parameter, and according to the distribution parameter fitting to obtain a target distribution function, including: carrying out maximum likelihood estimation according to the account number with the payment operation and consumption values corresponding to the accounts with the payment operation, so as to obtain a third distribution parameter; and carrying out lognormal distribution fitting according to the third distribution parameter to obtain a target distribution function.

Specifically, the computer device uses the number of the first objects for executing the payment operation as the sample number n, and uses the consumption value corresponding to each account with the payment operation as the target operation data x _i Distribution parameters based on log-normal distributionAnd->And (3) calculating to obtain a third distribution parameter.

Illustratively, for the consumption index, the computer device is based on the number n of first objects performing the payment operation, and the consumption value x corresponding to each account in which the payment operation exists _i Maximum likelihood estimation is carried out to obtain a third distribution parameterAnd->The specific calculation formula may refer to the above embodiment, and will not be described herein.

In the above embodiment, by fitting the consumption value to the account of the first object having consumption through the log-normal distribution, the overall consumption condition of the popularization information can be globally estimated based on the related data of the account having the payment operation by using the very limited first object, and the accurate evaluation of the consumption index can be realized.

After fitting to obtain a target distribution function, sampling the target distribution function according to sampling times to obtain a sampling result so as to realize index condition estimation of all objects. As stated previously, the number of samplings may be determined based on at least one of the number of implemented objects or the number of second objects. To this end, in some embodiments, as shown in fig. 4, the step of obtaining the number of samples may include:

Step S402, determining the index account number corresponding to the target evaluation index according to the actual operation data.

The index account number determined according to the actual operation data refers to the account number obtained by executing corresponding operation by the first object under the target evaluation index. Specifically, the computer device performs statistics according to the respective actual operation data of the first objects, so as to obtain the respective index account numbers corresponding to the first objects.

For example, when the target evaluation index is a registration index, the target operation is a registration operation; accordingly, the index account number corresponding to the registration index is obtained according to the number of registered accounts of each first object, for example, the average value of the number of registered accounts of the first object, and the like.

For another example, when the target evaluation index is a retention index, the target operation is a retention operation; accordingly, the index account number corresponding to the retention index can be obtained according to the account number in the retention state in the registered account of each first object, for example, the average value of the account numbers in the retention state of the first object, and the like.

For another example, when the target evaluation index is a consumption index, the target operation is a payment operation; accordingly, the index account number corresponding to the consumption index can be obtained according to the account number of the payment operation in the registered account of each first object, for example, the average value of the account numbers of the payment operation in the first object, and the like.

Step S404, determining the number of the first objects, and determining the actual operation proportion according to the index account number and the number of the first objects.

Specifically, the computer device determines the number of first objects, i.e., the number of objects to which the application promoted by the promotion information is installed. The index value condition of the second object under the target evaluation index is estimated by allowing the authorized first object to be under the target evaluation index. The index value condition is, for example, a registration rate, a retention day, or a consumption value.

In some embodiments, the computer device determines the actual operation scale based on the index account number and the number of first objects, comprising: and determining the ratio of the index account number to the number of the first objects, and taking the ratio as the actual operation proportion under the corresponding target evaluation index.

For example, taking the target evaluation index as the registration index, the computer device calculates the index account number U _{tracked,register} And the number N of the first objects _tracked And obtaining the actual registration proportion under the registration index based on the ratio of the two. The actual registration proportion characterizes the number of registration accounts corresponding to each first object. For example, the computer device may calculate by the following formula:

For another example, the target evaluation index is taken as a retention index, and the computer equipment calculates the number U of accounts according to the index _{tracked,retention} And the number N of the first objects _tracked And obtaining the actual retention ratio under the retention index based on the ratio of the two. The actual retention ratio characterizes the number of retention days corresponding to each account. For example, the computer device may calculate by the following formula:

for another example, using the target evaluation index as the consumption index, the computer device calculates the target evaluation index according to the index account U _{tracked,purchase} And the number N of the first objects _tracked Based on both ofThe ratio results in the actual consumption ratio at the consumption index. The actual consumption scale characterizes the consumption value corresponding to each account. For example, the computer device may calculate by the following formula:

step S406, determining the sampling times according to at least one of the number of the implemented objects or the number of the second objects and the actual operation proportion.

In some embodiments, the computer device determines the number of samples based on the number of implemented objects and the actual operating scale, including: the computer device takes as the number of samples the product of the number of implemented objects and the actual operating scale. In this manner, the overall index value case is obtained by calculating the index value case of the first object and directly generalizing the index value case for the first object to the entire object, that is, all the implemented objects.

For example, taking the target evaluation index as the registration index for illustration, the computer device calculates the actual operation scale under the registration index, i.e. the actual registration scale p _register Multiplying the number of implemented objects by N, and multiplying the product of the two by N p _register As the number of samples.

For another example, the target evaluation index is taken as a retention index for illustration, and the computer equipment is used for calculating the actual operation proportion under the retention index, namely the actual retention proportion p _retention Multiplying the number of implemented objects by N, and multiplying the product of the two by N p _retention As the number of samples.

For another example, the target evaluation index is taken as the consumption index, and the computer equipment performs the illustration according to the actual operation proportion under the consumption index, namely the actual consumption proportion p _purchase Multiplying the number of implemented objects by N, and multiplying the product of the two by N p _purchase As the number of samples.

In some embodiments, the computer device determines the number of samples based on the number of second objects and the actual operating scale, comprising: the computer device takes the product of the number of second objects and the actual operation scale as the number of samples. In this manner, estimating the index value condition of the portion of the object for the second object for which data is not available is achieved by calculating the index value condition of the first object and generalizing the index value condition for the first object to the second object for which data is not available.

Wherein the number N of the second objects _untracked Can be obtained by the difference between the number of implemented objects N and the number of first objects, i.e. N _untracked ＝N-N _tracked 。

For example, taking the target evaluation index as the registration index for illustration, the computer device calculates the actual operation scale under the registration index, i.e. the actual registration scale p _register Number N of implemented objects _untracked Multiplying to obtain the product N of the two _{untracked,register} ＝N _untracked *p _register As the number of samples.

For another example, the target evaluation index is taken as a retention index for illustration, and the computer equipment is used for calculating the actual operation proportion under the retention index, namely the actual retention proportion p _retention Number N of implemented objects _untracked Multiply and multiply the product of the two by N _{untracked,retention} ＝N _untracked *p _retention As the number of samples.

For another example, the target evaluation index is taken as the consumption index, and the computer equipment performs the illustration according to the actual operation proportion under the consumption index, namely the actual consumption proportion p _purchase Number N of implemented objects _untracked Multiply and multiply the product of the two by N _{untracked,purchase} ＝N _untracked *p _purchase As the number of samples.

In the above embodiment, the situation of index values under different evaluation indexes is calculated by using the very limited data of the first object, and the situation is expanded to the second object part or the whole object, so that the overall situation of the popularization information can be globally estimated, and the accurate evaluation of the throwing effect of each evaluation index is realized.

Wherein in some embodiments, when the sampling number is determined according to the number of implemented objects, sampling the target distribution function according to the sampling number, determining the index information of the target evaluation index based on the sampling result includes: sampling the target distribution function according to the sampling times, and counting based on the sampling results to obtain an estimated index value for the total implemented object; wherein, the single sampling result characterizes the estimated operation data of the single implemented object under the target evaluation index; the estimated index value of the full-scale implemented object is used as index information of the target estimated index.

Wherein the number of samples is determined based on the number of implemented objects, i.e. the number of samples is the number of implemented objects or the product of the number of implemented objects and the actual operation ratio calculated based on the number of first objects.

When the computer equipment samples the target distribution function for a plurality of times according to the sampling times, the sampling result obtained by each sampling represents the estimated operation data of a single implemented object under the target evaluation index. Wherein the estimated operation data is, for example, the number of registered accounts, the number of reserved days, or the consumption value of a single implemented object.

And the computer equipment performs statistics based on the sampling results of each time to obtain an estimated index value for the total implemented object. The estimated index value refers to an index condition of estimation for the total number of implemented objects under the target evaluation index, and may be, for example, the total number of registered accounts or the average number of registered accounts, the average number of retention days, or the average consumption value of the total number of implemented objects.

Therefore, the computer equipment can use the calculated estimated index value of the total implemented objects as the index information of the target estimated index so as to evaluate the release effect of the popularization information under the corresponding target estimated index. For example, the computer device determines whether the delivery effect of the promotion information can promote user conversion according to the average number of registered accounts of the total implemented objects estimated under the registration index.

In the above embodiment, by sampling based on the number of implemented objects, an estimated index value of the total implemented objects is obtained, that is, the effect of putting promotion information can be estimated directly as index information under the target estimated index, and global estimation can be performed by using a small portion of the data of the first object, thereby realizing accurate estimation of the effect of putting each estimated index.

In other embodiments, when the sampling number is determined according to the number of second objects, sampling the target distribution function according to the sampling number, determining the index information of the target evaluation index based on the sampling result includes: sampling the target distribution function according to the sampling times, and counting based on the sampling results to obtain an estimated index value aiming at the second object; wherein, the single sampling result characterizes the estimation operation data of the single second object under the target estimation index; determining an actual index value of the first object under the target evaluation index according to the actual operation data of the first object; and obtaining the index information of the target evaluation index according to the actual index value of the first object and the estimated index value of the second object.

Wherein the number of samples is determined based on the number of second objects, i.e. the number of samples is the number of second objects or the product of the number of second objects and the actual operation ratio calculated based on the number of first objects.

When the computer equipment samples the target distribution function for a plurality of times according to the sampling times, the sampling result obtained by each sampling represents the estimated operation data of the single second object under the target evaluation index. Wherein the estimated operation data is, for example, the number of registered accounts, the number of reserved days, or the consumption value of the single second object.

And the computer equipment performs statistics based on the sampling results of each time to obtain an estimated index value aiming at the second object. The estimated index value refers to an estimated index condition of the second object under the target evaluation index, and may be, for example, the total number of registered accounts or the average number of registered accounts, the average number of reserved days, or the average consumption value of the second object.

For example, when the target evaluation index is a registration index, statistics is performed based on the sampling results to obtain an evaluation index value for the second object, including: and carrying out summation statistics based on the sampling results to obtain the estimated registered account number aiming at the second object. The computer device obtains an estimated registered account number for the second object by summing according to the registered account number of the single second object obtained from each sampling result, wherein the estimated registered account number is the total registered account number.

Alternatively, in some embodiments, when the target evaluation index is a registration index, performing statistics based on the sampling results to obtain an evaluation index value for the second object includes: and carrying out mean statistics based on the sampling results to obtain the estimated registered account number aiming at the second object. The computer equipment obtains the estimated registered account number for the second object through averaging according to the registered account number of the single second object obtained from each sampling result, wherein the estimated registered account number is the average registered account number.

In another example, when the target evaluation index is a retention index, statistics is performed based on each sampling result to obtain an evaluation index value for the second object, including: and carrying out mean statistics based on the sampling results of each time to obtain the estimated retention days aiming at the second object. The computer device obtains estimated retention days for the second object by averaging according to the retention days of the single second object obtained from each sampling result, wherein the estimated retention days are average retention days.

For another example, when the target evaluation index is a consumption index, statistics is performed based on the sampling results to obtain an evaluation index value for the second object, including: and carrying out mean statistics based on the sampling results to obtain an estimated consumption value for the second object. The computer device obtains an estimated consumption value for the second object by averaging according to the consumption value of the single second object obtained from each sampling result, wherein the estimated consumption value is an average consumption value. In the above embodiment, the estimation index value of the second object can be quickly estimated by performing statistics, so that the index value of all implemented objects can be quickly estimated according to the actual index value of the first object and the estimation index value of the second object, and the calculation efficiency is high.

Thus, the computer equipment obtains the index information of the target evaluation index based on the ratio of the integrated index value of the actual index value of the first object and the calculated estimated index value of the second object and the number of the implemented objects.

For example, taking the target evaluation index as a registration index for illustration, the computer device calculates the sum of the actual registration account number of the first object and the estimated registration account number of the second object, and uses the ratio of the sum to the number of implemented objects as the average registration rate under the registration index, thereby evaluating the global registration condition under the registration index.

For another example, taking the target evaluation index as a retention index for illustration, the computer device calculates the sum of the actual retention days of the first object and the estimated retention days of the second object, and uses the ratio of the sum of the two to the number of implemented objects as the average retention days under the retention index, so as to evaluate the global retention condition under the retention index.

For another example, taking the target evaluation index as the consumption index for illustration, the computer equipment calculates the sum of the actual consumption value of the first object and the estimated consumption value of the second object, and uses the ratio of the sum to the number of the implemented objects as the average consumption value under the retention index, so as to evaluate the global consumption condition under the consumption index.

In the above embodiment, by sampling based on the number of the second objects to obtain the estimated index value of the second object, and calculating the index information of all the implemented objects under the global angle based on the actual index value of the first object and the estimated index value of the second object, global estimation can be performed by using a small amount of data of the first object, so as to realize accurate estimation of the delivery effect of each estimated index.

It should be noted that, for the retention index, statistics and evaluation are generally performed on the retention days in the preset period. In the process of sampling the fitted distribution function based on the sampling times, the sampling result needs to be subjected to stage processing so as to avoid that the sampling result exceeds the actual range. For example, if the average retention days in seven days are counted and evaluated, the maximum value of the average retention days in seven days is 6, and the excess portion needs to be cut off. In some embodiments, the computer device may perform subsequent calculations on the excess portions as a function of the maximum value to obtain as many data samples as possible. Alternatively, the computer device may discard the excess portion to maintain data accuracy.

As stated earlier, a promotional message may be delivered to multiple promotional platforms. To this end, in some embodiments, the promotional information carries a promotional platform identifier that is used to indicate the promotional platform to which the promotional information is to be placed. For example, the computer device may determine, according to information such as a promotion platform identifier carried by promotion information and a timestamp in actual operation data of the first object, that the downloading operation of the first object is derived from promotion information triggered on which promotion platform. Among them, the attribution modes include, but are not limited to, last click attribution, etc.

Correspondingly, the popularization information release effect evaluation method provided by the embodiment of the application further comprises the following steps: respectively acquiring the throwing effect of the popularization information on each popularization platform; and determining the throwing proportion of the popularization information in each popularization platform according to the throwing effect corresponding to each popularization platform.

Specifically, the computer equipment can determine a popularization platform corresponding to the popularization information in an attribution mode, and calculate index information of the popularization information under various evaluation indexes, so that the effect of throwing the popularization information on the popularization platform is evaluated. For different popularization platforms, the computer equipment can respectively calculate and evaluate, so that the throwing effect of the same popularization information on different popularization platforms is obtained.

Therefore, the computer equipment can determine the throwing proportion of the popularization information in each popularization platform according to the throwing effect corresponding to each popularization platform. For example, according to the respective corresponding release effect of each popularization platform, the computer device adjusts the popularization platform to be released continuously, and the release specific gravity of the popularization platform not released any more can be zero. For a popularization platform with good throwing effect, the computer equipment can heighten the throwing proportion of the popularization information in the popularization platform, and the like.

Illustratively, as shown in FIG. 5, taking a computer device as server 501 as an example, server 501 may be a single server or a cluster of distributed servers. The server 501 puts the same promotion information on different promotion platforms, such as promotion platform a, promotion platform B … …, etc., and the terminal 502 provides visual display of the promotion information. The computer device sets the throwing proportion P1 of the popularization platform a, the throwing proportion P2 … … of the popularization platform B and the like according to the throwing effect corresponding to each popularization platform.

The throwing effect of the same popularization information on each popularization platform is calculated respectively, and the throwing proportion of the popularization information in each popularization platform is adjusted according to the throwing effect corresponding to each popularization platform.

The application also provides an application scene, which applies the popularization information delivery effect evaluation method. Specifically, the application of the popularization information delivery effect evaluation method in the application scene is as follows: the server puts the promotion information on each promotion platform, and a user can browse and watch the promotion information on the promotion platform through the terminal. Based on the promotion information triggered by the user, the terminal jumps to an application program product downloading interface corresponding to the promotion information, downloads and installs the application program product, and can be called as the terminal of the user at this time being an implemented object of the promotion information. The user's terminal can send the data such as download or installation to the server for the server to count. The server can evaluate the statistical and popularization information throwing effect according to a preset evaluation period. The preset evaluation period is, for example, one day, one week, or the like.

When a preset evaluation period is reached, the server determines an implemented object corresponding to the popularization information after the popularization information is put in, and acquires actual operation data obtained by operating the first object based on the popularization information. The server then fits according to the actual operation data to obtain a target distribution function corresponding to the target evaluation index, and samples the target distribution function according to the sampling times. Finally, the server determines index information of the target evaluation index based on the sampling result, so that the release effect of the popularization information is evaluated based on the index information.

Of course, the method is not limited to this, and the popularization information delivery effect evaluation provided by the application can also be applied to other application scenes, such as popularization information flow optimization, machine learning training and prediction applied to popularization delivery, and the like.

The following is illustrated by way of an example. The popularization information release effect evaluation method provided by the embodiment of the application can be applied to application program popularization information release scenes installed in terminal equipment based on an iOS operating system. First, the user may choose whether to allow the application to obtain data based on his own needs. The user may be free to choose to allow the application to obtain its operational data within the application, or may choose to reject, i.e., not allow the application to obtain its operational data within the application.

Illustratively, a plurality of SDKs, namely a SKAdNetwork SDK, a three-party attribution SDK, and an SDK used by the application itself to store operation data, are built in the application promoted by the promotion information. The SKAdNetwork framework has two access modes, one is to directly access the related suite provided by the iOS-based operating system by itself, and the other is to update the three-party attribution SDK to a version with the SKAdNetwork framework. As the most basic data feedback, the SKAdNetwork SDK can provide the installation quantity of the application programs corresponding to each promotion information in each evaluation period after being attributed. Wherein the evaluation period is for example one day. The three-party attribution platform acquires attribution information reported by the three-party attribution SDK through the server, wherein the attribution information is used for indicating from which popularization platform the installed operation is specifically sourced. The SDK used by the application itself to store the operation data reports the operation and the corresponding operation data generated in the application to the background server of the application, so as to record the operations such as registration, activity, or payment generated in the application.

Taking an application program promoted by promotion information as a game APP as an example, for game manufacturers, the estimation of the release effect of the promotion information mainly takes registration rate, average retention days, average consumption value and the like as estimation indexes.

As shown in fig. 6, the computer device acquires the installation number acquired and reported by the SKAdNetwork SDK, acquires data acquired and reported by the three-party attribution SDK, and acquires data acquired and reported by the SDK used by the application itself to store the operation data, so that the computer device performs statistics on the related data of the first object. The data collected and reported by the three-party attribution SDK is, for example, attribution results indicating from which promotion platform the first object having the promoted application installed originated. The data collected and reported by the SDK used by the application itself to store the operation data is, for example, a behavior log, where the behavior log records various operation data of the first object in the application, and each operation data usually carries a timestamp.

After the underlying data is acquired, the computer device determines a promotion platform to which the promotion information is put based on the attribution result, and counts related data of the first object brought by the promotion information put on the promotion platform based on the behavior log, wherein the related data of the first object includes, but is not limited to, the number of the first object, the number of registered accounts of the first object, the number of accounts of the first object in a retention state, the number of accounts of the first object in a presence payment operation, and the like.

For the statistical obtained actual operation data of the first object, the computer equipment estimates based on a maximum likelihood method, respectively estimates corresponding distribution parameters for different evaluation indexes, and fits corresponding distribution functions based on the estimated distribution parameters. The computer device estimates the relevant data of the second object by sampling the fitted distribution function, wherein the relevant data of the second object includes, but is not limited to, the number of the second objects, the number of registered accounts of the second objects, the number of accounts of the second objects in a retention state, the number of accounts of the second objects in a presence payment operation, and the like.

Thus, based on at least one of the statistical result of the actual operation data of the first object and the statistical result (not shown) of the estimated related data or related data of the second object, the computer device calculates the index information under different evaluation indexes, and evaluates the delivery effect of the popularization information delivered at the popularization platform according to the index information.

For each popularization platform, the computer equipment can evaluate the throwing effect under each index respectively, so that the throwing effect of the same popularization information corresponding to different popularization platforms is obtained.

Therefore, the popularization information release effect evaluation method provided by the embodiment of the application can provide accurate estimation of the release effect of the released popularization information, and further can provide data support for decision optimization of the popularization information.

Illustratively, as shown in fig. 7, the server executing the method for evaluating the effect of delivering the promotional information is a publisher of the promotional information, and the publisher can promote a plurality of promotional information on different promotion platforms and distinguish each promotional information according to the promotional information identifier carried by each promotional information. The promotion information identification is, for example, the number of the promotion information or a watermark. The installation number collected by the SKAdNetwork SDK is reported to a provider of the hardware environment, for example, a terminal equipment manufacturer or an operating system manufacturer. The publisher of the promotion information can realize data transmission with the provider of the hardware environment through communication connection between the servers. And reporting the data acquired by the three-party attribution SDK to a third-party attribution platform party. The publisher of the promotion information can realize data transmission with the third party attribution platform through communication connection between the servers. The application program itself is used for storing the data collected by the SDK of the operation data and reporting the data to the provider of the application program. The provider of the application may be a distributor of the promotion information or another party independent of the distributor of the promotion information. The publisher of the promotion information can realize data transmission with the provider of the application program through communication connection between the servers.

In some embodiments, the overall estimation calculation flow may be as shown in fig. 8, for example. In the first step, when a user browses or watches the promotion information push, downloads and opens the promoted APP by triggering the promotion information, the APP asks the user whether to allow the APP to acquire related operation data based on the privacy protection policy. The user is free to choose whether to allow or reject. If the user allows, the APP can acquire whether the user performs operations such as registration, activity, payment and the like in the APP, and operation data such as the number of active days, payment amount and the like.

In the second step, the computer device estimates the distribution of each index and the corresponding parameters based on the collected actual operation data. Taking the registration index as an example, the registration account number of the registered object can be fitted for the poisson distribution, and the distribution parameters of the poisson distribution can be obtained through maximum likelihood estimation.

In the third step, the computer device obtains the overall installation number of the promotion information, that is, the number of implemented objects, based on the SKAdNetwork framework.

In the fourth step, the computer equipment performs random sampling for a certain sampling times based on the distribution and the parameters obtained by the second step and on the overall installation number obtained by the third step, so that overall index evaluation of the popularization information throwing effect is obtained.

According to the popularization information effect evaluation method provided by the embodiment of the application, the overall evaluation of the indexes related to the overall throwing effect of the popularization information is carried out by utilizing the data of the very limited partial authorized users, so that the accurate evaluation of each index can be realized, the evaluation is not required to be carried out based on a SKAdNetwork frame specially designed complex conversion value mode, and the efficiency is higher.

It should be understood that, although the steps in the flowcharts related to the above embodiments are sequentially shown as indicated by arrows, these steps are not necessarily sequentially performed in the order indicated by the arrows. The steps are not strictly limited to the order of execution unless explicitly recited herein, and the steps may be executed in other orders. Moreover, at least some of the steps in the flowcharts described in the above embodiments may include a plurality of steps or a plurality of stages, which are not necessarily performed at the same time, but may be performed at different times, and the order of the steps or stages is not necessarily performed sequentially, but may be performed alternately or alternately with at least some of the other steps or stages.

Based on the same inventive concept, the embodiment of the application also provides a popularization information delivery effect evaluation device for realizing the above related popularization information delivery effect evaluation method. The implementation scheme of the solution to the problem provided by the device is similar to the implementation scheme recorded in the method, so the specific limitation in the embodiments of the one or more promotion information delivery effect evaluation devices provided below can be referred to the limitation of the promotion information delivery effect evaluation method hereinabove, and will not be repeated here.

In one embodiment, as shown in fig. 9, there is provided a promotion information delivery effect evaluation apparatus 900, including: a determining module 901, an acquiring module 902, a fitting module 903 and a sampling module 904, wherein:

a determining module 901, configured to determine an implemented object corresponding to the promotion information after delivery; the implemented objects include a first object and a second object.

An obtaining module 902, configured to obtain actual operation data obtained by operating the first object based on the promotion information; wherein operation data obtained by the operation of the second object based on the promotion information is unknown.

The fitting module 903 is configured to fit the target distribution function corresponding to the target evaluation index according to the actual operation data.

A sampling module 904 for obtaining a number of samples, the number of samples being determined based on at least one of a number of implemented objects or a number of second objects.

The sampling module 904 is further configured to sample the target distribution function according to the sampling times, and determine index information of the target evaluation index based on the sampling result; the index information is used for evaluating the release effect of the popularization information.

In some embodiments, the fitting module is further configured to obtain a number of samples based on the first object performing the target operation; the target operation has a corresponding relation with the target evaluation index; determining target operation data corresponding to the target operation from the actual operation data; estimating to obtain distribution parameters according to the number of samples and target operation data, and fitting to obtain a target distribution function according to the distribution parameters; the target distribution function corresponds to a target evaluation index.

In some embodiments, the target evaluation metrics comprise registration metrics, the target operations comprise account registration operations, the sample number comprises a number of first objects performing the account registration operations, and the target operation data comprises an account registration number corresponding to each of the first objects performing the account registration operations; the fitting module is further used for carrying out maximum likelihood estimation according to the number of the first objects executing the account registration operation and the account registration number corresponding to each first object executing the account registration operation, so as to obtain a first distribution parameter; and carrying out poisson distribution fitting according to the first distribution parameters to obtain a target distribution function.

In some embodiments, the target evaluation index comprises a retention index, the target operation comprises a retention operation, the sample number comprises the number of accounts in a retention state, and the target operation data comprises retention days respectively corresponding to the accounts in the retention state; the fitting module is also used for carrying out maximum likelihood estimation according to the number of accounts in the retention state and the retention days corresponding to the accounts in the retention state respectively to obtain a second distribution parameter; and carrying out poisson distribution fitting according to the second distribution parameters to obtain a target distribution function.

In some embodiments, the target evaluation index comprises a consumption index, the target operation comprises a payment operation, the sample number comprises a number of accounts for which the payment operation is present, and the target operation data comprises consumption values respectively corresponding to the accounts for which the payment operation is present; the fitting module is further used for carrying out maximum likelihood estimation according to the number of accounts with payment operations and consumption values corresponding to the accounts with the payment operations, so as to obtain a third distribution parameter; and carrying out lognormal distribution fitting according to the third distribution parameter to obtain a target distribution function.

In some embodiments, the sampling module is further configured to determine an index account number corresponding to the target evaluation index according to the actual operation data; determining the number of the first objects, and determining the actual operation proportion according to the index account number and the number of the first objects; the number of samplings is determined according to at least one of the number of implemented objects or the number of second objects and the actual operation scale.

In some embodiments, when the sampling number is determined according to the number of implemented objects, the sampling module is further configured to sample the target distribution function according to the sampling number, and obtain an estimated index value for the total implemented objects based on statistics of sampling results of each time; wherein, the single sampling result characterizes the estimated operation data of the single implemented object under the target evaluation index; the estimated index value of the full-scale implemented object is used as index information of the target estimated index.

In some embodiments, when the sampling number is determined according to the number of the second objects, the sampling module is further configured to sample the target distribution function according to the sampling number, and calculate an estimated index value for the second object based on the sampling results of each time; wherein, the single sampling result characterizes the estimation operation data of the single second object under the target estimation index; determining an actual index value of the first object under the target evaluation index according to the actual operation data of the first object; and obtaining the index information of the target evaluation index according to the actual index value of the first object and the estimated index value of the second object.

In some embodiments, the sampling module is further configured to, when the target evaluation index is a registration index, perform summation statistics based on the sampling results to obtain an estimated registration account number for the second object.

In some embodiments, the sampling module is further configured to, when the target evaluation index is a retention index, perform mean statistics based on each sampling result to obtain an estimated retention day for the second object.

In some embodiments, the sampling module is further configured to perform mean statistics based on each sampling result when the target evaluation index is a consumption index, to obtain an estimated consumption value for the second object.

In some embodiments, the target evaluation index is any one of a plurality of preset evaluation indexes, and the device further comprises an evaluation module, configured to obtain index information of each evaluation index, where each evaluation index is preset with standard index information; based on the difference between the index information of each evaluation index and the corresponding standard index information, the overall throwing effect aiming at the popularization information is evaluated.

In some embodiments, promotion information is used to promote an application, and the implemented object is an object on which the application is installed; the actual operation data is acquired and reported by a data reporting interface built in the application program.

In some embodiments, the promotion information carries a promotion platform identification, which is used for indicating a promotion platform to which the promotion information is put; the device also comprises an optimizing module, a control module and a control module, wherein the optimizing module is used for respectively acquiring the throwing effect of the popularization information on each popularization platform; and determining the throwing proportion of the popularization information in each popularization platform according to the throwing effect corresponding to each popularization platform.

All or part of each module in the popularization information delivery effect evaluation device can be realized by software, hardware and a combination thereof. The above modules may be embedded in hardware or may be independent of a processor in the computer device, or may be stored in software in a memory in the computer device, so that the processor may call and execute operations corresponding to the above modules.

In one embodiment, a computer device is provided, which may be a server or a terminal. The internal structure of the computer device will be described below by taking the computer device as a server as an example, and the internal structure can be shown in fig. 10. The computer device includes a processor, a memory, an Input/Output interface (I/O) and a communication interface. The processor, the memory and the input/output interface are connected through a system bus, and the communication interface is connected to the system bus through the input/output interface. Wherein the processor of the computer device is configured to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, computer programs, and a database. The internal memory provides an environment for the operation of the operating system and computer programs in the non-volatile storage media. The database of the computer device is used for storing operation data and the like. The input/output interface of the computer device is used to exchange information between the processor and the external device. The communication interface of the computer device is used for communicating with an external terminal through a network connection. The computer program, when executed by the processor, implements a promotional information delivery effect assessment method.

It will be appreciated by those skilled in the art that the structure shown in FIG. 10 is merely a block diagram of some of the structures associated with the present inventive arrangements and is not limiting of the computer device to which the present inventive arrangements may be applied, and that a particular computer device may include more or fewer components than shown, or may combine some of the components, or have a different arrangement of components.

In an embodiment, there is also provided a computer device comprising a memory and a processor, the memory having stored therein a computer program, the processor implementing the steps of the method embodiments described above when the computer program is executed.

In one embodiment, a computer-readable storage medium is provided, on which a computer program is stored which, when executed by a processor, carries out the steps of the method embodiments described above.

In an embodiment, a computer program product is provided, comprising a computer program which, when executed by a processor, implements the steps of the method embodiments described above.

It should be noted that, the user information (including but not limited to user equipment information, user personal information, etc.) and the data (including but not limited to data for analysis, stored data, presented data, etc.) related to the present application are information and data authorized by the user or sufficiently authorized by each party, and the collection, use and processing of the related data need to comply with the related laws and regulations and standards of the related country and region.

In addition, in the embodiment of the application, the user can select whether to allow the application program to acquire the data. The user may be free to choose to allow the application to obtain its operational data within the application, or may choose to reject, i.e., not allow the application to obtain its operational data within the application. Meanwhile, for the released promotion information, the user can allow pushing of the promotion information so as to browse or watch the promotion information. The user may also reject the pushing of promotional information, such as by an application interface to quickly close the pushed information, or reject a promotional announcement, etc.

Those skilled in the art will appreciate that implementing all or part of the above described methods may be accomplished by way of a computer program stored on a non-transitory computer readable storage medium, which when executed, may comprise the steps of the embodiments of the methods described above. Any reference to memory, database, or other medium used in embodiments provided herein may include at least one of non-volatile and volatile memory. The nonvolatile Memory may include Read-Only Memory (ROM), magnetic tape, floppy disk, flash Memory, optical Memory, high density embedded nonvolatile Memory, resistive random access Memory (ReRAM), magnetic random access Memory (Magnetoresistive Random Access Memory, MRAM), ferroelectric Memory (Ferroelectric Random Access Memory, FRAM), phase change Memory (Phase Change Memory, PCM), graphene Memory, and the like. Volatile memory can include random access memory (Random Access Memory, RAM) or external cache memory, and the like. By way of illustration, and not limitation, RAM can be in the form of a variety of forms, such as static random access memory (Static Random Access Memory, SRAM) or dynamic random access memory (Dynamic Random Access Memory, DRAM), and the like. The databases referred to in the embodiments provided herein may include at least one of a relational database and a non-relational database. The non-relational database may include, but is not limited to, a blockchain-based distributed database, and the like. The processor referred to in the embodiments provided in the present application may be a general-purpose processor, a central processing unit, a graphics processor, a digital signal processor, a programmable logic unit, a data processing logic unit based on quantum computing, or the like, but is not limited thereto.

The technical features of the above embodiments may be arbitrarily combined, and all possible combinations of the technical features in the above embodiments are not described for brevity of description, however, as long as there is no contradiction between the combinations of the technical features, they should be considered as the scope of the description.

The foregoing examples illustrate only a few embodiments of the application and are described in detail herein without thereby limiting the scope of the application. It should be noted that it will be apparent to those skilled in the art that several variations and modifications can be made without departing from the spirit of the application, which are all within the scope of the application. Accordingly, the scope of the application should be assessed as that of the appended claims.

Claims

1. The popularization information delivery effect evaluation method is characterized by comprising the following steps:

2. The method according to claim 1, wherein said fitting the actual operation data to obtain the target distribution function corresponding to the target evaluation index comprises:

acquiring the number of samples based on a first object performing a target operation; the target operation has a corresponding relation with the target evaluation index;

determining target operation data corresponding to the target operation from the actual operation data;

estimating to obtain distribution parameters according to the sample number and the target operation data, and fitting to obtain a target distribution function according to the distribution parameters; the target distribution function corresponds to the target evaluation index.

3. The method of claim 2, wherein the target evaluation metrics include a registration metric, the target operations include account registration operations, the sample number includes a number of first objects performing account registration operations, and the target operation data includes account registration numbers corresponding to the first objects performing account registration operations, respectively;

Estimating a distribution parameter according to the sample number and the target operation data, and fitting according to the distribution parameter to obtain a target distribution function, wherein the method comprises the following steps:

performing maximum likelihood estimation according to the number of the first objects for executing the account registration operation and the account registration number corresponding to each first object for executing the account registration operation, so as to obtain a first distribution parameter;

and carrying out poisson distribution fitting according to the first distribution parameters to obtain a target distribution function.

4. The method of claim 2, wherein the target evaluation index comprises a retention index, the target operation comprises a retention operation, the sample number comprises an account number in a retention state, and the target operation data comprises retention days respectively corresponding to each account in a retention state;

carrying out maximum likelihood estimation according to the number of the accounts in the retention state and the retention days corresponding to the accounts in the retention state respectively to obtain second distribution parameters;

And carrying out poisson distribution fitting according to the second distribution parameters to obtain a target distribution function.

5. The method of claim 2, wherein the target assessment indicator comprises a consumption indicator, the target operation comprises a payment operation, the sample number comprises a number of accounts for which a payment operation exists, and the target operation data comprises consumption values for which respective accounts for the payment operation exist;

carrying out maximum likelihood estimation according to the account number with the payment operation and consumption values corresponding to the accounts with the payment operation, so as to obtain a third distribution parameter;

and carrying out lognormal distribution fitting according to the third distribution parameter to obtain a target distribution function.

6. The method of claim 1, wherein the obtaining the number of samples comprises:

determining an index account number corresponding to the target evaluation index according to the actual operation data;

determining the number of the first objects, and determining an actual operation proportion according to the index account number and the number of the first objects;

And determining the sampling times according to at least one of the number of the implemented objects or the number of the second objects and the actual operation proportion.

7. The method according to claim 1, wherein when the sampling number is determined according to the number of implemented objects, the sampling the target distribution function according to the sampling number, determining the index information of the target evaluation index based on the sampling result, comprises:

sampling the target distribution function according to the sampling times, and counting based on sampling results to obtain an estimated index value for the total implemented object; wherein a single sampling result characterizes estimated operational data of a single implemented object under the target evaluation index;

and taking the estimated index value of the full-scale implemented object as index information of the target estimated index.

8. The method according to claim 1, wherein when the sampling number is determined according to the number of the second objects, the sampling the target distribution function according to the sampling number, determining the index information of the target evaluation index based on the sampling result, comprises:

Sampling the target distribution function according to the sampling times, and counting based on sampling results to obtain an estimated index value aiming at the second object; wherein a single sampling result characterizes estimated operation data of a single second object under the target evaluation index;

determining an actual index value of the first object under the target evaluation index according to the actual operation data of the first object;

and obtaining the index information of the target evaluation index according to the actual index value of the first object and the estimated index value of the second object.

9. The method of claim 8, wherein the counting based on the sub-sampling results to obtain the estimated index value for the second object comprises:

when the target evaluation index is a registration index, summing statistics is carried out based on each sampling result, so that the estimated registration account number aiming at the second object is obtained;

when the target evaluation index is a retention index, carrying out mean value statistics based on each sampling result to obtain the estimated retention days aiming at the second object;

and when the target evaluation index is a consumption index, carrying out mean value statistics based on each sampling result to obtain an estimated consumption value aiming at the second object.

10. The method according to any one of claims 1 to 9, wherein the target evaluation index is any one of a plurality of preset evaluation indexes, the method further comprising:

acquiring index information of each evaluation index, wherein standard index information is preset in each evaluation index;

based on the difference between the index information of each evaluation index and the corresponding standard index information, the overall throwing effect aiming at the popularization information is evaluated.

11. The method according to any one of claims 1 to 9, wherein the promotion information is used to promote an application, and the implemented object is an object on which the application is installed; the actual operation data is acquired and reported by a data reporting interface built in the application program.

12. The method according to any one of claims 1 to 9, wherein the promotion information carries a promotion platform identification for indicating a promotion platform to which the promotion information is to be delivered; the method further comprises the steps of:

respectively acquiring the release effect of the popularization information on each popularization platform;

and determining the throwing proportion of the popularization information in each popularization platform according to the throwing effect corresponding to each popularization platform.

13. A promotional information delivery effect assessment device, the device comprising:

14. A computer device comprising a memory and a processor, the memory storing a computer program, characterized in that the processor implements the steps of the method of any one of claims 1 to 12 when the computer program is executed.

15. A computer readable storage medium, on which a computer program is stored, characterized in that the computer program, when being executed by a processor, implements the steps of the method of any of claims 1 to 12.

16. A computer program product comprising a computer program, characterized in that the computer program, when being executed by a processor, implements the steps of the method of any one of claims 1 to 12.