Target user determination method, device, electronic device and storage medium

By determining the indicators of interest and setting thresholds on the live broadcast platform based on user characteristics and scientifically determining the target users, the problem of artificial experience dependence in the existing technology is solved, and the accuracy of user rewards and the maximum benefit of the live broadcast platform is improved.

CN114862432BActive Publication Date: 2024-11-26武汉斗鱼鱼乐网络科技有限公司
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202110157881.8
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2021-02-04
Publication Date
2024-11-26
Estimated Expiration
2041-02-04

AI Technical Summary

Technical Problem

When determining target users, existing live broadcast platforms rely on manual experience, lack scientificity, and are difficult to effectively manage and improve the accuracy of user rewards, making it difficult to maximize profits.

Method used

By determining the indicators of interest based on the user characteristics of candidate users on the live broadcast platform, and setting thresholds, the target users are scientifically determined, and the accuracy of user rewards and maximum benefits are improved.

Benefits of technology

The scientific determination of target users has been achieved, the accuracy of user rewards and the maximum benefit of live broadcast platforms has been improved.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN114862432B_ABST
    Figure CN114862432B_ABST
Patent Text Reader

Abstract

The embodiment of the present invention discloses a method, device, electronic device and storage medium for determining a target user, the method comprising: determining an interest index corresponding to the candidate user according to the user characteristics of the candidate user on the live broadcast platform; determining a reference degree of the candidate user as a target user according to the interest index; determining whether the candidate user is a target user according to the reference degree and a set threshold. Through the technical solution of the embodiment of the present invention, the purpose of scientifically determining the target user is achieved, the determination accuracy of the target user is improved, and it is conducive to maximizing the revenue of the live broadcast platform.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] Embodiments of the present invention relate to the field of computer technology, and in particular, to a method, device, electronic device and storage medium for determining a target user. Background Art

[0002] In order to extend the viewing time of users, live broadcast platforms usually give users some valuable props as rewards after they reach a certain viewing time, thereby encouraging users to watch longer and pay, thereby realizing the revenue conversion of the live broadcast platform. However, props themselves have costs, and live broadcast platforms need to consider how to maximize their revenue, so they need to balance the costs and benefits.

[0003] The traditional way to solve the above problems is to select the right people based on tags based on operational experience, and then reward these selected users.

[0004] However, the above method requires maintaining a large number of selection rules, which is difficult to manage, and the division of the population to be rewarded is mainly based on manual experience and is not scientific. Summary of the invention

[0005] The embodiments of the present invention provide a target user determination method, device, electronic device and storage medium, which achieve the purpose of scientifically determining the target user, improve the determination accuracy of the target user, and thus help maximize the revenue of the live broadcast platform.

[0006] In a first aspect, an embodiment of the present invention provides a method for determining a target user, the method comprising:

[0007] Determine the interest index corresponding to the candidate user according to the user characteristics of the candidate user on the live broadcast platform;

[0008] Determining a reference degree of the candidate user as a target user according to the interest indicator;

[0009] Whether the candidate user is a target user is determined according to the reference degree and a set threshold.

[0010] In a second aspect, an embodiment of the present invention further provides a target user determination device, the device comprising:

[0011] A first determination module is used to determine the interest index corresponding to the candidate user according to the user characteristics of the candidate user on the live broadcast platform;

[0012] A second determination module, configured to determine a reference degree of the candidate user as a target user according to the indicator of interest;

[0013] The third determination module is used to determine whether the candidate user is a target user according to the reference degree and a set threshold.

[0014] In a third aspect, an embodiment of the present invention further provides an electronic device, the electronic device comprising:

[0015] one or more processors;

[0016] A memory for storing one or more programs;

[0017] When the one or more programs are executed by the one or more processors, the one or more processors implement the target user determination method steps provided by any embodiment of the present invention.

[0018] In a fourth aspect, an embodiment of the present invention further provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps of a target user determination method as provided in any embodiment of the present invention.

[0019] The embodiments of the above invention have the following advantages or beneficial effects:

[0020] The target user determination method provided by the embodiment of the present invention includes: determining the interest index corresponding to the candidate user according to the user characteristics of the candidate user on the live broadcast platform; determining the reference degree of the candidate user as the target user according to the interest index; and determining whether the candidate user is the target user according to the reference degree and a set threshold. By combining the user characteristics of the candidate user on the live broadcast platform to determine whether the candidate user is the target user, the purpose of scientifically determining the target user is achieved, the determination accuracy of the target user is improved, and it is conducive to maximizing the revenue of the live broadcast platform. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] Figure 1 is a flow chart of a method for determining a target user provided by Embodiment 1 of the present invention;

[0022] Figure 2 is a flow chart of a method for determining a target user provided by Embodiment 2 of the present invention;

[0023] Figure 3 is a structural diagram of a target user determination device provided by Embodiment 3 of the present invention;

[0024] Figure 4 It is a structural schematic diagram of an electronic device provided in Embodiment 4 of the present invention. DETAILED DESCRIPTION

[0025] The present invention will be further described in detail below in conjunction with the accompanying drawings and embodiments. It is to be understood that the specific embodiments described herein are only used to explain the present invention, rather than to limit the present invention. It should also be noted that, for ease of description, only parts related to the present invention, rather than all structures, are shown in the accompanying drawings.

[0026] Embodiment 1

[0027] The technical solution of this embodiment is applicable to determining which users of the live broadcast platform should be awarded rewards, so as to encourage such users to pay to watch more live videos on the live broadcast platform, thereby maximizing the revenue of the live broadcast platform.

[0028] Specifically, the goal of the live broadcast platform is to maximize the probability of users paying through incentives, so the optimization problem can be defined and solved:

[0029]

[0030] Among them, E represents the expected loss of the live broadcast platform through the incentive measures, U represents the complete set of live broadcast platform users; i represents any platform user, represents the user characteristics of user i; Indicates whether to issue a reward, 0 means no reward, 1 means a reward is issued, Represents the expected loss that user i brings to the live streaming platform.

[0031] The constraint condition of the above optimal problem is that the expected loss E obtained by the live broadcast platform through the incentive measures cannot be greater than 0, otherwise the income of the live broadcast platform will be damaged, so E≤0. According to the above optimization problem, it is necessary to determine which users to issue rewards to and which users not to issue rewards to, that is, it is necessary to solve the problem of which users The value is 1, for which users The value is 0. For users who issue rewards, the greater the expectation of bringing benefits to the live broadcast platform, this can ensure that the live broadcast platform will eventually make a profit.

[0032] Based on the above optimization problem, this embodiment provides a method for determining a target user, where the target user may be a user to whom a reward is issued, that is, The value is 1 for the user.

[0033] Figure 1 A flowchart of a target user determination method provided in Embodiment 1 of the present invention. The method can be executed by a target user determination device, which can be implemented by software and / or hardware and is usually integrated into a server of a live broadcast platform.

[0034] like Figure 1 As shown, the method specifically comprises the following steps:

[0035] Step 110: Determine the interest index corresponding to the candidate user according to the user characteristics of the candidate user on the live broadcast platform.

[0036] Among them, the candidate user can be any user who has logged into the live broadcast platform.

[0037] The user characteristics include at least one of the following: effective viewing time ratio, active days, viewing payment conversion rate and recharge amount. Among them, the effective viewing time ratio is specifically effective viewing time / total viewing time, and the effective viewing time can refer to the time when the continuous viewing time reaches the time threshold. Usually, if the continuous viewing time reaches 5 minutes, then these 5 minutes are considered to be effective viewing time. Active days specifically refer to the number of days the user has logged in to the live broadcast platform. For example, if the user has logged in to the live broadcast platform continuously or intermittently for 30 days, then the user can be considered to have 30 days of active days. The viewing payment conversion rate can refer to the payment amount / total viewing time.

[0038] Among them, the effective viewing time ratio indicates the proportion of users who effectively watch live videos. If this ratio is not high, it means that the effective viewing time can be increased through certain rewards. The number of active days indicates the user's activity and the user's stickiness to the live broadcast platform. The higher the activity, the greater the probability of further payment. The viewing payment conversion rate indicates the user's ability to convert viewing behavior into payment behavior. The higher the user's viewing payment conversion rate, the stronger the user's ability to convert to payment behavior. The recharge amount indicates the user's ability to pay. The larger the recharge amount, the greater the probability of subsequent payment.

[0039] The indicators of interest include the probability of paying on the live broadcast platform after receiving the reward, the expected cost when the reward is issued and the user pays, the first expected benefit brought to the live broadcast platform after receiving the reward, and the second expected benefit brought to the live broadcast platform when no reward is received. Among them, the expected cost when the reward is issued and the user pays can refer to the cost consumed by issuing the reward to the user. The first expected benefit brought to the live broadcast platform after receiving the reward can refer to the fee paid by the user on the live broadcast platform after receiving the reward minus the cost consumed by the live broadcast platform for issuing the reward to the user.

[0040] By combining the user characteristics of candidate users to determine which users will be rewarded, the scientificity and accuracy of target user identification are improved, which is conducive to maximizing the revenue of the live broadcast platform.

[0041] Step 120: Determine the reference degree of the candidate user as a target user according to the interest indicator.

[0042] It is more complicated to solve the above optimization problem directly, but since the contribution of each user is very small compared to the entire user group, each user can be considered independently. If a user has a higher probability of paying, the expected income he brings to the live broadcast platform is as large as possible (the loss is as small as possible), then it is more beneficial for the live broadcast platform to issue rewards to this user. So we can set the reference degree for:

[0043]

[0044] in It represents the reference degree of issuing rewards to users with user characteristics x. represents the payment probability of a user with user feature x, represents the expected loss brought to the live broadcast platform by a user with user feature x. Obviously, for user i, The bigger, The smaller, The larger the value is, the greater the contribution to the positive benefits of the live broadcast platform. For user i, if the calculated reference degree If it is below a certain threshold, then a reward will be given, otherwise no reward will be given.

[0045] Step 130: Determine whether the candidate user is a target user according to the reference degree and a set threshold.

[0046] Optionally, if the reference degree of a candidate user is lower than a set threshold, the candidate user is determined as a target user, that is, a reward is issued to the candidate user.

[0047] The technical solution of this embodiment determines which users are to be rewarded by combining the user characteristics of candidate users, thereby improving the scientificity and accuracy of target user determination and facilitating the live broadcast platform to maximize revenue.

[0048] Embodiment 2

[0049] Figure 2 This is a flow chart of a method for determining a target user provided in Embodiment 2 of the present invention. Based on the above embodiment, this embodiment further optimizes the implementation method for determining the reference degree, aiming to improve the determination accuracy of the reference degree, thereby improving the determination accuracy of the target user. The explanations of the terms that are the same or corresponding to the above embodiment are not repeated here.

[0050] See also Figure 2 The target user determination method provided in this embodiment specifically includes the following steps:

[0051] Step 210: input the user characteristics into a classification model to obtain the probability of paying on the live broadcast platform after obtaining the reward; wherein the classification model is trained based on the user characteristics of users who pay on the live broadcast platform after the reward and the user characteristics of users who do not pay on the live broadcast platform after the reward.

[0052] Step 220: Determine the reference degree of the candidate user as a target user according to the interest indicator.

[0053] Step 230: Determine the candidate users whose reference degree is lower than the set threshold as target users.

[0054] Specifically, define the following two parameters:

[0055] C(x) and ; Where: C(x) represents the expected cost when rewards are issued and users pay; It indicates the expected benefits of user payment whether rewards are issued or not, where i=0 means no rewards are issued, and i=1 means rewards are issued. Specifically, It indicates the first expected benefit brought to the live broadcast platform after receiving the reward. It indicates the second expected benefit brought to the live broadcast platform when no reward is received.

[0056] Therefore, setting the reference It can be expressed as:

[0057]

[0058] in, Indicates the probability of paying on the live broadcast platform after receiving the reward, Indicates the probability of paying on the live streaming platform without receiving any reward.

[0059] Define s(x) to represent the probability of paying on the live broadcast platform after receiving the reward, then:

[0060]

[0061] then:

[0062] Among them, s(x) represents the probability of paying on the live broadcast platform after receiving the reward. It indicates the first expected benefit brought to the live broadcast platform after receiving the reward. It represents the second expected benefit brought to the live broadcast platform when there is no reward, and C(x) represents the expected cost when rewards are issued and users pay.

[0063] The above-mentioned indicators of interest (s(x), R1(x), R0(x), C(x)) can be estimated by sampling experimental data. A portion of users are randomly sampled for online experiments, and rewards are given when their viewing time meets certain conditions. The above-mentioned indicators of interest are calculated according to the following method.

[0064] (1) Select positive and negative samples. Positive samples are users who pay after being rewarded, and negative samples are users who do not pay after being rewarded. Train the classification model CM(x). The output of the classification model CM(x) is used as the probability s(x) of paying on the live broadcast platform after receiving the reward.

[0065] (2) Select users who pay after being rewarded, use the user features x and the cost c of rewarding the user as samples to train the regression model RM1(x), and use the output of the regression model as the expected cost C(x) when the reward is issued and the user pays.

[0066] (3) Select users who make payments after receiving the reward, use the user features x and the payment amount r of the user as samples to train the regression model RM2(x), and use the output of the regression model as the first expected benefit R1(x) brought to the live broadcast platform after receiving the reward.

[0067] (4) Select users who have not received any rewards but have paid, and use the user features x and the user's payment amount r as samples to train the regression model RM3(x). The output of the regression model is used as the second expected benefit R0(x) brought to the live broadcast platform when no rewards are received.

[0068] Exemplarily, determining the interested index corresponding to the candidate user according to the user characteristics of the candidate user on the live broadcast platform includes:

[0069] Input the user characteristics into a classification model to obtain the probability of paying on the live broadcast platform after receiving the reward; wherein the classification model is trained based on the user characteristics of users who pay on the live broadcast platform after receiving the reward and the user characteristics of users who do not pay on the live broadcast platform after receiving the reward;

[0070] Input the user characteristics into a first regression model to obtain the expected cost when the reward is issued and the user pays; wherein the first regression model is trained based on the user characteristics of the paying user on the live broadcast platform after the reward and the cost of rewarding the user;

[0071] Input the user characteristics into a second regression model to obtain a first expected benefit brought to the live broadcast platform after the reward is obtained; wherein the second regression model is trained based on the user characteristics of the paying user on the live broadcast platform after the reward and the payment amount of the user;

[0072] The user characteristics are input into a third regression model to obtain a second expected benefit brought to the live broadcast platform when no reward is obtained; wherein the third regression model is trained based on the user characteristics of users who have not obtained rewards but have paid on the live broadcast platform and the payment amount of the users.

[0073] Furthermore, since the above-mentioned indicators of interest are obtained based on the trained model, there are still some deviations. In order to further improve the calculation accuracy of the reference degree, the step of calibrating the above-mentioned indicators of interest can be added. Specifically, the probability s(x) of paying on the live broadcast platform after receiving the reward is taken as an example to illustrate the calibration method.

[0074] Exemplarily, the probability of paying on the live broadcast platform after the reward is calibrated based on the following algorithm:

[0075]

[0076] Where N(x) represents the set of N samples in the model training samples whose similarity with the user feature x reaches the set threshold, s is the user feature corresponding to any sample in the sample set, represents the estimated probability of paying on the live broadcast platform after receiving the reward through the above classification model according to user feature s, It represents the true value of the probability that the user will pay on the live broadcast platform after being rewarded with the corresponding feature s. The meaning is: If If established, The value of is -1, otherwise The value of is 1.

[0077] The idea of ​​the above calibration algorithm is to find the sample x' that is closest to sample x, and estimate the prediction deviation of sample x based on sample x'. The prediction deviation can be estimated by the classification model. and actual value Since the error can be positive or negative, the square root of the sum of squares of the average error is used. Measure. Decide whether to increase or decrease the adjustment value based on the original value (i.e. the value to be calibrated s(x), the value before calibration). If the average predicted value is higher than the average actual value, it means that the prediction result is too high, and the adjustment value needs to be reduced based on the original value to lower the predicted result; otherwise, it means that the prediction result is too low, and the adjustment value needs to be increased based on the original value. Finally, add the original value and the above calculation result to get the calibration value after calibration.

[0078] According to the above method, the expected cost C(x) when rewards are issued and users pay, the first expected benefit R1(x) brought to the live broadcast platform after receiving the reward, and the second expected benefit R0(x) brought to the live broadcast platform when no reward is received are calibrated to obtain the calibrated , and .

[0079] Furthermore, the determining, according to the interest indicator, the reference degree of the candidate user as the target user includes:

[0080] The reference degree is determined based on the following algorithm:

[0081]

[0082] Wherein, T represents the reference degree, The calibration value representing the probability of paying on the live broadcast platform after the reward, represents the calibration value of the expected cost when rewards are issued and users pay, It represents the calibration value of the first expected benefit brought to the live broadcast platform after receiving the reward. Represents the calibration value of the second expected benefit brought to the live broadcast platform when no reward is received. The reference degree is determined by the probability of payment on the live broadcast platform after the reward, the expected cost when the reward is issued and the user pays, the first expected benefit brought to the live broadcast platform after the reward is received, and the second expected benefit brought to the live broadcast platform when no reward is received. The problem can be transformed into the solution of the probability, the expected cost, the first expected benefit and the second expected benefit. The purpose and benefit of such a transformation is that these variables are related to Y=1, that is, the group of people who issued the reward. Therefore, it is possible to track the subsequent behavior of the group of people who issued the reward, thereby constructing different positive and negative samples, so as to achieve the purpose of solving the above variables using a regression model or a classification model and then determining the reference degree.

[0083] If the above reference degree is greater than the set threshold, the current candidate user will not be rewarded, that is, the current candidate user is not the target user to be rewarded; if the above reference degree is less than the set threshold, the current candidate user can be rewarded, that is, the current candidate user is the target user to be rewarded. Among them, the factors affecting the value of the set threshold may include: the budget of the live broadcast platform for the reward activity. If the budget is high, it means that more rewards can be issued. At this time, the set threshold can be appropriately increased to make more candidate users become the target users to be rewarded by the method; otherwise, the set threshold can be appropriately lowered.

[0084] The following is a detailed description of the method for determining the reference degree and the method for determining the target users:

[0085] Assume that the calibration values ​​of the above indicators of interest corresponding to the candidate user with user feature x are:

[0086]

[0087] then:

[0088] Assume that the threshold is set to -0.3. Since -0.5 is less than -0.3, the current candidate user is the target user, and it is recommended to issue a reward to the current candidate user.

[0089] The technical solution of this embodiment, based on the above embodiment, further optimizes the implementation method of determining the reference degree. Specifically, the prediction deviation is determined by the model prediction value of other user features x' closest to the user feature x. If the average prediction value is higher than the average actual value, it means that the prediction result is too high, and it is necessary to reduce the adjustment value on the basis of the original value to lower the predicted prediction result; otherwise, it means that the prediction result is too low, and it is necessary to increase the adjustment value on the basis of the original value. Finally, the original value and the above calculation result are added to obtain the calibration value after calibration, and then the reference degree is determined based on the calibration value, thereby improving the determination accuracy of the reference degree, and then improving the determination accuracy of the target user; by combining user characteristics and specific algorithms, the scientific nature of determining the target user is improved.

[0090] Embodiment 3

[0091] Figure 3 This is a schematic diagram of the structure of a target user determination device provided in Embodiment 3 of the present invention. The device specifically includes: a first determination module 310 , a second determination module 320 and a third determination module 330 .

[0092] Among them, the first determination module 310 is used to determine the interest index corresponding to the candidate user according to the user characteristics of the candidate user on the live broadcast platform; the second determination module 320 is used to determine the reference degree of the candidate user as the target user according to the interest index; the third determination module 330 is used to determine whether the candidate user is the target user according to the reference degree and a set threshold.

[0093] Based on the above technical solution, the user characteristics include at least one of the following: effective viewing time ratio, active days, viewing payment conversion rate and recharge amount.

[0094] Based on the above technical solution, the indicators of interest include the probability of paying on the live broadcast platform after receiving the reward, the expected cost when the reward is issued and the user pays, the first expected benefit brought to the live broadcast platform after receiving the reward, and the second expected benefit brought to the live broadcast platform when no reward is received.

[0095] Based on the above technical solution, the first determination module 310 includes:

[0096] The first determination unit is used to input the user characteristics into a classification model to obtain the probability of paying on the live broadcast platform after obtaining the reward; wherein the classification model is trained based on the user characteristics of users who pay on the live broadcast platform after the reward and the user characteristics of users who do not pay on the live broadcast platform after the reward.

[0097] On the basis of the above technical solution, the device further comprises: a calibration module, which is used to calibrate the probability of payment on the live broadcast platform after the reward based on the following algorithm:

[0098]

[0099] in, Indicates the calibration value to be treated The calibration value obtained after calibration, N (x) represents the set of N samples in the model training samples whose similarity with the user feature x reaches the set threshold, s is the user feature corresponding to any sample in the sample set, represents the estimated probability of paying on the live broadcast platform after receiving the reward through the above classification model according to user feature s, It represents the true value of the probability that the user pays on the live broadcast platform after receiving the reward corresponding to the user feature s.

[0100] On the basis of the above technical solution, the first determining module 310 further includes:

[0101] A second determination unit is used to input the user characteristics into a first regression model to obtain the expected cost when the reward is issued and the user pays; wherein the first regression model is trained based on the user characteristics of the paying user on the live broadcast platform after the reward and the cost of rewarding the user;

[0102] A third determination unit is used to input the user characteristics into a second regression model to obtain a first expected benefit brought to the live broadcast platform after the reward is obtained; wherein the second regression model is trained based on the user characteristics of the paying user on the live broadcast platform after the reward and the payment amount of the user;

[0103] The fourth determination unit is used to input the user characteristics into the third regression model to obtain the second expected benefit brought to the live broadcast platform when no reward is obtained; wherein the third regression model is trained based on the user characteristics of users who have not obtained rewards but have paid on the live broadcast platform and the payment amount of the users.

[0104] Based on the above technical solution, the second determination module 320 is specifically used for:

[0105] The reference degree is determined based on the following algorithm:

[0106]

[0107] Wherein, T represents the reference degree, The calibration value representing the probability of paying on the live broadcast platform after the reward, represents the calibration value of the expected cost when rewards are issued and users pay, It represents the calibration value of the first expected benefit brought to the live broadcast platform after receiving the reward. It represents the calibration value of the second expected benefit brought to the live broadcast platform when no reward is received.

[0108] The technical solution of this embodiment determines which users are to be rewarded by combining the user characteristics of candidate users, thereby improving the scientificity and accuracy of target user determination and facilitating the live broadcast platform to maximize revenue.

[0109] The target user determination device provided in the embodiment of the present invention can execute the target user determination method provided in any embodiment of the present invention, and has the corresponding functional modules and beneficial effects of executing the target user determination method.

[0110] Embodiment 4

[0111] Figure 4 A schematic diagram of the structure of an electronic device provided in Embodiment 4 of the present invention. Figure 4 A block diagram of an exemplary electronic device 12 suitable for use in implementing embodiments of the present invention is shown. Figure 4 The electronic device 12 shown is only an example and should not bring any limitation to the functions and scope of use of the embodiments of the present invention.

[0112] like Figure 4 As shown, the electronic device 12 is in the form of a general purpose computing electronic device. The components of the electronic device 12 may include, but are not limited to: one or more processors or processing units 16, a system memory 28, and a bus 18 connecting various system components (including the system memory 28 and the processing unit 16).

[0113] Bus 18 represents one or more of several types of bus structures, including a memory bus or memory controller, a peripheral bus, an accelerated graphics port, a processor or a local bus using any of a variety of bus architectures. By way of example, these architectures include, but are not limited to, an Industry Standard Architecture (ISA) bus, a Micro Channel Architecture (MAC) bus, an Enhanced ISA bus, a Video Electronics Standards Association (VESA) local bus, and a Peripheral Component Interconnect (PCI) bus.

[0114] The electronic device 12 typically includes a variety of computer system readable media. These media can be any available media that can be accessed by the electronic device 12, including volatile and non-volatile media, removable and non-removable media.

[0115] The system memory 28 may include computer system readable media in the form of volatile memory, such as random access memory (RAM) 30 and / or cache memory 32. The electronic device 12 may further include other removable / non-removable, volatile / non-volatile computer system storage media. By way of example only, the storage system 34 may be used to read and write non-removable, non-volatile magnetic media ( Figure 4 not shown, usually called a "hard drive"). Although Figure 4 Not shown in the figure, a disk drive for reading and writing to a removable non-volatile disk (e.g., a "floppy disk"), and an optical disk drive for reading and writing to a removable non-volatile optical disk (e.g., a CD-ROM, a DVD-ROM, or other optical media) may be provided. In these cases, each drive may be connected to the bus 18 via one or more data medium interfaces. The system memory 28 may include at least one program product having a set (e.g., at least one) of program modules that are configured to perform the functions of the various embodiments of the present invention.

[0116] A program / utility 40 having a set (at least one) of program modules 42 may be stored, for example, in system memory 28, such program modules 42 including, but not limited to, an operating system, one or more application programs, other program modules, and program data, each of which or some combination may include an implementation of a network environment. Program modules 42 generally perform the functions and / or methods of the embodiments described herein.

[0117] The electronic device 12 may also communicate with one or more external devices 14 (e.g., keyboards, pointing devices, displays 24, etc.), may communicate with one or more devices that enable a user to interact with the electronic device 12, and / or may communicate with any device that enables the electronic device 12 to communicate with one or more other computing devices (e.g., network cards, modems, etc.). This communication may be performed via an input / output (I / O) interface 22. In addition, the electronic device 12 may also communicate with one or more networks (e.g., local area networks (LANs), wide area networks (WANs), and / or public networks, such as the Internet) via a network adapter 20. As shown, the network adapter 20 communicates with other modules of the electronic device 12 via a bus 18. It should be understood that, although not shown in the figure, other hardware and / or software modules may be used in conjunction with the electronic device 12, including but not limited to: microcode, device drivers, redundant processing units, external disk drive arrays, RAID systems, tape drives, and data backup storage systems, etc.

[0118] The processing unit 16 executes various functional applications and determines the region of interest by running the program stored in the system memory 28, for example, implementing a target user determination method provided by the embodiment of the present invention, the method comprising:

[0119] Determine the interest index corresponding to the candidate user according to the user characteristics of the candidate user on the live broadcast platform;

[0120] Determining a reference degree of the candidate user as a target user according to the interest indicator;

[0121] Whether the candidate user is a target user is determined according to the reference degree and a set threshold.

[0122] Of course, those skilled in the art can understand that the processor can also implement the technical solution of the target user determination method provided by any embodiment of the present invention.

[0123] Embodiment 5

[0124] This embodiment 5 provides a computer-readable storage medium, on which a computer program is stored. When the program is executed by a processor, the target user determination method steps provided in any embodiment of the present invention are implemented. The method includes:

[0125] Determine the interest index corresponding to the candidate user according to the user characteristics of the candidate user on the live broadcast platform;

[0126] Determining a reference degree of the candidate user as a target user according to the interest indicator;

[0127] Whether the candidate user is a target user is determined according to the reference degree and a set threshold.

[0128] The computer storage medium of the embodiment of the present invention can adopt any combination of one or more computer-readable media. The computer-readable medium can be a computer-readable signal medium or a computer-readable storage medium. The computer-readable storage medium can be, for example, but not limited to: an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, device or device, or any combination of the above. More specific examples (non-exhaustive list) of computer-readable storage media include: an electrical connection with one or more wires, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the above. In this document, a computer-readable storage medium can be any tangible medium containing or storing a program, which can be used by an instruction execution system, device or device or used in combination with it.

[0129] Computer-readable signal media may include data signals propagated in baseband or as part of a carrier wave, which carry computer-readable program code. Such propagated data signals may take a variety of forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination of the above. Computer-readable signal media may also be any computer-readable medium other than a computer-readable storage medium, which may send, propagate, or transmit a program for use by or in conjunction with an instruction execution system, apparatus, or device.

[0130] The program code embodied on the computer readable medium may be transmitted using any appropriate medium, including but not limited to wireless, wireline, optical fiber cable, RF, etc., or any suitable combination of the foregoing.

[0131] Computer program code for performing the operation of the present invention may be written in one or more programming languages ​​or a combination thereof, including object-oriented programming languages ​​such as Java, Smalltalk, C++, and conventional procedural programming languages ​​such as "C" or similar programming languages. The program code may be executed entirely on the user's computer, partially on the user's computer, as a separate software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In the case of a remote computer, the remote computer may be connected to the user's computer through any type of network, including a local area network (LAN) or a wide area network (WAN), or may be connected to an external computer (e.g., via the Internet using an Internet service provider).

[0132] It should be understood by those skilled in the art that the modules or steps of the present invention described above can be implemented by a general-purpose computing device, they can be concentrated on a single computing device, or distributed on a network composed of multiple computing devices, optionally, they can be implemented by a program code executable by a computer device, so that they can be stored in a storage device and executed by the computing device, or they can be made into individual integrated circuit modules, or multiple modules or steps therein can be made into a single integrated circuit module for implementation. Thus, the present invention is not limited to any specific combination of hardware and software.

[0133] Note that the above are only preferred embodiments of the present invention and the technical principles used. Those skilled in the art will understand that the present invention is not limited to the specific embodiments described herein, and that various obvious changes, readjustments and substitutions can be made by those skilled in the art without departing from the scope of protection of the present invention. Therefore, although the present invention has been described in more detail through the above embodiments, the present invention is not limited to the above embodiments, and may include more other equivalent embodiments without departing from the concept of the present invention, and the scope of the present invention is determined by the scope of the appended claims.

Claims

1. A method for determining a target user, characterized in that: include: Determine the interest index corresponding to the candidate user according to the user characteristics of the candidate user on the live broadcast platform; Determining a reference degree of the candidate user as a target user according to the interest indicator; Determining whether the candidate user is a target user according to the reference degree and a set threshold; The indicators of interest include the probability of paying on the live broadcast platform after receiving the reward, the expected cost when the reward is issued and the user pays, the first expected benefit brought to the live broadcast platform after receiving the reward, and the second expected benefit brought to the live broadcast platform when no reward is received; The determining of the interest index corresponding to the candidate user according to the user characteristics of the candidate user on the live broadcast platform includes: Input the user characteristics into a classification model to obtain the probability of paying on the live broadcast platform after receiving the reward; wherein the classification model is trained based on the user characteristics of users who pay on the live broadcast platform after receiving the reward and the user characteristics of users who do not pay on the live broadcast platform after receiving the reward; The probability of paying on the live broadcast platform after the reward is calibrated based on the following algorithm: Among them, s ′ (x) represents the calibration value obtained after calibrating the calibration value s(x), N(x) represents the N sample set in the model training sample whose similarity with the user feature x reaches the set threshold, s is the user feature corresponding to any sample in the sample set, CM(s) represents the estimated probability of paying on the live broadcast platform after the reward obtained by the above classification model according to the user feature s, r(s) represents the true value of the probability of paying on the live broadcast platform after the reward corresponding to the user feature s. It should be noted that The meaning is: If If established, The value of is -1, otherwise The value of is 1; The determining, according to the interest indicator, the reference degree of the candidate user as the target user includes: The reference degree is determined based on the following algorithm: Wherein, T represents the reference degree, s′(x) represents the calibration value of the probability of payment on the live broadcast platform after the reward, C′(c) represents the calibration value of the expected cost when the reward is issued and the user pays, R′1(x) represents the calibration value of the first expected benefit brought to the live broadcast platform after the reward, and R0′(x) represents the calibration value of the second expected benefit brought to the live broadcast platform when no reward is received, wherein C′(x), R′1(x) and R0′(x) are obtained based on the calculation method of s′(x).

2. The method according to claim 1, characterized in that: The user characteristics include at least one of the following: effective viewing time ratio, active days, viewing payment conversion rate and recharge amount.

3. The method according to claim 1, characterized in that: The determining of the interest index corresponding to the candidate user according to the user characteristics of the candidate user on the live broadcast platform includes: The user characteristics are input into a first regression model to obtain the expected cost when the reward is issued and the user pays; wherein the first regression model is trained based on the user characteristics of the paying user on the live broadcast platform after the reward and the cost of rewarding the user.

4. The method according to claim 1, characterized in that The determining of the interest index corresponding to the candidate user according to the user characteristics of the candidate user on the live broadcast platform includes: The user characteristics are input into a second regression model to obtain a first expected benefit brought to the live broadcast platform after the reward; wherein the second regression model is trained based on the user characteristics of the paying users on the live broadcast platform after the reward and the payment amount of the user.

5. The method according to claim 1, characterized in that The determining of the interest index corresponding to the candidate user according to the user characteristics of the candidate user on the live broadcast platform includes: The user characteristics are input into a third regression model to obtain a second expected benefit brought to the live broadcast platform when no reward is obtained; wherein the third regression model is trained based on the user characteristics of users who have not obtained rewards but have paid on the live broadcast platform and the payment amount of the users.

6. A target user determination device, characterized in that: include: A first determination module is used to determine the interest index corresponding to the candidate user according to the user characteristics of the candidate user on the live broadcast platform; A second determination module, configured to determine a reference degree of the candidate user as a target user according to the indicator of interest; A third determination module, configured to determine whether the candidate user is a target user according to the reference degree and a set threshold; The indicators of interest include the probability of paying on the live broadcast platform after receiving the reward, the expected cost when the reward is issued and the user pays, the first expected benefit brought to the live broadcast platform after receiving the reward, and the second expected benefit brought to the live broadcast platform when no reward is received; The first determining module includes: A first determination unit is used to input the user characteristics into a classification model to obtain the probability of paying on the live broadcast platform after obtaining the reward; wherein the classification model is trained based on the user characteristics of users who pay on the live broadcast platform after the reward and the user characteristics of users who do not pay on the live broadcast platform after the reward; The calibration module is used to calibrate the probability of payment on the live broadcast platform after the reward based on the following algorithm: Among them, s ′ 9x) represents the calibration value obtained after calibrating the calibration value s9x), N(x) represents the N sample set in the model training sample whose similarity with the user feature x reaches the set threshold, s is the user feature corresponding to any sample in the sample set, CM(s) represents the estimated probability of paying on the live broadcast platform after the reward obtained by the above classification model according to the user feature s, r(s) represents the true value of the probability of paying on the live broadcast platform after the reward corresponding to the user feature s. It should be noted that The meaning is: If If established, The value of is -1, otherwise The value of is 1; The determining, according to the interest indicator, the reference degree of the candidate user as the target user includes: The second determination module is specifically configured to determine the reference degree based on the following algorithm: Wherein, T represents the reference degree, s ′ (x) represents the calibration value of the probability of paying on the live broadcast platform after the reward, C ′ (x) represents the calibration value of the expected cost when rewards are issued and users pay, R ′ 1(x) represents the calibration value of the first expected benefit brought to the live broadcast platform after receiving the reward, and R0′(x) represents the calibration value of the second expected benefit brought to the live broadcast platform when no reward is received. ′ (x), R ′ 1(x) and R0′(x), based on s ′ (x) is obtained by the calculation method.

7. An electronic device, characterized in that: The electronic device comprises: one or more processors; A memory for storing one or more programs; When the one or more programs are executed by the one or more processors, the one or more processors implement the target user determination method as described in any one of claims 1 to 5.