WO2022237477A1

WO2022237477A1 - Data calibration method and apparatus, and computer device and readable storage medium

Info

Publication number: WO2022237477A1
Application number: PCT/CN2022/087839
Authority: WO
Inventors: 李少波
Original assignee: 腾讯科技（深圳）有限公司
Priority date: 2021-05-11
Filing date: 2022-04-20
Publication date: 2022-11-17
Also published as: CN115330428A

Abstract

A data calibration method and apparatus, and a computer device and a readable storage medium. The data calibration method comprises: acquiring resource attribute information of a target service resource under N resource attribute types; performing clustering on a historical service resource set on the basis of S combination types, so as to obtain H historical service resource subsets; performing aggregation statistics processing on the basis of the H historical service resource subsets, so as to obtain an aggregated data set; according to the resource attribute information, acquiring, from the aggregated data set, an effective conversion quantity and an effective joint factor for the target service resource; and determining a calibration coefficient according to the effective conversion quantity and the effective joint factor, and calibrating an estimated joint factor of the target service resource according to the calibration coefficient, wherein the estimated joint factor is determined according to an estimated conversion rate and an industry factor of the target service resource. By means of the method, a deviation between an estimated value and an actual value of advertisement display costs can be reduced, thereby improving the estimation accuracy.

Description

A data calibration method, device, computer equipment and readable storage medium

This application claims the priority of the Chinese patent application submitted to the China Patent Office on May 11, 2021, with the application number 202110513300X and the application name "A data calibration method, device, computer equipment, and readable storage medium", all of which The contents are incorporated by reference in this application.

technical field

This application relates to the field of Internet technology, and in particular to data calibration technology in the field of advertising.

Background technique

With the rise of the Internet, online advertising has become the main profit method for major advertising platforms and advertisers. Before advertising, the advertising platform needs to consider the benefits of all aspects (including users, advertisers and advertising platforms), and then estimate the display cost of the advertisement to be placed (ECPM, Effective Cost per Mile), that is, the estimated advertising platform The advertising fee that can be obtained by exposing an advertisement to be delivered.

At present, for optimized behavior bidding (oCPA, Optimized Cost per Action) advertisements, advertisers usually need to give a conversion bid (targetCpa, target Cost per Action), that is, the cost that is expected to be paid once the advertisement is converted, wherein conversion refers to the completion of the advertisement The optimization behavior expected by the master, such as completing an application registration through advertisements, etc. The ECPM estimation of oCPA needs to estimate the relationship between the exposure amount and the conversion number, and then determine the ECPM according to the taregtCpa combined with the estimated relationship. However, there is often a certain deviation between the estimated value of ECPM obtained by the existing technology and the actual value, and the estimation accuracy is low.

Contents of the invention

The embodiments of the present application provide a data calibration method, device, computer equipment, and readable storage medium, which can reduce the deviation between the estimated value and the actual value of the advertising display fee, and improve the estimated accuracy.

On the one hand, an embodiment of the present application provides a data calibration method executed by a computer device, including:

Obtain the resource attribute information of the target business resource under N resource attribute types; N is a positive integer;

Each historical business resource in the historical business resource set is clustered based on S combination types to obtain H historical business resource subsets; the resource attribute types corresponding to each combination type belong to N resource attribute types; a historical business resource The historical resource attribute combinations of each historical business resource in the resource subset are the same, and a historical resource attribute combination is associated with a resource attribute type corresponding to a combination type; S is a positive integer less than or equal to N; H is a positive integer;

Based on the H historical business resource sub-sets, the aggregated statistical processing is performed separately to obtain the aggregated data set; the aggregated data set includes the conversion numbers and joint factors corresponding to each of the H historical business resource sub-sets; the conversion numbers are based on the corresponding historical business resource sub-sets The conversion number of each historical business resource in the set is determined, and the joint factor is determined according to the estimated conversion rate and industry factor of each historical business resource in the corresponding historical business resource subset;

According to the resource attribute information, the effective conversion number and effective joint factor for the target business resource are obtained in the aggregated data set, and the calibration coefficient is determined according to the effective conversion number and effective joint factor;

The estimated joint factor of the target business resource is calibrated according to the calibration coefficient; the estimated joint factor is determined according to the estimated conversion rate and the industry factor of the target business resource.

An embodiment of the present application provides a data calibration device on the one hand, including:

An acquisition module, configured to acquire resource attribute information of target business resources under N resource attribute types; N is a positive integer;

The division module is used to cluster the historical business resources in the historical business resource set based on S combination types to obtain H historical business resource subsets; the resource attribute type corresponding to each combination type belongs to N resource attributes type; the historical resource attribute combination of each historical business resource in a historical business resource subset is the same, and a historical resource attribute combination is associated with a resource attribute type corresponding to a combination type; S is a positive integer less than or equal to N; H is a positive integer;

The aggregation statistics module is used to perform aggregate statistics processing based on the H historical business resource sub-sets respectively to obtain an aggregated data set; the aggregated data set includes the respective conversion numbers and joint factors corresponding to the H historical business resource sub-sets; the conversion numbers are based on The conversion number of each historical business resource in the corresponding historical business resource subset is determined, and the joint factor is determined according to the corresponding estimated conversion rate and industry factor of each historical business resource in the corresponding historical business resource subset;

The effective data determination module is used to obtain the effective conversion number and effective joint factor for the target business resource in the aggregated data set according to the resource attribute information;

The calibration module is used to determine the calibration coefficient according to the effective conversion number and the effective joint factor, and to calibrate the estimated joint factor of the target business resource according to the calibration coefficient; the estimated joint factor is determined according to the estimated conversion rate and industry factor of the target business resource of.

An embodiment of the present application provides a computer device, including: a processor, a memory, and a network interface;

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

An embodiment of the present application provides a computer-readable storage medium, where a computer program is stored in the computer-readable storage medium, and the computer program is adapted to be loaded by a processor and execute the method in the embodiment of the present application.

Embodiments of the present application provide a computer program product or computer program on the one hand, the computer program product or computer program includes computer instructions, the computer instructions are stored in a computer-readable storage medium, and the processor of the computer device reads from the computer-readable storage The medium reads the computer instruction, and the processor executes the computer instruction, so that the computer device executes the method in the embodiment of the present application.

According to the resource attribute information of the target business resource, the embodiment of the present application can obtain the effective conversion number and the effective combination factor determined according to the historical business resources related to the target business resource, and then use the effective conversion number and the effective combination factor to calculate the target The estimated joint factor of the service resource is calibrated, and then, according to the calibrated estimated joint factor, the estimated value of the display fee of the target service resource is adjusted. Wherein, the estimated joint factor is determined according to the estimated conversion rate of the target business resource and the industry factor. Among them, the estimated conversion rate and the industry factor are the main factors for calculating the display fee of the target business resource. By calibrating the estimated joint factor, the overall calculation error caused by the inaccurate estimated conversion rate or industry factor to the target business resource display cost can be reduced, thereby reducing the gap between the estimated value and the actual value of the target business resource display cost deviation to improve the prediction accuracy.

Description of drawings

FIG. 1 is a schematic diagram of a system architecture provided by an embodiment of the present application;

Fig. 2a is a schematic diagram of a data calibration scenario provided by an embodiment of the present application;

Fig. 2b is a schematic diagram of a data calibration scenario provided by an embodiment of the present application;

Fig. 3 is a schematic flow chart of a data calibration method provided by an embodiment of the present application;

Fig. 4a is a schematic diagram of a scenario for dividing historical service resource sets provided by an embodiment of the present application;

Fig. 4b is a schematic diagram of a determination scenario of an effective conversion number and an effective combination factor provided by the embodiment of the present application;

Fig. 5 is a schematic flow chart of a data calibration method provided by an embodiment of the present application;

Fig. 6 is a schematic flow chart of an analysis method for determining the combination factor of display costs provided by the embodiment of the present application;

Fig. 7 is a schematic structural diagram of a data calibration device provided by an embodiment of the present application;

Fig. 8 is a schematic structural diagram of a computer device provided by an embodiment of the present application.

Detailed ways

The technical solutions in the embodiments of the present application will be clearly and completely described below in conjunction with the accompanying drawings in the embodiments of the present application. Obviously, the described embodiments are only some of the embodiments of the present application, not all of them. Based on the embodiments in this application, all other embodiments obtained by persons of ordinary skill in the art without creative efforts fall within the protection scope of this application.

Please refer to FIG. 1 , which is a schematic diagram of a system architecture provided by an embodiment of the present application. The system architecture may include a service server 100 and a user terminal cluster. The user terminal cluster may include: user terminal 200a, user terminal 200b, user terminal 200c, ..., user terminal 200n, and multiple user terminals in the user terminal cluster may exist The communication connection, for example, there is a communication connection between the user terminal 200a and the user terminal 200b, and there is a communication connection between the user terminal 200a and the user terminal 200c. Meanwhile, any user terminal in the user terminal cluster may have a communication connection with the service server 100 , for example, there is a communication connection between the user terminal 200 a and the service server 100 . The above-mentioned communication connection is not limited to the connection method, and may be directly or indirectly connected by wired communication, wireless communication, or other methods, which are not limited in this application.

Each user terminal in the user terminal cluster can install a target application, and when the target application runs on each user terminal, it can perform data interaction with the service server 100 respectively. Wherein, the target application may include an application with the function of displaying data information such as text, image, audio or video (such as game application, video editing application, social application, instant messaging application, live broadcast application, short video application, video application, music application, etc. , shopping application, novel application, payment application, browser, etc.) in one or more applications. The service server 100 may respond to the promotion request for the service resource, and send the service resource to one or more user terminals in the above-mentioned user terminal cluster. Wherein, the business resources may be resources used to disseminate product or service information to consumers or users, such as job recruitment advertisements, commodity sales advertisements, movie promotion advertisements, game recommendation advertisements, and the like. After receiving the service resource sent by the service server 100, one or more user terminals can load and display the service resource through the target application, and then collect the user's operation behavior on the service resource through the user terminal, one or more user terminals The terminal returns the display information and operation behavior information of the service resource to the service server 100 as service data information. Wherein, the display information may include information such as the display times and display duration of the service resource, and the operation behavior may include no operation behavior, early closing behavior, resource clicking behavior, and the like. The service server 100 receives service data information returned by one or more user terminals, and records it in a log for the service resource.

It should be noted that the service server 100 may respond to promotion requests of multiple service resources, and send multiple service resources to one or more user terminals in the user terminal cluster at the same time period. For example, the service server 100 can send service resource A to user terminal 200a, user terminal 200b, and user terminal 200c in the user terminal cluster, and send service resource B to user terminal 200b and user terminal 200b in the user terminal cluster within one day. 200n. Send the service resource C to the user terminal 200c in the user terminal cluster. The target applications installed in multiple user terminals receiving the same service resource may be different. For example, the target application in the user terminal 200a is a short video application X, the target application in the user terminal 200b is a social application Y, the target application in the user terminal 200c is a live application Z, the user terminal 200a, the user terminal 200b, and the user terminal 200c Both can receive the service resource A transmitted from the service server 100, and then display the service resource A through their respective target applications. Among them, the short video application X, the social application Y and the live broadcast application Z belong to different types of promotion applications, which can be called different site collections. For example, the short video application X can be site collection 1, and the social application Y can be site collection 2. Live application Z can be site set 3.

Correspondingly, when the service server 100 responds to the promotion request of the target service resource (the target service resource is the service resource sent to the user terminal above), it will predict the ECPM of the target service resource according to the conversion bid of the target service resource. , and then send the target service resource to the corresponding user terminal. There are often differences between the predicted ECPM and the actual ECPM. Therefore, in the initial promotion stage of the target service resource, the service server 100 will perform data calibration on the estimated joint factor of the ECPM that affects the target service resource, and then pass the calibrated forecast. The ECPM of the target business resource is adjusted based on the estimated conversion rate and joint factors, so that the adjusted ECPM is closer to the actual value. The specific process of data calibration is as follows:

The business server 100 will obtain the resource attribute information of the target business resource under N resource attribute types, and then cluster each historical business resource in the historical business resource set based on the S combination types to obtain H historical business resource subsets , and perform aggregation statistical processing based on the H historical business resource sub-sets to obtain the aggregated data set, and then the business server 100 will obtain the effective conversion number and effective joint factor for the target business resource in the aggregated data set according to the resource attribute information, according to The effective conversion number and the effective joint factor determine the calibration coefficient, and finally the estimated joint factor of the target business resource is calibrated according to the calibration coefficient. Wherein, N is a positive integer. Among them, the resource attribute types in each combination type belong to N resource attribute types; the historical resource attribute combinations of each historical business resource in a historical business resource subset are the same; a historical resource attribute combination corresponds to a resource attribute of a combination type Types are associated; S is a positive integer less than or equal to N; H is a positive integer. Among them, the aggregation data set includes the conversion numbers and joint factors corresponding to the H historical business resource sub-sets respectively; Each historical business resource in the historical business resource subset is determined by the corresponding estimated conversion rate and industry factor. Wherein, the estimated joint factor is determined according to the estimated conversion rate of the target business resource and the industry factor.

It can be understood that the method provided in the embodiment of the present application can be executed by a computer device, and the computer device includes but is not limited to a terminal or a server. Among them, the server can be an independent physical server, or a server cluster or distributed system composed of multiple physical servers, and can also provide cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communication , middleware services, domain name services, security services, CDN, and cloud servers for basic cloud computing services such as big data and artificial intelligence platforms.

It can be understood that the above-mentioned devices (such as the above-mentioned service server 100, user terminal 200a, user terminal 200b, user terminal 200c, ..., user terminal 200n) may be nodes in a distributed system, wherein the distributed system may be a district A block chain system, the block chain system may be a distributed system formed by connecting multiple nodes through network communication. Among them, the peer-to-peer (P2P, Peer To Peer) network that can be formed between nodes, the P2P protocol is an application layer protocol that runs on the Transmission Control Protocol (TCP, Transmission Control Protocol) protocol. In a distributed system, any form of computer equipment, such as servers, user terminals and other electronic equipment, can become a node in the blockchain system by joining the peer-to-peer network.

Wherein, the user terminal 200a, the user terminal 200b, the user terminal 200c, ..., and the user terminal 200n in FIG. POS (Point Of Sales, point of sale) machines, wearable devices (such as smart watches, smart bracelets, etc.), etc.

In order to facilitate the understanding of the above data calibration process, the calibration of the estimated combination factor of the target service resource W by the service server is used as an example for illustration.

Please refer to FIG. 2a-FIG. 2b together, which are schematic diagrams of a data calibration scenario provided by an embodiment of the present application. The realization process of the data calibration scene can be carried out in the service server 100 shown in FIG. 1, and can also be carried out in the user terminal (the One or more) can also be executed by the user terminal cluster and the service server in cooperation, which is not limited here. The embodiment of the present application takes the user terminal cluster and the service server 100 as an example for illustration.

As shown in Figure 2a, the service server 100 sends the historical service resource A1, the historical service resource A2, and the historical service resource A3 in the historical service resource set 300 to the user terminal 200a, the user terminal 200b, the user terminal 200c, ..., the user respectively. One or more user terminals in terminal 200n. Among them, historical business resource A1 can be a lipstick sale advertisement of brand x that advertiser A wants to promote, historical business resource A2 can be a game application recommendation advertisement of brand y that advertiser B wants to promote, and historical business resource A3 can be an advertisement The game application recommendation advertisement of the brand y that the main A wants to promote; in short, each historical business resource in the historical business resource collection can be the advertising resource of the product or service information that the advertiser wants to spread to consumers or users. Wherein, one historical service resource can be sent to different user terminals, and one user terminal can receive different historical service resources. For example, historical service resource A1 can be sent to user terminal 200a and user terminal 200b, and user terminal 200a can receive historical service resources. Resource A2, historical business resource A3. Then, each user terminal displays the received historical service resources through its installed target application. Wherein, the target application refers to the application that can display the advertisement resources sent by the service server 100, and may include one or more different types of applications. As shown in FIG. 2a, the target application may include application B1, application B2, application B3, ... . Application Bn, wherein application B1 may be a live broadcast application, application B2 may be a social application, application B3 may be a short video application, and so on. The target application in each user terminal is a carrier for displaying service resources. The same application can be installed in different user terminals. For example, user terminal 200a can be installed with application B1 and application B2, and user terminal 200b can also be installed with application B1.

As shown in FIG. 2a, the service server 100 also receives service data information for each historical service resource returned by each user terminal. Wherein, the service data information may include display information of historical service resources and operation behavior information for the historical service resources. For example, user a has a binding relationship with user terminal 200a. After user terminal 200a receives historical service resource A1, it displays historical service resource A1 through application B1 that user a is using. User a browses historical service resource A1. , is very interested in the lipstick promoted by the historical business resource A1, clicks on the purchase link provided by the historical business resource A1, and jumps to the shopping interface. The user terminal 200a will record the number and time of displaying the historical service resource A1 through the application B1, and also record operations such as clicking or closing the historical service resource A1 by the user a, and send it to the service server 100 together as service data information. The service server 100 will summarize and count the service data information of the historical service resource sent by each user terminal in the user terminal cluster, and record it in the log. For example, the target application of the historical service resource A1 in 100 user terminals has been displayed once , then the exposure corresponding to the historical business resource A1 in the log is 100, and 50 users click the purchase link provided by the historical business resource A1, then the corresponding click volume of the historical business resource A1 in the log is 50, and the corresponding click rate is 50 %. Correspondingly, if user a purchases the lipstick on the shopping interface, it can be considered that the historical business resource A1 has completed a conversion. The conversion of historical business resources refers to the specific optimization goal selected by the advertiser in the process of delivering business resources. For example, the conversion of the above-mentioned historical business resources A2 may be the user completing the registration of the game application. It should be noted that user terminals may not be able to collect data such as lipstick purchase data and game application registration data through the target application. Therefore, the number of conversions of historical business resources mostly depends on the feedback from the advertiser. The business server 100 receives the conversion data. will also be written to the log. Assuming that the conversion data sent back by the advertiser is that the number of conversions of the above-mentioned historical business resource A1 is 10, then the conversion rate corresponding to the historical business resource A1 is the percentage of the number of conversions to the number of clicks, that is, 20%.

It can be understood that the advertising platform completes the promotion of business resources for the advertiser through the service server 100, and naturally needs to charge the advertiser. For business resources that are charged based on the number of conversions, such as oCPA advertisements, the advertiser will give a conversion bid for the business resource (targetCpa, target Cost per Action), that is, the cost that the business resource is expected to convert once paid, and the advertiser is willing to pay The desired total cost for is conversions*targetCpa. However, the service server 100 charges based on the exposure amount*ECPM. In order to ensure that the actual charge is the same as the advertiser's expected total charge, the value of ECPM should be (conversion number/exposure amount)*targetCpa. Among them, the business server 100 can obtain the specific value of the exposure and targetCpa, but the conversion number depends on the feedback from the advertiser, and the business server 100 cannot determine the specific value of the conversion number before putting in business resources. Different prediction models predict the estimated click rate, estimated conversion rate, industry factors and other factors of the business resource, and then get the estimated conversion by exposure * estimated click rate * estimated conversion rate * industry factor * other factors and further obtain the ECPM of the business expense.

In the process of deploying business resources, there may be a deviation between the ECPM calculated based on the various factors predicted in the above method and the actual expected ECPM value, because the product of the estimated conversion rate and industry factors is different from the actual conversion rate and industry The error between factor products may be relatively large, so the service server 100 will calibrate the estimated joint factor when deploying service resources. Among them, the estimated joint factor is the product of the estimated conversion rate and the industry factor. Then calculate the new ECPM based on the calibrated estimated joint factor, and then charge the advertiser based on the new ECPM* exposure, so that the charge is closer to the total cost that the advertiser expects to pay for the business resource.

It will be described by taking the service server 100 as an example to calibrate the estimated joint factor of the newly launched target service resource by using the historical service resource in the above historical service resource set.

The service server 100 will obtain the resource attribute information associated with the target service resource and N resource attribute types, assuming that the N resource attribute types include advertisers and brands, as shown in Figure 2b, the service server 100 will obtain the target service resource C1 and the advertisement The resource attribute information associated with the owner and the brand, assuming that the advertiser of the target business resource C1 is B, which belongs to brand x. The service server 100 divides the historical service resource set based on the S combination types to obtain H historical service resource subsets. Wherein, each combination type includes one or more resource attribute types, and the resource attribute types in each combination type belong to the above N resource attribute types, as shown in Figure 2b, it can be assumed that the S combination types include [advertiser ] and [Brand], it is understandable that one of the S combination types can also be [Advertiser, Brand], and the setting of the combination type can be set according to the actual situation, here only [Advertiser] and [Brand] two combination types will be explained. The business server 100 will first divide the historical business resource set 300 according to the combination type of [advertiser]. It can be seen from the above that the advertiser of the historical business resource A1 and the historical business resource A3 is both A, and the advertiser of the historical business resource A2 is B, therefore, based on the combination type of [advertiser], the service server 100 can obtain a historical service resource subset 311 (including advertiser A's historical service resources) and a historical service resource subset 312 (including advertiser A's B’s historical business resource); similarly, based on the combination type of [brand], the historical business resource set 300 is divided, and the business server 100 can obtain the historical business resource subset 321 and the historical business resource subset 322, wherein the historical business resource Resource subset 321 includes historical business resource A1 belonging to brand x, and historical business resource subset 322 includes historical business resource A2 and historical business resource A3 belonging to brand y.

After dividing the historical service resource set 300 according to each combination type, the service server 100 performs aggregate statistical processing on each historical service resource subset to obtain an aggregated data set 400 . The aggregated data set includes conversion numbers and joint factors corresponding to each historical business resource subset. As shown in Figure 2b, m1 and n1 are the conversion number and joint factor corresponding to the historical business resource subset 311 respectively, m2 and n2 are the conversion number and joint factor corresponding to the historical business resource subset 312 respectively, and m3 and n3 are the historical business resource subset 312 respectively. The conversion number and combination factor corresponding to the business resource subset 321, m4 and n4 are respectively the conversion number and the combination factor corresponding to the historical business resource subset 322. Among them, the conversion number m1 can be the sum of the corresponding conversion numbers of the historical business resource A1 and the historical business resource A2, and the joint factor n1 is based on the estimated conversion rate and industry factor of the historical business resource A1 and the estimated historical business resource A2 The conversion rate and industry factors are determined; among them, the estimated conversion rate refers to the estimated value of the probability of conversion after the historical business resource is clicked, and the industry factor is the adjusted ECPM that can be used by business resources belonging to a specific industry factor. Similarly, the conversion number m2, the conversion number m3 and the conversion number m4 can also be determined according to the conversion numbers corresponding to the historical business resources included in their respective corresponding historical business resource subsets, the joint factor n2, the joint factor n3 and the joint factor n4 It may also be determined according to the estimated conversion rates and industry factors corresponding to the historical business resources included in their respective corresponding historical business resource subsets.

Then, the business server 100 will obtain the corresponding conversion number and joint factor in the aggregated data set 400 according to the resource attribute information. For example, the advertiser of the target business resource C1 is B. Among the obtained historical business resource subsets, the target business resource C1 corresponds to the historical business resource subset 312, and the business server 100 will obtain the conversion number m2 and the joint factor n2 corresponding to the historical business resource subset 312. Similarly, because the target business resource C1 brand is x, the business server will also determine that the target business resource C1 corresponds to the historical business resource subset 321, and then obtain the conversion number m3 and the joint factor n3 corresponding to the historical business resource subset 321, and then the business server 100 will calculate the conversion number m2 , joint factor n2, conversion number m3 and joint factor n3, determine the effective conversion number and effective joint factor, and then determine the calibration coefficient according to the effective conversion number and effective joint factor. Finally, the service server 100 calibrates the estimated combination factor of the target business resource C1 according to the calibration coefficient, wherein the estimated combination factor is determined according to the estimated conversion rate and the industry factor of the target business resource C1. By calibrating the estimated joint factor of the target business resource C1, the overall error caused by the estimated conversion rate of the target business resource C1 and industry factors can be reduced, so that a more accurate estimate can be obtained when calculating the ECPM, and the ECPM estimate can be improved. estimated accuracy.

Please refer to FIG. 3 . FIG. 3 is a schematic flowchart of a data calibration method provided in an embodiment of the present application. The method is executed by the computer equipment in FIG. 1, and the computer equipment may be the service server 100 in FIG. 1, or may be a user terminal in the user terminal cluster in FIG. 200c and user terminal 200n). As shown in FIG. 3 , the data calibration method may include the following steps S101-S105.

Step S101, acquiring resource attribute information of the target service resource under N resource attribute types; N is a positive integer.

Specifically, the target business resource may be an advertisement resource that the advertiser wants to disseminate product or service information to consumers or users, for example, an oCPA advertisement, and the presentation form of the oCPA advertisement may include text, pictures, videos, and so on. The essence of oCPA advertising is to pay according to user behavior. Advertisers can select specific optimization goals (such as activation of mobile applications, order placement on shopping websites, etc.) , and timely and accurately return the advertising conversion data to the advertising platform, the computer equipment will use the predictive model to estimate the fee that should be charged for each exposure, that is, get the estimated value of ECPM, and finally deduct the fee according to the exposure and ECPM.

In order to make the consumption data closer to the expected total cost of the advertiser and ensure the interests of both the advertiser and the advertising platform, the computer device will also adjust the estimated joint factor of the oCPA according to the data information of other advertisements related to this oCPA advertisement , to further adjust the ECPM. Among them, the consumption data is the total fee charged by the advertising platform to the advertiser, and the consumption data is equal to the product of the exposure and the ECPM; the expected total fee is the fee that the advertiser expects to pay, and the expected total fee is equal to the product of the targetCpa and the number of conversions. Among them, the estimated joint factor is the product of the estimated conversion rate and the industry factor, and the ECPM of the oCPA advertisement can be obtained according to the estimated joint factor, the estimated click rate, other factors and targetCpa. Wherein, the estimated conversion rate is the predicted probability that the oCPA advertisement will be converted after being clicked, that is, the probability that the oCPA advertisement will achieve the advertiser's optimization goal after being clicked. Among them, industry factors are usually designed for a specific industry or a specific group of people in a certain industry to enhance the effect of a specific industry or a specific group of people; PCVR compensating factor for e-commerce and strengthening factor for high-conversion crowd of direct e-commerce. Wherein, the estimated click-through rate is the predicted probability of the oCPA advertisement being clicked after being exposed. Among them, other factors refer to other factors that can affect ECPM, such as billing ratio factors, price adjustment factors, risk control factors, and so on.

Specifically, the N resource attribute types may include one or more of advertisers, product brands, products, site collections, new and old advertisements, and may also include other resource attribute types, such as same group, same region, same orientation, etc. , here only takes N resource attribute types as the above five resource attribute types as an example for illustration. The computer device obtains the resource attribute information of the target business resource under N resource attribute types, which may be, obtains advertiser information, product brand information, product information, site collection information, old and new advertisement information of the target business resource. For example, the advertiser of lipstick advertisement N0 is Xiaoming, the product brand is x, the product is lipstick, and the site collection is 27, which is a new advertisement. The strategy used to define the new and old advertisements can be that the advertisements that are only exposed today and have not been exposed before are new advertisements, and other advertisements are old advertisements; it can also be that the advertisements that have been exposed in the current 2 days are new advertisements, and other advertisements are old advertisements ; Old and new ads are defined without limitation.

Step S102, clustering the historical business resources in the historical business resource set based on the S combination types to obtain H historical business resource subsets; the resource attribute type corresponding to each combination type belongs to the N resource attributes type; the historical resource attribute combination of each historical business resource in a historical business resource subset is the same, and a historical resource attribute combination is associated with a resource attribute type corresponding to a combination type; S is a positive integer less than or equal to N; H is a positive integer.

Specifically, the set of historical business resources includes several historical business resources, wherein the historical business resources may be other advertising resources that have been placed by the above-mentioned advertising platforms. Assume that S combination types include combination type Mi, i is a positive integer less than or equal to S, the historical business resource set includes historical business resource Td, and d is a positive integer less than or equal to the total number of historical business resources in the historical business resource set , then the computer equipment clusters each historical business resource in the historical business resource set based on S combination types, and obtains the process of H historical business resource subsets, which can be: determine the resource attribute type contained in the combination type Mi is the target resource attribute type; the historical resource attribute information associated with the historical business resource Td in the historical business resource set and the target resource attribute type is determined as the historical resource attribute combination of the historical business resource Td; in the historical business resource set, the historical resource Each historical business resource with the same resource attribute combination is added to the same historical business resource sub-set to obtain one or more historical business resource sub-sets corresponding to the combination type Mi; use one or more historical business resource sub-sets corresponding to each combination type collections to form H sub-collections of historical business resources.

Specifically, the resource attribute types in the N resource attribute types can be divided into dimensional attribute types and granular attribute types, and different dimensional attribute types can be combined with all granular attribute types to obtain a combined type. For example, among advertisers, commodity brands, commodities, site collections, and new and old advertisements, advertisers, commodity brands, and commodities are dimension attribute types selected from the perspectives of the ownership, audience, and content of advertising resources, and the attribution of advertising resources Refers to the advertiser who created the advertisement. This advertiser reflects the group to which the advertisement resource belongs. All the advertisement resources under the same advertiser belong to the same advertiser, so they are similar to a certain extent; Advertisement The audience of the resource refers to the target of the advertisement. Among the various resource attribute types of the advertisement resource itself, the commodity brand can be used to represent the audience of the advertisement resource. The audience of the advertisement resource of the same commodity brand has certain similarities; The content of the resource refers to the product promoted by the advertising resource, and this content embodies the essence of the advertising resource itself. Therefore, the historical business resource set in the historical business resource set can be divided from the three dimensions of the advertiser, the commodity brand, and the commodity respectively. In the resource attribute type, site sets, new and old advertisements can be used as granular attribute types, which can be used to further divide the historical business resource collection obtained through the division of dimension attribute types. Therefore, the combination types of advertiser, product brand, product, site collection, and new and old advertisements can be obtained as [advertiser, site collection, new and old advertisement], [product brand, site collection, new and old advertisement], [product, site collection, new and old advertisement] advertise].

Specifically, taking [advertiser, site collection, new and old advertisement] as an example, the above-mentioned clustering of the historical service resource set according to the combination type Mi to obtain the historical service resource subset is described as an example. Please also refer to FIG. 4a. FIG. 4a is a schematic diagram of a scenario of clustering historical service resource collections provided by an embodiment of the present application. As shown in Figure 4a, the historical business resource set 400 includes historical business resource N1, historical business resource N2, and historical business resource N3, wherein the advertiser of historical business resource N1 is Xiaojia, and the site set 27 is a new advertisement; The advertiser of the business resource N2 is Xiao B, and the site set 29 is a new advertisement; the advertiser of the historical business resource N3 is Xiao A, and the site set 27 is an old advertisement. The computer device will use the three resource attribute types of advertiser, site collection, and new and old advertisements in [advertiser, site collection, new and old advertisement] as the target resource attribute type, and then associate historical business resources with the target resource attribute type Historical resource attribute information is determined as the historical resource attribute combination of historical business resources, then the historical resource attribute combination of historical business resource N1 is [Xiao Jia, 27, new]; the historical resource attribute combination of historical business resource N2 is [Xiao B , 29, new]; the historical resource attribute combination of historical business resource N3 is [Xiaojia, 27, new]. The combination of historical resource attributes of historical business resource N1 and historical business resource N3 is the same, therefore, as shown in Figure 4a, through [advertisers, site collections, old and new advertisements], computer equipment can be divided into historical business resource N1 and historical business resource The historical business resource subset 4001 of N3 includes the historical business resource subset 4002 of historical business resource N2. According to each combination type, the above-mentioned division process is performed on the historical service resource set, and a total of H historical service resource subsets can be obtained.

Step S103, performing aggregation statistical processing based on the H historical business resource subsets respectively to obtain an aggregated data set; the aggregated data set includes the conversion numbers and joint factors corresponding to each of the H historical business resource subsets; the The number of conversions is determined based on the number of conversions of each historical business resource in the corresponding historical business resource subset, and the associated joint factor is based on the estimated conversion rate and industry factor of each historical business resource in the corresponding historical business resource subset definite.

Specifically, a historical business resource corresponds to a conversion number, an estimated conversion rate, and an industry factor, and the joint factor of the historical business resource can be determined according to the estimated conversion rate and the industry factor; Aggregate the number of conversions corresponding to historical business resources to obtain the total number of conversions corresponding to all historical business resources in the historical business resource subset; perform aggregate statistics on the joint factor corresponding to each historical business resource in a historical business resource subset , the total joint factor corresponding to all historical business resources in the historical business resource subset can be obtained; then the above-mentioned total conversion number and total joint factor are respectively used as the conversion number and joint factor corresponding to the historical business resource subset, and added to the aggregated data concentrated.

Step S104, according to the resource attribute information, obtain the effective conversion number and effective combination factor for the target service resource in the aggregated data set.

Specifically, the computer device will extract S resource attribute combinations from the resource attribute information based on the S combination types, and then search the conversion numbers and joint factors corresponding to each of the S resource attribute combinations in the aggregation data set. Assuming that S resource attribute combinations include resource attribute combination Za, and a is a positive integer less than or equal to S, the process of finding the conversion numbers and joint factors corresponding to each of the S resource attribute combinations in the aggregation data set can be: In the historical resource attribute combination corresponding to the historical business resource subset, find the historical business resource subset that is the same as the resource attribute combination Za as the matching subset; obtain the conversion number and the joint factor corresponding to the matching subset in the aggregated data set as The conversion number and joint factor corresponding to the resource attribute combination Za.

Taking the above combination type as [advertiser, site collection, new and old advertisement] as an example, assuming that the resource attribute combination extracted from the resource attribute information of the target business resource based on this combination type is [Xiaojia, 27, new], the above is based on In the historical business resource subset 4001 and the historical business resource subset 4002 of [advertiser, site collection, new and old advertisement], the historical resource attribute combination corresponding to the historical business resource subset 4001 is the same as the resource attribute combination, so the computer device will The historical service resource subset 4001 is used as a matching subset, and then the conversion numbers and joint factors corresponding to the matching subset are obtained from the aggregated data set as the conversion numbers and joint factors corresponding to the resource attribute combination. Finally, the computer device will determine the effective conversion numbers and effective combination factors for the target business resource from the corresponding conversion numbers and combination factors of the S resource attribute combinations.

Specifically, the computer device will determine the effective conversion number and effective combination factor for the target business resource among the corresponding conversion numbers and combination factors of the S resource attribute combinations, which can be: according to the priorities of the S combination types, Determine the conversion numbers corresponding to the S resource attribute combinations and the priority of the joint factor; according to the consumption data corresponding to the S resource attribute combinations, determine the conversion numbers and the effectiveness of the joint factor corresponding to the S resource attribute combinations; Among the conversion numbers and joint factors corresponding to each resource attribute combination, the effective conversion numbers and joint factors are determined as candidate conversion numbers and candidate joint factors, and the candidate conversion numbers and candidate joint factors with the highest priority are used as targets for the target business. The number of effective conversions and effective joint factors for the resource. Wherein, the priority of the combination type can be set according to the actual situation, for example, the priority can be set as [advertiser], [commodity brand], [commodity] from high to low. Wherein, the conversion number and the effectiveness of the joint factor corresponding to a resource attribute combination can be determined according to whether the total consumption data of the historical service resource subset corresponding to the resource attribute combination is greater than four times the targetCpa.

For ease of understanding, please also refer to FIG. 4b. FIG. 4b is a schematic diagram of a determination scenario of an effective conversion number and an effective combination factor provided by an embodiment of the present application. As shown in Figure 4b, the historical business resource subset corresponding to the candidate conversion number m1 and the candidate joint factor n1 is the historical business resource subset Z1, and the combination type corresponding to the historical business resource subset Z1 is [advertiser]; the candidate conversion number m2 The historical business resource subset corresponding to the candidate joint factor n2 is the historical business resource subset Z2, and the combination type corresponding to the historical business resource subset Z2 is [commodity brand]; the candidate conversion number m3 and the historical business resource corresponding to the candidate joint factor n3 The subset is the historical business resource subset Z3, and the combination type corresponding to the historical business resource subset Z3 is [commodity].

Assuming that the priority of combination types from high to low is [advertiser], [commodity brand], [commodity], when the computer equipment determines the number of effective conversions and effective joint factors, it will first determine the historical business resource subset Z1 Whether the total consumption data of all historical business resources is greater than four times the total targetCpa of all historical business resources in the historical business resource subset Z1; Joint factor; if not, the computer device will determine whether the total consumption data of all historical business resources in the historical business resource subset Z2 is greater than four times the total targetCpa of all historical business resources in the historical business resource subset Z2, if yes, the computer device The candidate conversion number m2 and the candidate joint factor n2 will be used as the effective conversion number and the effective joint factor; if not, the computer device will determine whether the total consumption data of all historical business resources in the historical business resource subset Z3 is greater than four times the historical business resources The total targetCpa of all historical business resources in the subset Z3, if yes, the computer device will use the candidate conversion number m3 and the candidate joint factor n3 as the effective conversion number and effective joint factor.

Step S105, determine a calibration coefficient according to the effective conversion number and the effective combination factor, and calibrate the estimated combination factor of the target service resource according to the calibration coefficient; the estimated combination factor is based on the target service resource The resource's estimated conversion rate and industry factors are determined.

Specifically, the calibration coefficient can be calculated by the following formula (1):

cali_rate=Conv_valid/PCVRMulFactor_valid Formula (1)

Among them, cali_rate is the calibration coefficient, Conv_valid is the effective conversion number,

PCVRMulFactor_valid is the valid joint factor.

Specifically, the calibration process can refer to the following formula (2):

New_PCVRMulFactor＝old_PCVRMulFactor*cali_rate formula (2)

Among them, New_PCVRMulFactor is the estimated joint factor after calibration, and old_PCVRMulFactor is the estimated joint factor before calibration. The estimated joint factor before calibration is determined based on the estimated conversion rate and industry factor of the target business resource. Among them, the estimated conversion rate and industry factors of the target business resources can be predicted through corresponding prediction models.

Optionally, the above calibration process can only be applied to the initial promotion stage of the target business resource. At this time, after the computer device receives the calibration request for the target business resource, it will first determine the promotion stage of the target business resource; if the promotion stage of the target business resource In the initial promotion stage, the above-mentioned calibration process is executed in response to the calibration request of the target business resource. Because for the target business resources, in the initial promotion stage of the launch, since there is no historical data of the day (or the historical data is insufficient), its calibration cannot rely on the data of the target business resources themselves, and it is necessary to comprehensively consider the data of other historical business resources. . The initial promotion stage can be defined as a promotion stage in which the number of conversions of target business resources is less than or equal to 2 or the consumption is less than or equal to 2 times the targetCpa.

Using the method provided in the embodiment of the present application, the resource attribute information of the target business resource under N resource attribute types can be obtained, and then each historical business resource in the historical business resource set is clustered based on the S combination types to obtain H historical business resource sub-sets, and based on the H historical business resource sub-sets, perform aggregate statistical processing respectively to obtain an aggregated data set, and then the business server 100 will obtain the effective conversion data for the target business resource in the aggregated data set according to the resource attribute information and the effective joint factor, determine the calibration coefficient according to the effective conversion number and the effective joint factor, and finally calibrate the estimated joint factor of the target business resource according to the calibration coefficient. Wherein, N is a positive integer. Among them, the resource attribute types corresponding to each combination type belong to N resource attribute types; the historical resource attribute combinations of each historical business resource in a historical business resource subset are the same, and a historical resource attribute combination corresponds to a resource of a combination type Attribute types are associated; S is a positive integer less than or equal to N; H is a positive integer. Among them, the aggregated data set includes the conversion numbers and joint factors corresponding to each of the H historical business resource subsets; The estimated conversion rate and industry factor corresponding to each historical business resource in the historical business resource subset are determined. Wherein, the estimated joint factor is determined through the estimated conversion rate of the target business resource and the industry factor.

Using the method provided in this application, based on the resource attribute information of the target business resource, the effective conversion number and effective joint factor determined according to the historical business resources related to the target business resource can be obtained, and then the effective conversion number and effective joint factor can be used Calibrate the estimated joint factor of the target business resource, and adjust the estimated display cost of the target business resource according to the calibrated estimated joint factor, which can reduce the deviation between the estimated display cost and the actual value, Improve forecast accuracy.

Further, please refer to FIG. 5 , which is a schematic flowchart of a data calibration method provided by an embodiment of the present application. The method is executed by the computer equipment in FIG. 1, and the computer equipment can be the service server 100 in FIG. terminal 200c and user terminal 200n). As shown in FIG. 5 , the data calibration method may include the following steps S201-S208.

Step S201, acquiring resource attribute information of the target service resource under N resource attribute types; N is a positive integer.

Step S202, clustering the historical business resources in the historical business resource set based on the S combination types, to obtain H historical business resource subsets; the H historical business resource subsets include the historical business resource subset Nj, j is a positive integer less than or equal to H.

Specifically, for the specific implementation process of step S201 and step S02, reference may be made to the description of step S101 and step S102 in the above embodiment corresponding to FIG. 3 , which will not be repeated here.

Step S203, determine the first unit conversion number and the first unit combination factor corresponding to the historical business resource subset Nj; the first unit conversion number is based on the historical business resources in the historical business resource subset Nj The number of conversions within a unit of time is determined, and the first unit joint factor is determined according to the estimated conversion rate and industry factor of each historical business resource in the historical business resource subset Nj within the first unit of time .

Specifically, the computer device will determine each historical business resource in the historical business resource subset Nj as the historical business resource to be counted, and obtain the log information of the historical business resource to be counted; The number of conversions, estimated conversion rate, and industry factor within a unit of time; multiply the estimated conversion rate of the historical business resources to be counted within the first unit of time and the industry factor to obtain the historical business resources to be counted in the first unit of time Furthermore, the summation processing is performed on the joint factors of the historical business resources to be counted within the first unit duration to obtain the first unit joint factor corresponding to the historical business resource subset Nj. The computer equipment sums the conversion numbers of the historical business resources to be counted within the first unit duration to obtain the first unit conversion numbers corresponding to the historical business resource subset Nj.

Wherein, the duration of the first unit may be one minute, one hour, two hours, etc., which is not limited here. Wherein, the end time of the first unit duration is usually the current system time. For example, when the duration of the first unit is one hour, and the current system time is 9:00, then the number of conversions, estimated conversion rates, and industry factors of the historical business resources to be counted within the last hour refer to the period from 8:00 to During the time period of 9:00, the number of conversions, estimated conversion rates, and industry factors of historical business resources are to be counted.

Step S204, determine the second unit conversion number and the second unit combination factor corresponding to the historical business resource subset Nj; the second unit conversion number is based on the historical business resources in the historical business resource subset Nj The number of conversions within two units of time is determined, and the second unit joint factor is determined according to the estimated conversion rate and industry factor of each historical business resource in the historical business resource subset Nj within the second unit of time ; The second unit duration is greater than the first unit duration.

Specifically, the computer device will determine the second unit duration of the historical service resource subset Nj, wherein the second unit duration can be one hour, two hours, the whole day, etc. It should be noted that the second unit duration should be greater than the above The first unit duration. Then, the computer device divides the second unit duration based on the first unit duration to obtain at least two statistical periods; the duration of each statistical period is less than or equal to the first unit duration. For example, the first unit duration is one hour, and the second unit duration may be a whole day. Wherein, the whole day refers to the period between 0:00 today and the current system time. For example, the current system time is 7:00, and the second unit duration refers to the duration between 0:00 and 7:00 today. The computer device divides the second unit time according to the first unit time, and can get 0:00-1:00, 1:00-2:00, 2:00-3:00, 3:00-4:00, 4 :00-5:00, 5:00-6:00, 6:00-7:00 six statistical periods, each statistical period is one hour long. Then, the computer device will obtain the conversion number, estimated conversion rate, and industry factor of the historical business resources to be counted within each statistical period from the above log information; for each statistical period, the historical business resources to be counted within the statistical period The estimated conversion rate and the industry factor are multiplied to generate the joint factor of the historical business resources to be counted in the statistical period; then, the joint factors of the historical business resources to be counted in the statistical period are summed , to obtain the statistical period joint factor of the historical service resource subset Nj within the statistical period. For each statistical period, the computer equipment sums the conversion numbers of the historical business resources to be counted within the statistical period, and obtains the conversion numbers of the historical business resource subset Nj in each statistical period. Finally, according to the time attenuation strategy, the statistical period conversion number and statistical period joint factor of the historical business resource subset Nj in each statistical period are processed, and the second unit conversion number and the second unit conversion number corresponding to the historical business resource subset Nj are obtained. joint factor.

Specifically, it is assumed that the above-mentioned at least two statistical periods include a statistical period Lk, and k is a positive integer less than or equal to the total number of the at least two statistical periods; the start time of the statistical period Lk is earlier than the statistical period Lk+1 , then according to the time attenuation strategy mentioned above, the statistical period conversion number and the statistical period joint factor of the historical business resource subset Nj in each statistical period are processed, and the second unit conversion number and the second unit conversion number corresponding to the historical business resource subset Nj are obtained. The process of two joint factors can be: according to the time decay factor, the total quantity of at least two statistical periods, and the forward order of the start time of the statistical period Lk in at least two statistical periods, the statistics within the period Lk The number of conversions in the statistical period and the joint factor of the statistical period are respectively attenuated to obtain the number of attenuation conversions and the joint factor of attenuation; the number of attenuation conversions in each statistical period is summed to obtain the second unit corresponding to the historical business resource subset Nj The number of conversions; the attenuation joint factors in each statistical period are summed to obtain the second unit joint factor corresponding to the historical service resource subset Nj.

If the first unit duration is one hour and the second unit duration is a whole day, the above determination process can be referred to the following formulas (3) and (4):

Among them, Conv_advertiser_day refers to the second unit conversion number of the historical business resource subset Nj, that is, the total conversion number of the day; I is the current moment (hour); lambda is the time decay coefficient, which can be valued according to the actual situation, such as 0.05; Conv_advertiser_hour is The number of statistical period conversions within the statistical period Lk; PCVRMulFactor_advertiser_day refers to the second unit joint factor of the historical business resource subset Nj, that is, the total joint factor of the day, that is, the sum of the total (PCVR*industry factors); PCVRMutor_advertiser_hour _k is the statistical period Statistical period joint factor within Lk.

Step S205, obtaining the first unit consumption data of the historical business resource subset Nj; according to the first unit consumption, the first unit conversion number, the first unit combination factor, the second unit conversion number and the second unit combination factor , to determine the conversion number and joint factor corresponding to the historical business resource subset Nj.

Specifically, if the first unit consumption data belongs to sufficient consumption data, the first unit conversion number is taken as the conversion number corresponding to the historical business resource subset Nj, and the first unit joint factor is taken as the joint factor corresponding to the historical business resource subset Nj ; If the first unit consumption data belongs to insufficient consumption data, the second unit conversion number is used as the conversion number corresponding to the historical business resource subset Nj, and the second unit joint factor is used as the joint factor corresponding to the historical business resource subset Nj. Wherein, the first unit consumption data refers to the sum of the consumption data of each historical service resource in the historical service resource subset Nj within the first unit duration. The above process can be referred to formula (5) and formula (6):

Among them, Conv_advertiser refers to the conversion number corresponding to the historical service resource subset Nj, and PCVRMulFactor_advertiser refers to the joint factor corresponding to the historical service resource subset Nj.

Optionally, the process for the computer device to determine whether the first unit consumption data belongs to the sufficient consumption data may be as follows: obtaining the conversion transaction value data, and determining the sufficient data threshold according to the conversion transaction value data; if the first unit consumption data is greater than the sufficient data threshold , it is determined that the first unit consumption data belongs to sufficient consumption data; if the first unit consumption data is less than or equal to the sufficient data threshold, it is determined that the first unit consumption data belongs to insufficient consumption data. Wherein, the conversion transaction value data is the above-mentioned conversion bid targetCpa. Wherein, the sufficient data threshold may be equal to 4 times targetCpa.

Step S206, when the conversion numbers and joint factors corresponding to the H historical business resource subsets are obtained, generate an aggregated data set including the respective conversion numbers and joint factors corresponding to the H historical business resource subsets.

Step S207, according to the resource attribute information, obtain the effective conversion number and effective combination factor for the target service resource in the aggregated data set.

Step S208, determine a calibration coefficient according to the effective conversion number and the effective combination factor, and calibrate the estimated combination factor of the target service resource according to the calibration coefficient; the estimated combination factor is based on the target service The resource's estimated conversion rate and industry factors are determined.

Specifically, for the specific implementation process of steps S207 to S208, reference may be made to steps S104 to S105 in the above embodiment corresponding to FIG. 3 , which will not be repeated here.

Through the method provided by the embodiment of this application, in the process of calibrating the estimated joint factor of the target business resource, the first unit conversion number and the first unit corresponding to each historical business subset Nj can be obtained when determining the aggregated data set Joint factor, second unit conversion number and second unit joint factor, and then by judging whether the first unit consumption data is sufficient consumption data, from the first unit conversion number corresponding to each historical business resource subset and the first unit joint From the factor, the second unit conversion number and the second unit joint factor, select the conversion number and the joint factor of the historical business resource subset, and then obtain the aggregated data set. By adopting the method provided in the embodiment of the present application, the timeliness and effectiveness of the conversion numbers and joint factors in the aggregated data set can be improved, thereby improving the accuracy of the calibration coefficients and finally improving the prediction accuracy.

Further, please refer to FIG. 6 . FIG. 6 is a schematic flowchart of an analysis method for determining a combination factor of display costs provided by an embodiment of the present application. The method is executed by the computer equipment in FIG. 1, and the computer equipment can be the service server 100 in FIG. terminal 200c and user terminal 200n). As shown in FIG. 6, the data calibration method may include the following steps S301-S308.

Step S301, in the set of historical business resources, according to the division granularity of historical business resources, the historical business resources whose expected consumption data is smaller than the actual consumption data are regarded as resources to be processed, and the resources to be processed are added to the set of resources to be processed; The resource set to be processed includes resource Sr to be processed, and r is a positive integer less than or equal to the total number of resources to be processed in the resource set to be processed.

Specifically, the computer device will divide the historical business resources with the same granularity information in the historical business resource set together, where the granularity information is the information associated with the division granularity of the historical resources; and then obtain the expected Historical business resources whose consumption data is smaller than the actual consumption data are regarded as pending resources, and the pending resources are added to the pending resource collection. Wherein, the granularity information of the historical service resources included in each resource to be processed is the same. Among them, the granularity of historical business resources can be resource granularity, account granularity, group granularity, etc. Business resource O2, historical business resource O3 of group A, and historical business resource O4 of group A, then the historical business resources after division are {historical business resource O1, historical business resource O3, historical business resource O4} and {historical business resource O2 }, if only the expected consumption data of the historical business resource O3 is greater than the actual consumption data, the final two resources to be processed, one resource to be processed includes the historical business resource O1 and the historical business resource O4, and the other resource to be processed includes the historical business resource Resource O2.

Specifically, the expected consumption data refers to the expected total cost GMV (GMV, Guaranteed Minimum Value) mentioned in the embodiment corresponding to Figure 3 above, that is, the cost that the advertiser expects to pay. If the advertiser bids targetCpa according to the conversion, then according to the formula ( 7) GMV can be calculated:

GMV＝targetCpa*conversion number formula Formula (7)

The actual consumption data refers to the consumption (Cost) mentioned in the above-mentioned embodiment corresponding to Figure 3, and the consumption data can be calculated according to formula (8):

Cost＝ECPM*exposure formula Formula (8)

Among them, ECPM is the display fee. In an ideal situation, Cost = GMV, which can achieve mutual benefits between the advertising platform and the advertiser, that is, the advertising platform neither charges more money nor charges less money. However, as the advertising system becomes more and more complex and there are more and more evolutionary versions, GMV<Cost often means that the expected consumption data is less than the actual consumption data. This situation is that the platform overcharges, resulting in over-costs, which is not good for advertisers. Therefore, when analyzing the joint factor of the display fee, the computer device can select historical service resources whose expected consumption data is smaller than the actual consumption data in the historical service data set.

Step S302, according to the price adjustment factor, risk control factor, estimated conversion rate, actual conversion rate, estimated cost ratio factor, actual billing ratio factor, estimated click rate, actual click rate and industry of the resource Sr to be processed factor to determine the control value, conversion rate ratio, billing ratio factor ratio, click-through rate ratio and industry factor of the resource Sr to be processed.

Specifically, the above ECPM can be calculated by formula (9), namely:

ECPM＝targetCpa*all_factor*pcvr*pctr formula (9)

Among them, targetCpa is the conversion bid of the advertiser, and all_factor is a comprehensive factor, including (price adjustment factor*risk control factor)*gsp_factor*industry_factor. Among them, (price adjustment factor * risk control factor) can be regarded as a whole, that is, the regulation factor for targetCpa; gap_factor refers to the billing ratio factor, usually caused by gsp; industry_factor refers to industry factors, and different industries have different industry factors, such as Industry factors in direct e-commerce include e-commerce industry factors and crowd weighting factors; pcvr refers to estimated conversion rate; pctr refers to estimated click-through rate. By comparing the above formula (7), formula (8) and formula (9), we can see that GMV<Cost is caused by the high product of (price adjustment factor*risk control factor)*gsp_factor*industry_factor*pcvr*pctr. The more accurate the factor itself is, the smaller the impact it will have on ECPM. However, these factors have their own characteristics, and the accuracy of their predictions is also different. It is necessary to analyze the accuracy of each factor from the data of the corresponding resources to be processed.

Specifically, the computer device will first determine the estimated conversion rate, estimated cost ratio factor, and estimated click rate of the resource Sr to be processed; then obtain the actual price adjustment factor, actual risk control factor, and Actual conversion rate, actual billing ratio factor, actual click-through rate, and actual industry factor. Wherein, the estimated conversion rate, the estimated cost ratio factor, and the estimated click-through rate can be determined through respective corresponding prediction models, and the prediction model can be a deep learning model. Then, the computer device can determine the control value of the resource Sr to be processed according to the product of the actual price adjustment factor and the actual risk control factor; the computer device can determine the conversion rate ratio of the resource Sr to be processed according to the ratio of the estimated conversion rate to the actual conversion rate ; The computer device can determine the ratio of the billing ratio factor of the resource Sr to be processed according to the ratio of the expected billing ratio factor to the actual billing ratio factor; the computer device can determine the resource to be processed according to the ratio of the estimated click rate to the actual click rate The click rate ratio of Sr; the computer device can determine the industry factor value of the resource Sr to be processed according to the industry factor.

Step S303, acquiring a value range.

Specifically, the ideal values of the above control value, conversion rate ratio, billing ratio factor ratio, click-through rate ratio and industry factor are all 1, so the range including 1 can be divided to obtain the value range. A feasible area division method is [0 0.5 0.7. 0.9 1.0 1.1 1.3 ∞], where ∞ means infinite. The divided value range may include [0, 0.5), [0.5, 0.7), [0.7, 0.9), [0.9, 1.0), [1.0, 1.1), [1.1, 1.3), [1.3, ∞).

Step S304, in the set of resources to be processed, obtain resources to be processed whose conversion number is greater than or equal to the first conversion threshold and whose conversion number is less than the second conversion threshold, as the first conversion resource; the second conversion threshold is greater than The first conversion threshold.

Specifically, the first conversion threshold and the second conversion threshold corresponding to different historical business resource division granularities may be different, and in different industrial advertising scenarios, the first conversion threshold and the second conversion threshold may also be different.

Step S305, according to the control value, conversion rate ratio, billing ratio factor ratio, click-through rate ratio, industry factor and the value range corresponding to the first conversion resource, determine the first control analysis ratio, the first conversion rate analysis Ratio, First Billing Ratio Factor Analysis Ratio, First Click Rate Analysis Ratio, and First Industry Factor Analysis Ratio.

Specifically, assuming that the value interval includes the value interval Gb, and b is a positive integer less than or equal to the total number of value intervals, then the computer device calculates the first conversion resource corresponding to the control value, conversion rate ratio, billing ratio factor ratio, Click-through rate ratio, industry factor and value interval, the process of determining the first control analysis ratio, the first conversion rate analysis ratio, the first billing ratio factor analysis ratio, the first click-through rate analysis ratio and the first industry factor analysis ratio, It may be as follows: the computer device determines the first resource quantity of the resource to be processed whose corresponding control value in the first conversion resource belongs to the value interval Gb, and the ratio of the first resource quantity to the total quantity of the resource to be processed in the first conversion resource, As the first regulation and analysis ratio corresponding to the value interval Gb; the computer device determines that the conversion rate ratio corresponding to the low-conversion resource belongs to the second resource quantity of the resource to be processed in the value interval Gb, and the second resource quantity and the first conversion resource The ratio of the total quantity of resources to be processed is used as the first conversion rate analysis ratio corresponding to the value range Gb; the computer device determines that the corresponding billing ratio factor ratio in the first conversion resource belongs to the value range Gb of the resources to be processed Three resource quantities, the ratio of the third resource quantity and the total quantity of resources to be processed in the first conversion resource is used as the first billing ratio factor analysis ratio corresponding to the value interval Gb; the computer device determines the corresponding first conversion resource The click rate ratio belongs to the fourth resource quantity of resources to be processed in the value interval Gb, and the ratio of the fourth resource quantity to the total quantity of resources to be processed in the first conversion resource is used as the first click rate analysis corresponding to the value interval Gb Ratio; the computer equipment determines that the corresponding industry factor in the low conversion resource belongs to the fifth resource quantity of the resource to be processed in the value interval Gb, and the ratio of the fifth resource quantity and the total quantity of the resource to be processed in the first conversion resource is taken as The factor analysis ratio of the first industry corresponding to the value interval Gb.

Step S306, in the set of resources to be processed, obtain the resources to be processed whose conversion number is greater than or equal to the second conversion threshold, as the second conversion resources.

Step S307, according to the control value, conversion rate ratio, billing ratio factor ratio, click-through rate ratio, industry factor and the value range corresponding to the second conversion resource, determine the second control analysis ratio and the second conversion rate analysis Ratio, Second Billing Ratio Factor Analysis Ratio, Second Click Rate Analysis Ratio, and Second Industry Factor Analysis Ratio.

Specifically, for the specific implementation of step S306 and step S307, reference may be made to the above-mentioned step S304 and step S306, which will not be repeated here.

Step S308, analyzing the first control analysis ratio, the first conversion rate analysis ratio, the first billing ratio factor analysis ratio, the first click-through rate analysis ratio, the first industry factor analysis ratio, the second regulation analysis ratio, the second Conversion rate analysis ratio, the second billing ratio factor analysis ratio, the second click-through rate analysis ratio and the second industry factor analysis ratio are analyzed and processed to determine the impact factor used to adjust the expected income of the display; the impact factor includes estimated conversion rate and an industry factor, and the estimated conversion rate and the industry factor are jointly used to generate the estimated joint factor.

In order to better understand the above step S301-step S308, the analysis of the advertisement set in the direct-operated e-commerce scenario at the resource granularity, account granularity, and group granularity is used as an example to illustrate.

First, analyze the deviation process of each factor from the perspective of resource granularity as follows:

About the price adjustment factor * risk control factor: To study the scope of the price adjustment factor, the price adjustment factor * risk control factor can be taken as a whole. First, take out all the advertisements with GMV<Cost (that is, the above-mentioned resources to be processed), which are divided into the following two situations:

Advertisements with conversions greater than or equal to 1 and less than 6 (i.e. the first conversion resource):

The average value of price adjustment factor * risk control factor (that is, the above-mentioned control value): 1.0055626343375126.

The proportion of the control value falling in the corresponding value range is shown in Table 1 below:

Table 1

Ads with conversions greater than or equal to 6 (that is, the second conversion resource):

The average value of price adjustment factor * risk control factor (that is, the above-mentioned control value): 1.0151019986530374.

The proportion of the control value falling in the corresponding value range is shown in Table 2 below:

Table 2

About the estimated conversion rate pcvr: investigate the accuracy of pcvr estimation, compare pcvr and the actual conversion rate of the advertisement, and determine the closeness of the conversion rate ratio to 1. The closer the conversion rate ratio is to 1, the closer the pcvr is to the actual conversion rate, and the higher the accuracy of pcvr.

First, take out all the advertisements with GMV<Cost, which are divided into the following two situations:

Mean conversion rate ratio: 1.2929644284281734.

The proportion of the conversion rate ratio falling within the corresponding value range is shown in Table 3 below:

table 3

Mean conversion rate ratio: 1.0258310512901163.

The proportion of the conversion rate ratio falling within the corresponding value range is shown in Table 4 below:

Table 4

Regarding the billing ratio factor gsp_factor: investigate the accuracy of gsp_factor, divide the estimated gsp_factor (the above-mentioned estimated cost ratio factor) by the actual statistical gsp_factor (the above-mentioned actual billing ratio factor) to obtain the ratio of the billing ratio factor, and determine How close the billing ratio factor is to 1. The closer the billing ratio factor ratio is to 1, the higher the accuracy of the estimated gsp_factor is.

The average ratio of billing ratio factor: 1.005334324931539.

The ratio of the billing ratio factor falls within the corresponding value range as shown in Table 5 below:

table 5

The average ratio of the billing ratio factor: 0.9984545757710465.

The ratio of the billing ratio factor falls within the corresponding value range as shown in Table 6 below:

Table 6

About the estimated click-through rate pctr: Investigate the accuracy of pctr estimates, calculate the ratio of the estimated pctr to the actual click-through rate, obtain the click-through rate ratio, and determine the degree to which the click-through rate ratio is close to 1. The closer the click-through rate ratio is to 1, the higher the accuracy of the estimated pctr.

Average CTR ratio: 1.030886270371121.

The proportion of the click-through rate ratio falling within the corresponding value range is shown in Table 7 below:

Table 7

Average CTR ratio: 1.0017268891863382.

The proportion of the click-through rate ratio falling within the corresponding value range is shown in Table 8 below:

Table 8

About the industry factor: After calculating the industry factor (e-commerce industry factor * crowd weighting factor), determine the degree to which the industry factor is close to 1.

The mean of the industry factor: 1.1719941122774267.

The proportion of industry factors falling within the corresponding value range is shown in Table 9 below:

Table 9

The mean of the industry factor: 1.1646948281407095.

The proportion of industry factors falling in the corresponding value range is shown in Table 10 below:

Table 10

From the above data, it can be clearly seen that the inaccurate estimation of the pcvr factor is very obvious in advertisements with low conversion numbers. The proportion of pcvr falling in [1.3, ∞) is as high as 0.35306299, and the average value of pcvr is also greater than 1, indicating that pcvr It is often greater than the actual conversion rate, which will bring a certain degree of explosive cost, that is, the actual consumption is much greater than the expected total cost; however, in the advertisement with a high conversion number, the pcvr estimation is relatively accurate, and the pcvr falls in [0.7, 0.9), The proportions of [0.9, 1.0), [1.0, 1.1) are relatively high, and the mean value of pcvr is also relatively close to 1. Whether the industry factor is in the advertisement with low conversion number or the advertisement with high conversion number, the proportion of [1.1, 1.3) and [1.3, ∞) is relatively large, and the average value of the industry factor is obviously greater than 1, which will lead to explosive costs . According to the proportion and average value of other factors in the value range, it can be determined that their impact on ECPM is small, which can be understood as a higher prediction accuracy.

Then analyze the deviation process of each factor from the perspective of account granularity as follows:

About the price adjustment factor * risk control factor: To study the scope of the price adjustment factor, the price adjustment factor * risk control factor can be taken as a whole. First, take out all accounts with GMV<Cost (one or more advertisements corresponding to the same account, that is, the resource to be processed above), which can be divided into the following two cases:

Accounts with a conversion number greater than or equal to 1 and less than 10 (that is, the conversion number of an advertisement corresponding to an account is greater than or equal to 1 and less than 10):

The average value of price adjustment factor * risk control factor (that is, the above-mentioned control value): 1.0270681225915883.

The proportion of the control value falling within the corresponding value range is shown in Table 11 below:

Table 11

Accounts with conversions greater than or equal to 10:

The average value of price adjustment factor * risk control factor (that is, the above-mentioned control value): 1.0163434608061954.

The proportion of the control value falling within the corresponding value range is shown in Table 12 below:

Table 12

About the estimated conversion rate pcvr: investigate the accuracy of advertising pcvr estimates under the same account, compare the estimated pcvr with the actual conversion rate, and determine how close the conversion rate ratio is to 1. The closer the conversion rate ratio is to 1, the closer the pcvr is to the actual conversion rate, and the higher the accuracy of pcvr.

First, take out all accounts with GMV<Cost, which can be divided into the following two situations:

Accounts with conversions greater than or equal to 1 and less than 10:

Mean conversion rate ratio: 1.5795300001182175.

The proportion of the conversion rate ratio falling within the corresponding value range is shown in Table 13 below:

Table 13

Accounts with conversions greater than or equal to 10:

Mean conversion ratio: 1.1302273285339024.

The proportion of the conversion rate ratio falling within the corresponding value range is shown in Table 14 below:

Table 14

About the billing ratio factor gsp_factor: investigate the accuracy of gsp_factor, and compare the estimated gsp_factor (the above-mentioned estimated cost ratio factor) with the actual statistical gsp_factor (the above-mentioned actual billing ratio factor) to obtain the ratio of the billing ratio factor, Determines how close the billing ratio factor ratio is to 1. The closer the billing ratio factor ratio is to 1, the higher the accuracy of the estimated gsp_factor is.

Accounts with conversions greater than or equal to 1 and less than 10:

The average ratio of billing ratio factor: 1.0063416604925397.

The ratio of the billing ratio factor falling within the corresponding value range is shown in Table 15 below:

Table 15

Accounts with conversions greater than or equal to 10:

The average ratio of billing ratio factor: 1.0013801378217713.

The ratio of the billing ratio factor falling within the corresponding value range is shown in Table 16 below:

Table 16

About the estimated click-through rate pctr: investigate the accuracy of pctr estimation, compare the estimated pctr with the actual click-through rate, obtain the click-through rate ratio, and determine the degree to which the click-through rate ratio is close to 1. The closer the click-through rate ratio is to 1, the higher the accuracy of the estimated pctr.

Accounts with conversions greater than or equal to 1 and less than 10:

Average CTR ratio: 1.030300842027766.

The proportion of the click-through rate ratio falling within the corresponding value range is shown in Table 17 below:

Table 17

Accounts with conversions greater than or equal to 10:

Average CTR ratio: 1.0189707247565336.

The proportion of the click-through rate ratio falling within the corresponding value range is shown in Table 18 below:

Table 18

Accounts with conversions greater than or equal to 1 and less than 10:

The mean of the industry factor: 1.057864621100866.

The proportion of industry factors falling within the corresponding value range is shown in Table 19 below:

Table 19

Accounts with conversions greater than or equal to 10:

The mean of the industry factor: 1.0942537338421843.

The proportion of industry factors falling within the corresponding value range is shown in Table 20 below:

Table 20

From the above data, it can be clearly seen that the inaccurate estimation of the pcvr factor is very obvious in accounts with low conversion numbers, and it is overestimated, which will bring a certain degree of explosion costs; in accounts with higher conversion numbers In , the situation is slightly better, but there are also large deviations. The industry factor also has a certain deviation whether it is low conversion number or high conversion number, and it is obviously larger than other factors.

Finally, the process of analyzing the deviation of each factor from the perspective of group granularity is as follows:

About the price adjustment factor * risk control factor: To study the scope of the price adjustment factor, the price adjustment factor * risk control factor can be taken as a whole. First, take out all groups with GMV<Cost (one or more advertisements corresponding to the same group, that is, the resource to be processed above), which can be divided into the following two situations:

A group whose conversion number is greater than or equal to 1 and less than 10 (that is, the conversion number of an advertisement corresponding to a group is greater than or equal to 1 and less than 10):

The average value of price adjustment factor * risk control factor (that is, the above-mentioned control value): 1.034216183755146.

The proportion of the control value falling within the corresponding value range is shown in Table 21 below:

Table 21

Groups with conversions greater than or equal to 10:

The average value of price adjustment factor * risk control factor (that is, the above-mentioned control value): 1.0296622775279651.

The proportion of the control value falling within the corresponding value range is shown in Table 22 below:

Table 22

About the estimated conversion rate pcvr: investigate the accuracy of the estimated pcvr of advertising under the same group, compare the estimated pcvr with the actual conversion rate, and determine the closeness of the conversion rate ratio to 1. The closer the conversion rate ratio is to 1, the closer the pcvr is to the actual conversion rate, and the higher the accuracy of pcvr.

First, take out all the GMV<Cost groups, which are divided into the following two situations:

Groups with conversion numbers greater than or equal to 1 and less than 10:

Mean conversion ratio: 1.5079355403667978.

The proportion of the conversion rate ratio falling within the corresponding value range is shown in Table 23 below:

Table 23

Groups with conversions greater than or equal to 10:

Mean conversion rate ratio: 1.1648124771184638.

The proportion of the conversion rate ratio falling within the corresponding value range is shown in Table 24 below:

Table 24

Groups with conversion numbers greater than or equal to 1 and less than 10:

Billing ratio factor ratio mean: 1.007077182551767.

The proportion of the billing ratio factor falling within the corresponding value range is shown in Table 25 below:

Table 25

Groups with conversions greater than or equal to 10:

The average ratio of billing ratio factor: 1.0015587163937008.

The ratio of the billing ratio factor falls within the corresponding value range as shown in Table 26 below:

Table 26

About the estimated click-through rate pctr: investigate the accuracy of pctr estimates, compare the estimated pctr with the actual click-through rate, obtain the click-through rate ratio, and determine the degree to which the click-through rate ratio is close to 1. The closer the click-through rate ratio is to 1, the higher the accuracy of the estimated pctr.

Groups with conversion numbers greater than or equal to 1 and less than 10:

Average CTR ratio: 1.0260427036634276.

The proportion of the click-through rate ratio falling within the corresponding value range is shown in Table 27 below:

Table 27

Groups with conversions greater than or equal to 10:

Average CTR ratio: 1.0238124641852109.

The proportion of the click-through rate ratio falling within the corresponding value range is shown in Table 28 below:

Table 28

Groups with conversion numbers greater than or equal to 1 and less than 10:

The mean of the industry factor: 0.9840837406933998.

The proportion of industry factors falling within the corresponding value range is shown in Table 29 below:

Table 29

Groups with conversions greater than or equal to 10:

The mean of the industry factor: 1.0061536337623453.

The proportion of industry factors falling within the corresponding value range is shown in Table 30 below:

Table 30

From the perspective of subgroups, the main factor causing the low GMV/COST is the overestimation of pcvr, and the deviations of other factors are relatively small.

Through the above-mentioned process of analyzing the deviation of each factor from the perspective of three granularities, it is found that the factors that most affect ECPM include PCVR estimation and industry factors. In order to make the cost of the advertisement close to the advertiser's bid, it is necessary to jointly calibrate the PCVR and the industry factor, that is, the product of the PCVR and the industry factor is used as the joint factor, and then it is calibrated by the method in the embodiment corresponding to Figure 3 above.

Through the method provided by the embodiment of this application, the factors that have a greater impact on ECPM in advertisements in a specific industry scenario can be analyzed from different granularities, and then the product of the analyzed factors can be used as a joint factor, through the above Figure 3 The method provided in the corresponding embodiment calibrates the joint factors, so as to obtain a more accurate ECPM value and improve the prediction accuracy of ECPM.

Further, please refer to FIG. 7 , which is a schematic structural diagram of a data calibration device provided by an embodiment of the present application. The above-mentioned data calibration device may be a computer program (including program code) running on a computer device, for example, the data calibration device is an application software; the device may be used to execute the corresponding steps in the method provided by the embodiment of the present application. As shown in FIG. 7 , the data calibration device may include: an acquisition module 11 , a division module 12 , an aggregation statistics module 13 , a valid data determination module 14 and a calibration module 15 .

An acquisition module 11, configured to acquire resource attribute information of target business resources under N resource attribute types; N is a positive integer;

The division module 12 is used to cluster the historical business resources in the historical business resource set based on the S combination types to obtain H historical business resource subsets; the resource attribute type corresponding to each combination type belongs to N resources Attribute type; the historical resource attribute combination of each historical business resource in a historical business resource subset is the same, and a historical resource attribute combination is associated with a resource attribute type corresponding to a combination type; S is a positive integer less than or equal to N; H is positive integer;

Aggregation statistical module 13, is used for carrying out aggregation statistical processing based on H historical business resource sub-collections respectively, obtains aggregated data set; Aggregate data set includes H historical business resource sub-collections respectively corresponding conversion number and joint factor; This conversion number is It is determined according to the respective conversion numbers of each historical business resource in the corresponding historical business resource subset, and the joint factor is determined according to the corresponding estimated conversion rate and industry factor of each historical business resource in the corresponding historical business resource subset;

Effective data determination module 14, is used for according to resource attribute information, obtains the effective transformation number and the effective joint factor for target business resource in aggregate data set;

The calibration module 15 is used to determine the calibration coefficient according to the effective conversion number and the effective joint factor, and calibrate the estimated joint factor of the target business resource according to the calibration coefficient; the estimated joint factor is based on the estimated conversion rate and industry factor of the target business resource definite.

Wherein, the specific function implementation of the acquisition module 11, the division module 12, the aggregation statistics module 13, the effective data determination module 14 and the calibration module 15 can refer to the specific description of steps S101-step S105 in the corresponding embodiment in FIG. to repeat.

Wherein, the S combination types include the combination type Mi, i is a positive integer less than or equal to S; the historical business resource set includes the historical business resource Td, and d is a positive integer less than or equal to the total number of historical business resources in the historical business resource set ;

Referring to FIG. 7 again, the dividing module 12 may include: a combination determining unit 121 and a subset determining unit 122 .

A combination determining unit 121, configured to determine the resource attribute type contained in the combination type Mi as the target resource attribute type;

The combination determining unit 121 is further configured to determine the historical resource attribute information associated with the historical business resource Td in the historical business resource set and the target resource attribute type as the historical resource attribute combination of the historical business resource Td;

The sub-set determining unit 122 is configured to add each historical business resource with the same historical resource attribute combination to the same historical business resource sub-set in the historical business resource set, and obtain one or more historical business resource sub-sets corresponding to the combination type Mi gather;

The subset determination unit 122 is further configured to use one or more historical service resource subsets corresponding to each combination type to form H historical service resource subsets.

For the specific function implementation manners of the combination determining unit 121 and the subset determining unit 122, reference may be made to the specific description of step S102 in the embodiment corresponding to FIG. 3 , which will not be repeated here.

Wherein, the H historical business resource subsets include the historical business resource subset Nj, j is a positive integer less than or equal to H;

Please refer to FIG. 7 again, the aggregation statistics module 13 may include: a first determination unit 131 , a second determination unit 132 , a consumption acquisition unit 133 , a sufficient determination unit 134 and a data set generation unit 135 .

The first determining unit 131 is configured to determine the first unit conversion number and the first unit combination factor corresponding to the historical business resource subset Nj; the first unit conversion number is based on the historical business resources in the historical business resource subset Nj in the first The number of conversions per unit time is determined, and the first unit joint factor is determined according to the estimated conversion rate and industry factor of each historical business resource in the historical business resource subset Nj within the first unit time;

The second determination unit 132 is configured to determine the second unit conversion number and the second unit combination factor corresponding to the historical business resource subset Nj; the second unit conversion number is based on the historical business resources in the historical business resource subset Nj in the second The number of conversions per unit time is determined, and the second unit joint factor is determined according to the estimated conversion rate and industry factor of each historical business resource in the historical business resource subset Nj within the second unit time; the second unit time is greater than the first a unit of time;

A consumption acquisition unit 133, configured to acquire the first unit consumption data of the historical service resource subset Nj;

The sufficient judgment unit 134 is configured to use the first unit conversion number as the conversion number corresponding to the historical business resource subset Nj if the first unit consumption data belongs to the sufficient consumption data, and use the first unit combination factor as the historical business resource subset Nj The corresponding joint factor;

The sufficient judging unit 134 is further configured to use the second unit conversion number as the conversion number corresponding to the historical business resource subset Nj if the first unit consumption data belongs to insufficient consumption data, and use the second unit joint factor as the historical business resource sub-set The joint factor corresponding to the set Nj;

The data set generation unit 135 is configured to generate an aggregated data set including the conversion numbers and the joint factors corresponding to the H historical business resource subsets respectively when the respective conversion numbers and joint factors corresponding to the H historical business resource subsets are obtained.

Wherein, the specific function implementation of the first determination unit 131, the second determination unit 132, the consumption acquisition unit 133, the sufficient determination unit 134, and the data set generation unit 135 can refer to the specifics of steps S203-step S206 in the corresponding embodiment in FIG. 5 description, and will not be repeated here.

Referring to FIG. 7 again, the first determination unit 131 may include: a first information acquisition subunit 1311 and a first data determination subunit 1312 .

The first information acquisition subunit 1311 is configured to determine each historical business resource in the historical business resource subset Nj as a historical business resource to be counted, and obtain log information of the historical business resource to be counted;

The first information obtaining subunit 1311 is further configured to obtain the conversion number, estimated conversion rate and industry factor of the historical business resource to be counted within the first unit duration from the log information;

The first data determination subunit 1312 is used to multiply the estimated conversion rate of the historical business resources to be counted within the first unit time length and the industry factor to obtain the joint factor of the historical business resources to be counted within the first unit time length; Summing the joint factors of the historical business resources to be counted within the first unit duration to obtain the first unit joint factor corresponding to the historical business resource subset Nj;

The first data determining subunit 1312 is configured to sum the conversion numbers of the historical business resources to be counted within the first unit duration to obtain the first unit conversion numbers corresponding to the historical business resource subset Nj.

Wherein, the specific function implementation manners of the first information acquiring subunit 1311 and the first data determining subunit 1312 can refer to the specific description of step S203 in the embodiment corresponding to FIG. 5 , which will not be repeated here.

Referring to FIG. 7 again, the second determination unit 132 may include: a second information acquisition subunit 1321 , a second data determination subunit 1322 , and an attenuation processing subunit 1323 .

The second information acquisition subunit 1321 is configured to determine each historical business resource in the historical business resource subset Nj as the historical business resource to be counted, and obtain the log information of the historical business resource to be counted;

The second information acquiring subunit 1321 is also used to determine the second unit duration;

The second information acquisition subunit 1321 is also used to divide the second unit duration based on the first unit duration to obtain at least two statistical periods; the duration of each statistical period is less than or equal to the first unit duration;

The second information obtaining subunit 1321 is also used to obtain the conversion number, estimated conversion rate and industry factor of each historical business resource to be counted within each statistical period from the log information;

The second data determination subunit 1322 is used for multiplying the estimated conversion rate of the historical business resource to be counted within the statistical time period and the industry factor for each statistical period to generate the historical business resource to be counted in The joint factor in the statistical period; summing the joint factors of the historical business resources to be counted in the statistical period to obtain the statistical period joint factor of the historical business resource subset Nj in the statistical period;

The second data determination subunit 1322 is also used to sum the conversion numbers of the historical business resources to be counted within the statistical period for each statistical period, and obtain the historical business resource subset Nj within the statistical period Conversions for the statistics period.

The attenuation processing subunit 1323 is used to process the statistical period conversion numbers and statistical period joint factors of the historical business resource subset Nj in each statistical period according to the time decay strategy, and obtain the first historical service resource subset Nj corresponding to Two-unit transformation number and second joint factor.

For the implementation of specific functions of the second information acquisition subunit 1321 , the second data determination subunit 1322 and the attenuation processing subunit 1323 , please refer to the specific description of step S204 in the embodiment corresponding to FIG. 5 , which will not be repeated here.

Wherein, at least two statistical periods include a statistical period Lk, k is a positive integer less than or equal to the total number of at least two statistical periods; the start time of the statistical period Lk is earlier than the statistical period Lk+1;

The attenuation processing subunit is specifically used to: according to the time attenuation factor, the total number of at least two statistical periods, and the forward sequence of the start time of the statistical period Lk in at least two statistical periods, the statistics within the statistical period Lk The number of period conversions and the joint factor of the statistical period are respectively attenuated to obtain the number of attenuation conversions and the joint factor of attenuation; the number of attenuation conversions in each statistical period is summed to obtain the second unit conversion corresponding to the historical business resource subset Nj number; the attenuation joint factors in each statistical period are summed to obtain the second unit joint factor corresponding to the historical service resource subset Nj.

Wherein, the first unit consumption data is the sum of the consumption data of each historical business resource in the historical business resource subset Nj within the first unit duration;

Referring to FIG. 7 again, the data calibration device 1 may further include: a consumption determination module 16 .

The consumption determination module 16 is used to obtain the conversion transaction value data, and determine the sufficient data threshold according to the conversion transaction value data;

The consumption determination module 16 is further configured to determine that the first unit consumption data belongs to sufficient consumption data if the first unit consumption data is greater than the sufficient data threshold;

The consumption determination module 16 is further configured to determine that the first unit consumption data belongs to insufficient consumption data if the first unit consumption data is less than or equal to the sufficient data threshold.

Wherein, the specific function implementation of the consumption determination module 16 can refer to the specific description of step S205 in the embodiment corresponding to FIG. 5 , which will not be repeated here.

Referring to FIG. 7 again, the valid data determination module 14 may include: a combination extraction unit 141 , a data search unit 142 and a valid determination unit 143 .

A combination extraction unit 141, configured to extract S resource attribute combinations from the resource attribute information based on the S combination types;

A data search unit 142, configured to search the conversion numbers and joint factors corresponding to each of the S resource attribute combinations in the aggregated data set;

The effective determination unit 143 is configured to determine the effective conversion number and the effective combination factor for the target service resource among the conversion numbers and the combination factors corresponding to the S resource attribute combinations.

For the implementation of specific functions of the combination extraction unit 141 , the data search unit 142 and the validity determination unit 143 , refer to the specific description of step S104 in the embodiment corresponding to FIG. 3 , and details are not repeated here.

Wherein, the S resource attribute combinations include the resource attribute combination Za, and a is a positive integer less than or equal to S;

Referring to FIG. 7 again, the data searching unit 142 may include: a matching determining subunit 1421 and a data obtaining subunit 1422 .

The matching determination subunit 1421 is used to search for the same historical service resource subset as the resource attribute combination Za in the historical resource attribute combinations corresponding to the H historical service resource subsets as the matching subset;

The data obtaining subunit 1422 is configured to obtain the conversion number and the combination factor corresponding to the matching subset in the aggregated data set as the conversion number and the combination factor corresponding to the resource attribute combination Za.

Wherein, the specific function implementation manners of the matching determination subunit 1421 and the data acquisition subunit 1422 can refer to the specific description of step S104 in the embodiment corresponding to FIG. 3 , and details are not repeated here.

Referring to FIG. 7 again, the validity determination unit 143 may include: a priority determination subunit 1431 , a validity determination subunit 1432 and a valid data determination subunit 1433 .

The priority determination subunit 1431 is configured to determine the priority of the conversion numbers and joint factors corresponding to each of the S resource attribute combinations according to the priorities of the S combination types;

The validity determining subunit 1432 is configured to determine the validity of the conversion numbers and joint factors corresponding to each of the S resource attribute combinations according to the consumption data corresponding to each of the S resource attribute combinations;

The effective data determination subunit 1433 is used to determine the effective conversion numbers and combination factors among the corresponding conversion numbers and combination factors of the S resource attribute combinations as candidate conversion numbers and candidate combination factors, and select the candidate conversion numbers with the highest priority The number of conversions and the candidate joint factor are used as the effective conversion number and effective joint factor for the target business resource.

For the implementation of specific functions of the priority determination subunit 1431 , the validity determination subunit 1432 and the valid data determination subunit 1433 , please refer to the specific description of step S104 in the embodiment corresponding to FIG. 3 , which will not be repeated here.

Please refer to Fig. 7 again, the data calibration device 1 may also include: a request receiving module 17 and a stage determining module 18.

A request receiving module 17, configured to receive a calibration request for a target service resource;

Stage determination module 18, used to determine the promotion stage of target business resources;

The stage determination module 18 is further configured to respond to the calibration request of the target service resource and execute the step of acquiring resource attribute information of the target service resource under N resource attribute types if the promotion stage of the target service resource is the initial promotion stage.

Wherein, the specific function implementation manners of the request receiving module 17 and the stage determining module 18 can refer to the specific description of step S105 in the embodiment corresponding to FIG. 3 , which will not be repeated here.

Referring to FIG. 7 again, the data calibration device 1 may further include: an impact factor determination module 19 .

The impact factor determination module 19 is used to, in the set of historical business resources, according to the division granularity of historical business resources, use the historical business resources whose expected consumption data is smaller than the actual consumption data as resources to be processed, and add the resources to be processed to the set of resources to be processed ; The resource set to be processed includes the resource Sr to be processed, and r is a positive integer less than or equal to the total number of resources to be processed in the resource set to be processed;

The impact factor determination module 19 is also used to adjust the price according to the resource Sr to be processed, the risk control factor, the estimated conversion rate, the actual conversion rate, the estimated cost ratio factor, the actual billing ratio factor, the estimated click rate, the actual Click rate and industry factor, determine the control value of the resource Sr to be processed, conversion rate ratio, billing ratio factor ratio, click rate ratio and industry factor;

The impact factor determination module 19 is also used to obtain the value range when the control value, conversion rate ratio, billing ratio factor ratio, click-through rate ratio and industry factor corresponding to each resource to be processed are determined;

The impact factor determination module 19 is also used to obtain resources to be processed whose conversion number is greater than or equal to the first conversion threshold and whose conversion number is less than the second conversion threshold in the set of resources to be processed, as the first conversion resource; the second conversion threshold is greater than first conversion threshold;

The impact factor determination module 19 is also used to determine the first control analysis ratio, the first conversion rate, and the first conversion rate according to the control value, conversion rate ratio, billing ratio factor ratio, click-through rate ratio, industry factor and value interval corresponding to the first conversion resource. Rate Analysis Ratio, First Billing Ratio Factor Analysis Ratio, First Click Rate Analysis Ratio, and First Industry Factor Analysis Ratio;

The impact factor determination module 19 is also used to acquire resources to be processed whose conversion number is greater than or equal to the second conversion threshold in the resource set to be processed, as the second converted resource;

The influence factor determination module 19 is also used to determine the second regulation analysis ratio, the second conversion rate analysis ratio, the second billing ratio factor analysis ratio, the second click rate analysis ratio and the second industry factor analysis ratio;

The impact factor determination module 19 is also used to analyze the ratio of the first regulation analysis, the first conversion rate analysis ratio, the first billing ratio factor analysis ratio, the first click-through rate analysis ratio, the first industry factor analysis ratio, and the second regulation analysis Ratio, the second conversion rate analysis ratio, the second billing ratio factor analysis ratio, the second click-through rate analysis ratio and the second industry factor analysis ratio are analyzed and processed to determine the impact factor used to adjust the expected income of the display; the impact factors include forecast The estimated conversion rate and industry factors are used together to generate an estimated joint factor.

For the specific function implementation of the impact factor determining module 19 , refer to the specific description of steps S301 - S308 in the embodiment corresponding to FIG. 6 , which will not be repeated here.

Further, please refer to FIG. 8 , which is a schematic structural diagram of a computer device provided by an embodiment of the present application. As shown in Figure 8, the data calibration device 1 in the above-mentioned embodiment corresponding to Figure 7 can be applied to the above-mentioned computer equipment 1000, and the above-mentioned computer equipment 1000 can include: a processor 1001, a network interface 1004 and a memory 1005, in addition, the above-mentioned computer equipment 1000 also includes: a user interface 1003 , and at least one communication bus 1002 . Wherein, the communication bus 1002 is used to realize connection and communication between these components. Wherein, the user interface 1003 may include a display screen (Display) and a keyboard (Keyboard), and the optional user interface 1003 may also include a standard wired interface and a wireless interface. Optionally, the network interface 1004 may include a standard wired interface and a wireless interface (such as a WI-FI interface). The memory 1005 can be a high-speed RAM memory, or a non-volatile memory, such as at least one disk memory. Optionally, the memory 1005 may also be at least one storage device located away from the aforementioned processor 1001 . As shown in FIG. 8 , the memory 1005 as a computer-readable storage medium may include an operating system, a network communication module, a user interface module, and a device control application program.

In the computer device 1000 shown in FIG. 8 , the network interface 1004 can provide a network communication function; the user interface 1003 is mainly used to provide an input interface for the user; and the processor 1001 can be used to call the device control application stored in the memory 1005 program to implement the data calibration method provided in the embodiment of this application.

It should be understood that the computer equipment 1000 described in the embodiment of the present application can execute the description of the data calibration method in the previous embodiments, and can also execute the description of the data calibration device 1 in the embodiment corresponding to FIG. 7 above. This will not be repeated here. In addition, the description of the beneficial effect of adopting the same method will not be repeated here.

In addition, it should be pointed out here that the embodiment of the present application also provides a computer-readable storage medium, and the above-mentioned computer-readable storage medium stores the computer program executed by the data calibration device 1 mentioned above. When the computer loads and executes the above-mentioned computer program, it can execute the description of the above-mentioned data calibration method in any of the above-mentioned embodiments, so details will not be repeated here. In addition, the description of the beneficial effect of adopting the same method will not be repeated here. For the technical details not disclosed in the embodiments of the computer-readable storage medium involved in the present application, please refer to the description of the method embodiments of the present application.

The above-mentioned computer-readable storage medium may be the data calibration device provided in any of the foregoing embodiments or an internal storage unit of the above-mentioned computer equipment, such as a hard disk or memory of the computer equipment. The computer-readable storage medium may also be an external storage device of the computer device, such as a plug-in hard disk equipped on the computer device, a smart memory card (smart media card, SMC), a secure digital (secure digital, SD) card, Flash card (flash card), etc. Further, the computer-readable storage medium may also include both an internal storage unit of the computer device and an external storage device. The computer-readable storage medium is used to store the computer program and other programs and data required by the computer device. The computer-readable storage medium can also be used to temporarily store data that has been output or will be output.

The above disclosures are only preferred embodiments of the present application, which certainly cannot limit the scope of the present application. Therefore, equivalent changes made according to the claims of the present application still fall within the scope of the present application.

Claims

A data calibration method, performed by a computer device, comprising:

Obtain the resource attribute information of the target business resource under N resource attribute types; N is a positive integer;

Based on the S combination types, each historical business resource in the historical business resource set is clustered to obtain H historical business resource subsets; the resource attribute type corresponding to each combination type belongs to the N resource attribute types; a The historical resource attribute combination of each historical business resource in the historical business resource subset is the same, and a historical resource attribute combination is associated with a resource attribute type corresponding to a combination type; S is a positive integer less than or equal to N; H is a positive integer;

Based on the H sub-sets of historical business resources, aggregate statistical processing is performed respectively to obtain an aggregated data set; the aggregated data set includes the conversion numbers and joint factors corresponding to each of the H historical business resource sub-sets; the conversion numbers are It is determined according to the respective conversion numbers of each historical business resource in the corresponding historical business resource subset, and the joint factor is determined according to the corresponding estimated conversion rate and industry factor of each historical business resource in the corresponding historical business resource subset ;

According to the resource attribute information, obtain the effective conversion number and effective combination factor for the target business resource in the aggregated data set;

Determine the calibration coefficient according to the effective conversion number and the effective joint factor, and calibrate the estimated joint factor of the target service resource according to the calibration coefficient; the estimated joint factor is based on the estimated joint factor of the target service resource Determined by estimated conversion rate and industry factors.
According to the method according to claim 1, the S combination types include combination types Mi, i is a positive integer less than or equal to S; the set of historical service resources includes historical service resources Td, and d is less than or equal to the historical service resource Td. A positive integer of the total number of historical business resources in the business resource collection;

According to the S combination types, each historical business resource in the historical business resource set is clustered respectively to obtain H historical business resource subsets, including:

Determining the resource attribute type contained in the combination type Mi as the target resource attribute type;

determining the historical resource attribute information associated with the historical business resource Td in the historical business resource set and the target resource attribute type as the historical resource attribute combination of the historical business resource Td;

In the set of historical business resources, each historical business resource with the same combination of historical resource attributes is added to the same historical business resource subset to obtain one or more historical business resource subsets corresponding to the combination type Mi;

One or more historical service resource subsets corresponding to each combination type are used to form the H historical service resource subsets.
The method according to claim 1, wherein the H historical service resource subsets include a historical service resource subset Nj, where j is a positive integer less than or equal to H;

The performing aggregated statistical processing based on the H historical service resource subsets respectively to obtain aggregated data sets, including:

Determine the first unit conversion number and the first unit joint factor corresponding to the historical business resource subset Nj; the first unit conversion number is based on the first unit duration of each historical business resource in the historical business resource subset Nj The first unit combination factor is determined according to the estimated conversion rate and industry factor of each historical business resource in the historical business resource subset Nj within the first unit duration;

Determine the second unit conversion number and the second unit joint factor corresponding to the historical business resource subset Nj; the second unit conversion number is based on the second unit duration of each historical business resource in the historical business resource subset Nj The conversion number within the second unit is determined, and the second unit combination factor is determined according to the estimated conversion rate and industry factor of each historical business resource in the historical business resource subset Nj within the second unit duration; the The second unit duration is greater than the first unit duration;

Acquiring the first unit consumption data of the historical service resource subset Nj;

If the first unit consumption data belongs to sufficient consumption data, the first unit conversion number is used as the conversion number corresponding to the historical business resource subset Nj, and the first unit combination factor is used as the historical business resource The joint factor corresponding to the subset Nj;

If the first unit consumption data belongs to insufficient consumption data, use the second unit conversion number as the conversion number corresponding to the historical business resource subset Nj, and use the second unit combination factor as the historical business The joint factor corresponding to the resource subset Nj;

When the conversion numbers and joint factors corresponding to the H historical business resource subsets are obtained, an aggregated data set including the respective conversion numbers and joint factors corresponding to the H historical business resource subsets is generated.
According to the method according to claim 3, said determining the first unit conversion number and the first unit combination factor corresponding to the historical service resource subset Nj includes:

Determining each historical business resource in the historical business resource subset Nj as the historical business resource to be counted, and obtaining the log information of the historical business resource to be counted;

Obtain the number of conversions, estimated conversion rates, and industry factors of the historical business resources to be counted within the first unit duration from the log information;

Multiply the estimated conversion rate of the historical business resources to be counted within the first unit of time by the industry factor to obtain the joint factor of the historical business resources to be counted within the first unit of time; Summing the combination factors of the historical service resources to be counted within the first unit duration to obtain the first unit combination factor corresponding to the historical service resource subset Nj;

The conversion numbers of each historical service resource to be counted within the first unit duration are summed to obtain the first unit conversion number corresponding to the historical service resource subset Nj.
The method according to claim 3, said determining the second unit conversion number and the second unit combination factor corresponding to the historical service resource subset Nj, comprising:

Determining each historical business resource in the historical business resource subset Nj as the historical business resource to be counted, and obtaining the log information of the historical business resource to be counted;

determining the second unit duration;

The second unit duration is divided based on the first unit duration to obtain at least two statistical periods; the duration of each statistical period is less than or equal to the first unit duration;

Obtain the number of conversions, estimated conversion rates, and industry factors of each of the historical business resources to be counted within each statistical period from the log information;

For each statistical period, the estimated conversion rate of the historical business resources to be counted within the statistical period and the industry factor are multiplied to generate the historical business resources to be counted within the statistical period Combining factors; summing the combining factors of each of the historical business resources to be counted in the statistical period to obtain the statistical period combining factors of the historical service resource subset Nj in the statistical period;

For each statistical period, the conversion numbers of the historical business resources to be counted within the statistical period are summed to obtain the statistical period of the historical business resource subset Nj within the statistical period number of conversions;

According to the time attenuation strategy, the conversion number of the statistical period and the joint factor of the statistical period of the historical business resource subset Nj in each statistical period are processed, and the second unit conversion number and the corresponding second unit conversion number of the historical business resource subset Nj are obtained. Second joint factor.
The method according to claim 5, the at least two statistical periods include a statistical period Lk, and k is a positive integer less than or equal to the total number of the at least two statistical periods; the start time of the statistical period Lk is earlier During the statistical period Lk+1;

According to the time attenuation strategy, the conversion number of the statistical period and the joint factor of the statistical period in each statistical period of the historical service resource subset Nj are respectively processed to obtain the second unit corresponding to the historical service resource subset Nj Conversions and second syndication factors, including:

According to the time decay factor, the total number of the at least two statistical periods, and the forward order of the start time of the statistical period Lk in the at least two statistical periods, the statistics in the statistical period Lk The period conversion number and the statistical period joint factor are respectively attenuated to obtain the attenuation conversion number and the attenuation joint factor;

Summing the attenuation conversion numbers in each statistical period to obtain the second unit conversion numbers corresponding to the historical service resource subset Nj;

The attenuation combination factors in each statistical period are summed to obtain the second unit combination factor corresponding to the historical service resource subset Nj.
According to the method according to claim 3, the first unit consumption data is the sum of the consumption data of each historical service resource in the historical service resource subset Nj within the first unit duration;

The method also includes:

Acquiring converted transaction value data, and determining a sufficient data threshold according to the converted transaction value data;

If the first unit consumption data is greater than the sufficient data threshold, determine that the first unit consumption data belongs to the sufficient consumption data;

If the first unit consumption data is less than or equal to the sufficient data threshold, it is determined that the first unit consumption data belongs to the insufficient consumption data.
According to the method according to claim 1, said acquisition of effective conversion numbers and effective joint factors for said target business resources in said aggregated data set according to said resource attribute information comprises:

Extracting S resource attribute combinations from the resource attribute information based on the S combination types;

Find the conversion numbers and joint factors corresponding to each of the S resource attribute combinations in the aggregated data set;

Among the conversion numbers and joint factors corresponding to the S resource attribute combinations, the effective conversion numbers and effective joint factors for the target service resource are determined.
According to the method according to claim 8, the S resource attribute combinations include a resource attribute combination Za, where a is a positive integer less than or equal to S;

The searching for the corresponding conversion numbers and joint factors of the S resource attribute combinations in the aggregated data set includes:

In the historical resource attribute combinations corresponding to the H historical service resource subsets, search for a historical service resource subset identical to the resource attribute combination Za as a matching subset;

The conversion numbers and combination factors corresponding to the matching sub-sets are obtained from the aggregated data set as the conversion numbers and combination factors corresponding to the resource attribute combination Za.
According to the method according to claim 9, the determination of the effective conversion number and the effective combination factor for the target service resource among the corresponding conversion numbers and combination factors of the S resource attribute combinations includes:

According to the priorities of the S combination types, determine the conversion numbers and the priorities of the joint factors corresponding to the S resource attribute combinations;

According to the consumption data corresponding to each of the S resource attribute combinations, determine the validity of the conversion numbers and the joint factors corresponding to each of the S resource attribute combinations;

Determine the valid conversion numbers and joint factors among the corresponding conversion numbers and joint factors of the S resource attribute combinations as candidate conversion numbers and candidate joint factors, and use the candidate conversion numbers and candidate joint factors with the highest priority as The number of effective conversions and the effective joint factor for the target business resource.
The method according to claim 1, said method further comprising:

receiving a calibration request for the target service resource;

Determining the promotion stage of the target business resource;

If the promotion stage of the target service resource is the initial promotion stage, the step of acquiring resource attribute information of the target service resource under N resource attribute types is performed in response to the calibration request of the target service resource.
The method according to claim 1, said method further comprising:

In the set of historical business resources, according to the division granularity of historical business resources, the historical business resources whose expected consumption data is less than the actual consumption data are used as resources to be processed, and the resources to be processed are added to the set of resources to be processed; The resource set to be processed includes a resource to be processed Sr, where r is a positive integer less than or equal to the total number of resources to be processed in the resource set to be processed;

According to the price adjustment factor, risk control factor, estimated conversion rate, actual conversion rate, estimated cost ratio factor, actual billing ratio factor, estimated click rate, actual click rate, and industry factor of the resource Sr to be processed, determine The control value, conversion rate ratio, billing ratio factor ratio, click-through rate ratio and industry factor of the resource Sr to be processed;

Get the value range;

In the set of resources to be processed, resources to be processed whose conversion number is greater than or equal to a first conversion threshold and whose conversion number is less than a second conversion threshold are acquired as first conversion resources; the second conversion threshold is greater than the first conversion threshold a conversion threshold;

According to the control value, conversion rate ratio, billing ratio factor ratio, click rate ratio, industry factor and the value range corresponding to the first conversion resource, determine the first control analysis ratio, the first conversion rate analysis ratio, and the second conversion rate analysis ratio. 1. Billing ratio factor analysis ratio, first click-through rate analysis ratio and first industry factor analysis ratio;

In the set of resources to be processed, acquire resources to be processed whose conversion numbers are greater than or equal to the second conversion threshold as second converted resources;

Determine the second control analysis ratio, the second conversion rate analysis ratio, the second conversion rate analysis ratio, and the second conversion rate analysis ratio according to the adjustment value, conversion rate ratio, billing ratio factor ratio, click-through rate ratio, industry factor, and the value range corresponding to the second conversion resource. The second billing ratio factor analysis ratio, the second click-through rate analysis ratio and the second industry factor analysis ratio;

For the first control analysis ratio, the first conversion rate analysis ratio, the first billing ratio factor analysis ratio, the first click-through rate analysis ratio, the first industry factor analysis ratio, the second control analysis ratio, and the second conversion rate analysis Ratio, the second billing ratio factor analysis ratio, the second click-through rate analysis ratio and the second industry factor analysis ratio are analyzed and processed to determine the influencing factors used to adjust the expected income of the display; the influencing factors include estimated conversion rate and industry factor, the estimated conversion rate and the industry factor are jointly used to generate the estimated joint factor.
A data calibration device, comprising:

An acquisition module, configured to acquire resource attribute information of target business resources under N resource attribute types; N is a positive integer;

The division module is used to cluster the historical business resources in the historical business resource set based on the S combination types to obtain H historical business resource subsets; the resource attribute type corresponding to each combination type belongs to the N Resource attribute type; the historical resource attribute combination of each historical business resource in a historical business resource subset is the same, and a historical resource attribute combination is associated with a resource attribute type corresponding to a combination type; S is a positive integer less than or equal to N; H is a positive integer;

An aggregation statistics module, configured to perform aggregation statistics processing based on the H historical business resource subsets respectively to obtain an aggregated data set; the aggregated data set includes the respective conversion numbers and joint factors of the H historical business resource subsets ; The conversion number is determined according to the respective conversion numbers of each historical business resource in the corresponding historical business resource subset, and the joint factor is based on the corresponding estimated conversion of each historical business resource in the corresponding historical business resource subset rate and industry factors;

an effective data determination module, configured to obtain the effective conversion number and effective combination factor for the target business resource in the aggregated data set according to the resource attribute information;

A calibration module, configured to determine a calibration coefficient according to the effective conversion number and the effective joint factor, and calibrate the estimated joint factor of the target service resource according to the calibration coefficient; the estimated joint factor is based on the Determined by the estimated conversion rate of the target business resource and industry factors.
A computer device, comprising: a processor, a memory, and a network interface;

The processor is connected to the memory and the network interface, wherein the network interface is used to provide a network communication function, the memory is used to store program codes, and the processor is used to call the program codes to execute The method according to any one of claims 1-12.
A computer-readable storage medium, wherein a computer program is stored in the computer-readable storage medium, and the computer program is adapted to be loaded by a processor and execute the method according to any one of claims 1-12.
A computer program product comprising instructions, which, when run on a computer, cause the computer to implement the method according to any one of claims 1-12.