A method for sorting exposure in a live broadcast room and related equipment

Abstract

The embodiment of the present invention provides a method and related equipment for ranking exposure of a live broadcast room, which can make the ranking of the live broadcast room more accurate and improve the enthusiasm of the anchor. The method includes: determining the state vector of the target live broadcast room at the first moment; estimating the number of false viewers of the target live broadcast room at the second moment according to the state vector; calculating the logarithm of the average viewing time of users of the target live broadcast room at the second moment; calculating the total number of people in the target live broadcast room at the second moment; iteratively calculating the first utility parameter of the live broadcast platform at the second moment; calculating the utility of the live broadcast platform at the second moment; determining the first utility parameter corresponding to the maximum utility of the live broadcast platform at the second moment as the second utility parameter; determining the second click-through conversion rate of the target live broadcast room at the second moment according to the second utility parameter and the state vector; and ranking the target live broadcast room for exposure at the second moment based on the second click-through conversion rate.

Description

Technical Field

[0001] The present invention relates to the field of information processing technology, and in particular to an exposure sorting method for a live broadcast room and related equipment. Background Art

[0002] In the live broadcast platform, in order to obtain better resource positions, live broadcast rooms often forge some fake views. These fake views will make the click-through rate of the live broadcast room very high in a short period of time. Through recommendation algorithms or manual operation rules, it is easy to obtain better resource positions at the next moment and thus obtain more resources. Over time, those live broadcast rooms that cheat obtain good resources for a longer period of time through a period of fake viewing behavior, while those live broadcast rooms that do not cheat will have lower exposure.

[0003] The above phenomenon illustrates the drawbacks of the current resource allocation in live broadcast rooms. First, it does not take into account the host’s use of false and cheating methods to influence the strategy for resource allocation. Second, it does not take into account the game process between the host and the platform. Summary of the invention

[0004] The embodiment of the present invention provides a method for ranking exposure of a live broadcast room and related equipment, which can make the ranking of the live broadcast room more accurate and improve the enthusiasm of the anchor.

[0005] A first aspect of an embodiment of the present invention provides an exposure ranking method for a live broadcast room, which is applied to a live broadcast platform. The method includes:

[0006] Determine a state vector of a target live broadcast room at a first moment, the state vector including the number of real viewers of the target live broadcast room, the number of fake viewers of the target live broadcast room, the ratio of the number of fake viewers of the target live broadcast room, the total number of people exposed to the target live broadcast room, the logarithm of the average viewing time of users in the target live broadcast room, and the first click conversion rate of the target live broadcast room;

[0007] Estimate the number of false viewers of the target live broadcast room at a second moment according to the state vector, where the second moment is a moment after the first moment, and the interval between the second moment and the first moment is a preset duration;

[0008] Calculate the logarithm of the average viewing time of users in the target live broadcast room at the second moment according to the logarithm of the average viewing time of users in the target live broadcast room;

[0009] Calculate the total number of people in the target live broadcast room at the second moment according to the number of false viewers in the target live broadcast room and the real number of viewers in the target live broadcast room;

[0010] Iteratively calculating a first utility parameter of the live broadcast platform at the second moment according to the logarithm of the average viewing time of users of the target live broadcast room at the second moment, the total number of people in the target live broadcast room at the second moment, and the state vector;

[0011] Calculate the utility of the live broadcast platform at the second moment according to the first utility parameter, the logarithm of the average viewing time of users at the second moment, the total number of people in the target live broadcast room at the second moment, and the state vector;

[0012] Determine the first utility parameter corresponding to the maximum utility of the live broadcast platform at the second moment as the second utility parameter;

[0013] Determine a second click-through conversion rate of the target live broadcast room at the second moment according to the second utility parameter and the state vector;

[0014] Based on the second click-through conversion rate, the target live broadcast room is ranked for exposure at the second moment.

[0015] Optionally, estimating the number of false viewers of the target live broadcast room at the second moment according to the state vector includes:

[0016] The number of fake viewers of the target live broadcast room at the second moment is estimated by the following formula:

[0017]

[0018] Wherein, t is the first moment, t+1 is the second moment, and f i t+1 is the number of false viewers of the target live broadcast room i at the second moment, is the state vector, Δ is the decision function, Obtained through the following formula:

[0019]

[0020] Among them, u is a k*1 vector, k is the dimension of the state vector of the target live broadcast room i, and α is a constant.

[0021] Optionally, calculating the logarithm of the average viewing time of users of the target live broadcast room at the second moment according to the logarithm of the average viewing time of users in the target live broadcast room includes:

[0022] The logarithm of the average viewing time of users in the target live broadcast room at the second moment is obtained by the following formula:

[0023]

[0024] Wherein, t is the first moment, t+1 is the second moment, is a normal distribution with a mean of τ is the standard deviation, is the logarithm of the average viewing time of users of the target live broadcast room at the second moment, is the logarithm of the average viewing time of users in the target live broadcast room at the first moment.

[0025] Optionally, the first utility parameter includes a parameter W t+1 And the parameter β t+1 , the iterative calculation of the first utility parameter of the live broadcast platform at the second moment according to the logarithm of the average viewing time of users of the target live broadcast room at the second moment, the total number of people in the target live broadcast room at the second moment, and the state vector includes:

[0026] The parameter W of the first utility parameter of the live broadcast platform at the second moment is obtained by iteration through the following formula: t +1 :

[0027]

[0028] Among them, W t+1 (k+1) is the parameter W in the first utility parameter t+1 The value at the kth iteration is, is the total number of people in the target live broadcast room i at the first moment, is the state vector, n is the number of live broadcast rooms in the live broadcast platform, τ is the standard deviation, is the logarithm of the average viewing time of users in the target live broadcast room at the first moment, and ξ is the learning rate;

[0029] The parameter β of the first utility parameter of the live broadcast platform at the second moment is obtained by iteratively obtaining the following formula: t+1 :

[0030]

[0031] Among them, β t+1 (k+1) is the parameter β in the second utility parameter t+1 The value at the kth iteration, τ is the standard deviation, is the logarithm of the average viewing time of users in the target live broadcast room at the first moment.

[0032] Optionally, calculating the utility of the live broadcast platform at the second moment according to the first utility parameter, the logarithm of the average viewing time of users at the second moment, the total number of people in the target live broadcast room at the second moment, and the state vector includes:

[0033] The utility of the live broadcast platform at the second moment is obtained by the following formula:

[0034]

[0035] Where n is the number of live broadcast rooms in the live broadcast platform, i is the target live broadcast room, t is the first moment, t+1 is the second moment, F t+1 is the utility of the live broadcast platform at the second moment, is the logarithm of the average viewing time of users of the target live broadcast room at the second moment, is the state vector, λ is a constant used to adjust the degree of negative impact, W t+1 and β t+1 is the first utility parameter.

[0036] Optionally, it is characterized in that the determining the second click-through conversion rate of the target live broadcast room at the second moment according to the second utility parameter and the state vector comprises:

[0037] The second click conversion rate is obtained by the following formula:

[0038]

[0039] Among them, ctr i t+1 is the second click conversion rate, i is the target live broadcast room, t+1 is the second time, is the state vector, as well as is the second utility parameter.

[0040] In summary, it can be seen that in the embodiment provided by the present invention, the exposure ranking device of the live broadcast room determines the state vector of the target live broadcast room at the first moment, and predicts the number of false viewers at the second moment, the logarithm of the average viewing time of users at the second moment, and the total number of people in the target live broadcast room at the second moment through the state vector of the live broadcast room, and then iteratively calculates the utility parameter of the live broadcast platform at the second moment, and calculates the utility of the live broadcast platform at the second moment according to the utility parameter at the second moment, and then takes the utility parameter corresponding to the maximum utility value of the live broadcast platform at the second moment in the iterative calculation as the optimal utility parameter to calculate the click-through conversion rate of the target live broadcast room at the second moment, and then performs exposure ranking of the target live broadcast room according to the click-through conversion rate at the second moment. It can be seen from this that the state vector of the live broadcast room at the first moment t can be used to determine the pit position ranking of the live broadcast room at the next moment t+1 (that is, the second moment), and the click-through conversion rate of the live broadcast room at the moment t+1 can be obtained through the pit position ranking, so that when sorting, the sorting of the live broadcast room is more accurate and the enthusiasm of the anchor is improved. BRIEF DESCRIPTION OF THE DRAWINGS

[0041] Figure 1 A schematic diagram of a flow chart of a method for ranking exposure in a live broadcast room provided by an embodiment of the present invention;

[0042] Figure 2 A virtual structural diagram of an exposure sorting device for a live broadcast room provided by an embodiment of the present invention;

[0043] Figure 3 A schematic diagram of the hardware structure of an exposure sorting device for a live broadcast room provided by an embodiment of the present invention;

[0044] Figure 4 A schematic diagram of an embodiment of an electronic device provided by an embodiment of the present invention;

[0045] Figure 5 A schematic diagram of an embodiment of a computer-readable storage medium provided in an embodiment of the present invention. DETAILED DESCRIPTION

[0046] The embodiment of the present invention provides a method for ranking exposure of a live broadcast room and related equipment, which can quickly recommend a live broadcast room to a user.

[0047] The following describes the exposure sorting method for a live broadcast room from the perspective of an exposure sorting device for the live broadcast room. The exposure sorting device for the live broadcast room may be a server or a service unit in the server.

[0048] See also Figure 1 , Figure 1 A schematic diagram of an embodiment of a method for ranking exposure in a live broadcast room provided by an embodiment of the present invention is provided. The method for ranking exposure in a live broadcast room is applied to a live broadcast platform. The method for ranking exposure in a live broadcast room includes:

[0049] 101. Determine a state vector of a target live broadcast room at a first moment.

[0050] In this embodiment, the exposure ranking device of the live broadcast room can determine the state vector of the target live broadcast room at the first moment, wherein the state vector includes the real number of viewers of the target live broadcast room, the number of fake viewers of the target live broadcast room, the proportion of the number of fake viewers of the target live broadcast room, the total number of people exposed to the target live broadcast room, the logarithm of the average viewing time of users in the target live broadcast room, and the first click conversion rate of the target live broadcast room. In other words, the exposure ranking device of the live broadcast room can construct the state characteristics of the target live broadcast room by means of a state vector, mainly describing the state characteristics of the live broadcast room through the following characteristic elements:

[0051] (r,f,fr,b,logw,ctrw);

[0052] Among them, ctrw = ctr*logw, ctr is the click-through conversion rate, which is calculated by dividing the number of users entering the live broadcast room by the number of users exposed to the live broadcast room;

[0053] logw is the logarithm of the average viewing time w of users in the live broadcast room;

[0054] ctrw is the click-through viewing time, which measures the final viewing time of users from exposure to entering the live broadcast room. It takes into account the click link and the viewing link, and represents the expected viewing time of users after exposure to the live broadcast room;

[0055] r refers to the actual number of viewers in the live broadcast room;

[0056] f refers to the fake number of viewers in the live broadcast room;

[0057] fr refers to the percentage of fake viewers in the live broadcast room, which is calculated by

[0058] b refers to the total number of people exposed in the live broadcast room;

[0059] Therefore, the state vector of the target live broadcast room i at the first time t can be constructed through the above state characteristics:

[0060]

[0061] in, is the state vector of the target live broadcast room i at time t, r i t is the actual number of viewers in the target live broadcast room at time t, f i t is the number of fake viewers in the target live broadcast room at time t, fr i t is the proportion of fake viewers in target live broadcast room i at time t, is the total number of exposures of the target live broadcast room i at time t, is the logarithm of the average viewing time of users in the target live broadcast room i at time t, is the first click conversion rate of target live broadcast room i at time t.

[0062] It is understandable that the reasons for selecting the above-mentioned characteristic elements as the description of the status of the live broadcast room are: the proportion of false viewers and the number of false viewers in the live broadcast room indicate the false viewing situation of the current live broadcast room; the number of real viewers, click-to-conversion duration and click-to-conversion rate of the live broadcast room indicate the click-to-conversion situation of the live broadcast room; the number of exposures indicates the resource situation of the current live broadcast room. Using the above-mentioned characteristics as the description of the status of the live broadcast room can more realistically reflect the live broadcast status of the live broadcast room. In other words, the selection of the above-mentioned characteristic elements is not a selection of human subjective factors, but a selection that conforms to natural laws and can more realistically reflect the live broadcast status of the live broadcast room.

[0063] 102. Estimate the number of false viewers of the target live broadcast room at the second moment according to the state vector.

[0064] In this embodiment, after determining the state vector of the target live broadcast room at the first moment, the exposure ranking device of the live broadcast room can estimate the number of false viewers of the target live broadcast room at the second moment according to the state vector, where the second moment is the moment after the first moment, and the interval between the second moment and the first moment is a preset duration. Specifically, the exposure ranking device of the live broadcast room can estimate the number of false viewers of the target live broadcast room at the second moment by the following formula:

[0065]

[0066] Among them, t is the first moment, t+1 is the second moment, and f i t+1 is the number of fake viewers of the target live broadcast room i at the second moment, is the state vector, Δ is the decision function, It can be obtained by the following formula:

[0067]

[0068] Among them, u is a k*1 vector, k is the dimension of the state vector of the target live broadcast room i (here k is 6, that is, the state vector of live broadcast room i is constructed from 6 dimensions), and α is a constant; in order to make the decision function as close as possible to the predicted number of false viewers, then:

[0069]

[0070] It should be noted that the principle of the calculation formula of the above decision function is: for live broadcast room i, the decision function calculates the number of false viewers of live broadcast room i at the second moment The predicted number of fake viewers is f i t+1, the difference between the predicted number of false viewers of live broadcast room i at the second moment and the calculated number of false viewers of live broadcast room i at the second moment is In order to make the decision function of all live broadcast rooms as close as possible to the predicted number of fake viewers, the sum of squares of the difference between the decision function of a single live broadcast and the predicted number of fake viewers is used to measure the degree of closeness. The smaller the sum of squares, the smaller the overall deviation between the two. The advantage of using squares is that the value obtained by the decision function may be greater than or less than the predicted number of fake viewers. Using squares can unify the two situations for measurement and improve the accuracy of the calculation.

[0071] In order to solve the above decision function, we need to obtain the parameters α and u, and set the initial values ​​of the parameters α and u to α(0) = α 0 ,u(0)=u 0 , and the parameters α and u are calculated by iteration through the following formula:

[0072]

[0073]

[0074] Among them, α(s) is the value of parameter α in the s-th iteration, u(s) is the value of parameter u in the s-th iteration, and φ is the learning rate, which is generally between 0.01 and 0.1 (of course, the specific value of φ can also be set according to actual conditions, and there is no specific limitation).

[0075] It is understandable that the live broadcast room will decide how many false views will be made in the next moment based on the current state. Simply put, if it is found that certain characteristic factors in the current state, such as logw, ctrw, etc., are low, the live broadcast room will increase the number of false viewers in the next period; the reason why it can be assumed that the decision function of the live broadcast room is a linear function is that each factor is directly related to the number of false viewers decided by the live broadcast room. For example, the lower the ctrw at time t, the higher the number of false viewers at time t+1.

[0076] It should be noted that the principle of the above iteration to calculate the iterative formula of parameters α and parameter u is: in order to make the objective function as small as possible, the parameters should be updated in the direction of the maximum gradient at each iteration, so that the minimum value of the function can be found as soon as possible. The direction of gradient descent can be obtained by taking the derivative of the objective function, so for parameter α (it can be understood that the inverse of parameter u is similar to the calculation method of parameter a, and parameter α is used as an example here), its derivative is From this we can get the derivative of parameter α at the s+1th iteration: After calculating the derivative, you need to multiply the derivative by the learning rate. The advantage of doing this is to control the speed of each learning. Learning too slowly will make it difficult to converge, and learning too fast may cause repeated oscillations and fail to find the minimum value.

[0077] 103. Calculate the logarithm of the average viewing time of users in the target live broadcast room at the second moment according to the logarithm of the average viewing time of users in the target live broadcast room.

[0078] In this embodiment, the exposure ranking device of the live broadcast room can calculate the logarithm of the average viewing time of users in the target live broadcast room at the second moment according to the logarithm of the average viewing time of users in the target live broadcast room. Specifically, the logarithm of the average viewing time of users in the target live broadcast room at the second moment can be obtained by the following formula:

[0079]

[0080] Among them, t is the first moment, t+1 is the second moment, is a normal distribution with a mean of τ is the standard deviation (the standard deviation τ is a constant greater than 0, and its value is generally between 0.1 and 0.5), is the logarithm of the average viewing time of users in the target live broadcast room at the second moment, is the logarithm of the average viewing time of users in the target live broadcast room at the first moment, It represents a normal distribution.

[0081] 104. Calculate the total number of people in the target live broadcast room at the second moment according to the number of false viewers in the target live broadcast room and the real number of viewers in the target live broadcast room.

[0082] In this embodiment, the exposure ranking device of the live broadcast room can calculate the total number of people in the target live broadcast room at the second moment according to the number of false viewers in the target live broadcast room and the number of real viewers in the target live broadcast room. Specifically, the total number of people in the target live broadcast room at the second moment can be obtained by the following formula:

[0083]

[0084] In the above formula, represents a normal distribution, is the total number of people in the target live broadcast room at the first moment (that is, For r i t With f i t The sum of the above), β is the standard deviation (β is a constant greater than 0, and the value of β is generally between 0.1 and 0.5).

[0085] It should be noted that the exposure ranking device of the live broadcast room can estimate the number of false viewers of the target live broadcast room at the second moment through step 102, calculate the logarithm of the average user viewing time of the target live broadcast room at the second moment through step 103, and calculate the total number of people in the target live broadcast room at the second moment through step 104. However, there is no restriction on the order of execution of these three steps. Step 102 can be executed first, step 103 can be executed first, step 104 can be executed first, or they can be executed simultaneously without specific limitation.

[0086] 105. Iteratively calculate the first utility parameter of the live broadcast platform at the second moment according to the logarithm of the average viewing time of users in the target live broadcast room at the second moment, the total number of people in the target live broadcast room at the second moment, and the state vector.

[0087] In this embodiment, after obtaining the logarithm of the average viewing time of users in the target live room at the second moment, the total number of people in the target live room at the second moment, and the state vector, the exposure ranking device of the live room can iteratively calculate the first utility parameter of the live platform at the second moment according to the logarithm of the average viewing time of users in the target live room at the second moment, the total number of people in the target live room at the second moment, and the state vector, wherein the first utility parameter includes parameter W t+1 And the parameter β t+1 Specifically, the parameter W of the live broadcast platform at the second moment can be calculated as follows: t+1 And the parameter β t+1 :

[0088]

[0089] Among them, W t+1 (k+1) is the parameter W in the first utility parameter t+1 The value at the kth iteration is, is the total number of people in the target live broadcast room i at the first moment, is the state vector, n is the number of live broadcast rooms in the live broadcast platform, τ is the standard deviation, is the logarithm of the average viewing time of users in the target live broadcast room at the first moment, ξ is the learning rate, and its value is generally between 0.01 and 0.1. Here, β is the standard deviation (a constant greater than 0, and its value is generally between 0.1 and 0.5);

[0090] The parameter β of the first utility parameter of the live broadcast platform at the second moment is obtained by iteratively obtaining the following formula: t+1 :

[0091]

[0092] Among them, β t+1 (k+1) is the parameter β in the second utility parameter t+1The value at the kth iteration, τ is the standard deviation, is the logarithm of the average viewing time of users in the target live broadcast room at the first moment, ξ is the learning rate, and its value is generally between 0.01-0.1.

[0093] It should be noted that when iterating through the formula to calculate β t+1 and W t+1 When , the parameter λ can also be iteratively calculated using the following formula:

[0094]

[0095] Among them, λ(k+1) parameter λ is the value taken at the kth iteration, ξ is the learning rate, and the value is generally between 0.01-0.1. Later, when performing iterative calculation of the utility of the target live broadcast room at the second moment, the value of parameter λ can also be updated each time, and the specific value is not limited.

[0096] It should be noted that the parameter W for obtaining the first utility parameter of the live broadcast platform at the second moment is t+1 and β t+1 The principle of the iterative formula is: in order to make the objective function as large as possible, the parameters should be updated in the direction of the maximum gradient at each iteration, so that the maximum value of the function can be found as soon as possible. The direction of gradient descent can be obtained by taking the derivative of the objective function, so for the parameter β, its derivative at time t is For parameter W, its derivative at time t is For the parameter β, after calculating the derivative, you need to multiply the derivative by the learning rate. The advantage of doing this is the speed of learning each time. Learning too slowly will make it difficult to converge, and learning too fast may cause repeated oscillations and fail to find the maximum value. The same is true for parameter W.

[0097] 106. Calculate the utility of the live broadcast platform at the second moment according to the first utility parameter, the logarithm of the average viewing time of users at the second moment, the total number of people in the target live broadcast room at the second moment, and the state vector.

[0098] In this embodiment, the exposure ranking device of the live broadcast room can calculate the utility of the live broadcast platform at the second moment by using the first utility parameter, the logarithm of the average viewing time of users at the second moment, the total number of people in the target live broadcast room at the second moment, and the state vector. Specifically, the utility of the live broadcast platform at the second moment can be obtained by the following formula:

[0099]

[0100] Where n is the number of live broadcast rooms in the live broadcast platform, i is the target live broadcast room, t is the first moment, t+1 is the second moment, and F t+1 For the utility of the live broadcast platform at the second moment, is the logarithm of the average viewing time of users in the target live broadcast room at the second moment, is the state vector, λ is a constant (λ is used to adjust the degree of negative impact), W t+1 and β t+1 is the first utility parameter.

[0101] That is, after each iteration in step 105, a first utility parameter (W t+1 and β t+1 ), and then substitute the first utility parameter into the formula in step 106 to obtain the utility of the target live broadcast room corresponding to the first utility parameter of this iteration at the second moment, and so on, until the iteration termination moment is reached. The number of iterations is the number of first utility parameters and the corresponding target live broadcast rooms at the second moment that can be obtained.

[0102] It should be noted that in the embodiment provided by the present invention, the total utility of the live broadcast platform at time T is defined and obtained by the following formula:

[0103]

[0104] Among them, U T is the sum of the utilities of the live broadcast platform between 1 and time t, λ is a constant used to adjust the degree of negative impact, and n is the total number of live broadcast rooms on the platform. The meaning of the above formula is: the number of users watching the real live broadcast room brings positive utility to the live broadcast platform, while the number of fake users watching the live broadcast platform brings negative utility to the live broadcast platform, so the negative utility is subtracted from the positive utility. The utility here can be expressed as the comprehensive viewing intensity by multiplying the number of viewers by the logarithm of the average viewing time. The live broadcast platform needs to make decisions continuously, take timely actions, and adjust the resource positions of the live broadcast room to maximize the total utility.

[0105] 107. Determine the first utility parameter corresponding to the time when the utility of the live broadcast platform at the second moment is the maximum as the second utility parameter.

[0106] In this embodiment, after the iteration is terminated, the exposure ranking device of the live broadcast room determines the first utility parameter corresponding to the maximum utility of the live broadcast platform at the second moment as the second utility parameter.

[0107] 108. Determine a second click-through conversion rate of the target live broadcast room at the second moment according to the second utility parameter and the state vector.

[0108] In this embodiment, for the live broadcast platform, the state vector of the live broadcast room at the first moment t can be used to determine the ranking of the slots of the live broadcast room at the next moment t+1 (that is, the second moment), and the click-through rate of the live broadcast room at moment t+1 can be obtained through the ranking of the slots. It is considered here that the click-through rate of the live broadcast room at moment t+1 is also a linear function of the state of the live broadcast room at moment t. The principle is that the click-through rate can be regarded as the user's viewing decision process, and whether or not to click is directly related to the state of the live broadcast room. For example, a large number of real viewers indicates that the live broadcast room is popular and has high live broadcast quality. Therefore, after obtaining the second utility parameter, the exposure ranking device of the live broadcast room can determine the second click-through conversion rate of the target live broadcast room at the second moment according to the second utility parameter and the state vector of the target live broadcast room at the first moment. Specifically, the second click-through conversion rate of the target live broadcast room at the second moment can be obtained by the following formula:

[0109]

[0110] Among them, ctr i t+1 is the second click conversion rate, i is the target live broadcast room, t+1 is the second moment, is the state vector, as well as is the second utility parameter.

[0111] 109. Based on the second click conversion rate, the target live broadcast rooms are ranked for exposure at the second moment.

[0112] In this embodiment, after obtaining the second click-through conversion rate, the exposure ranking device of the live broadcast room can perform exposure ranking of the target live broadcast room at the second moment based on the second click-through conversion rate, that is, according to the ctr i t+1 At the second moment, the target live broadcast room is sorted. The ctr i t+1 The higher it is, the more forward the exposure position of the target live broadcast room will be.

[0113] In summary, it can be seen that in the embodiment provided by the present invention, the exposure ranking device of the live broadcast room determines the state vector of the target live broadcast room at the first moment, and predicts the number of false viewers at the second moment, the logarithm of the average viewing time of users at the second moment, and the total number of people in the target live broadcast room at the second moment through the state vector of the live broadcast room, and then iteratively calculates the utility parameter of the live broadcast platform at the second moment, and calculates the utility of the live broadcast platform at the second moment according to the utility parameter at the second moment, and then takes the utility parameter corresponding to the maximum utility value of the live broadcast platform at the second moment in the iterative calculation as the optimal utility parameter to calculate the click-through conversion rate of the target live broadcast room at the second moment, and then performs exposure ranking of the target live broadcast room according to the click-through conversion rate at the second moment. It can be seen from this that the state vector of the live broadcast room at the first moment t can be used to determine the pit position ranking of the live broadcast room at the next moment t+1 (that is, the second moment), and the click-through conversion rate of the live broadcast room at the moment t+1 can be obtained through the pit position ranking, so that when sorting, the sorting of the live broadcast room is more accurate and the enthusiasm of the anchor is improved.

[0114] The following is a detailed explanation with examples:

[0115] For the live broadcast platform, live broadcast room i is any live broadcast room in the live broadcast platform. At time t, the state vector of live broadcast room i at time t can be obtained. Then the iterative process of the decision function is explained:

[0116] Assuming α(0)=1, u(0)=(1,1,5,0.01,1,1), For example, the learning rate φ = 0.05 Iterate to calculate the values ​​of α and u as follows:

[0117] First iteration:

[0118]

[0119] In the next iteration, the result of the above iteration is used for calculation. When the iteration termination condition is reached (the iteration termination condition may be, for example, parameter convergence or the number of iterations reaching the maximum value), the values ​​of α and u are finally calculated. After that, the values ​​of α and u can be used to estimate the number of false viewers of live broadcast room i at the second moment.

[0120] Since the click-through conversion rate of live broadcast room i at the second moment needs to be determined by the following formula to obtain the optimal W t+1 and β t+1 , so that F t+1 Max. at F t+1 When the number of live broadcast rooms is at its maximum, the live broadcast room can achieve the optimal ranking, and the exposure of the pit position is most reasonable at this time:

[0121]

[0122] Combined with the above F t+1 The iterative formula and the following formula are iteratively calculated, where n is the total number of live broadcast rooms in the live broadcast platform, is the logarithm of the average viewing time of users in live broadcast room i at the second moment, is the state vector of live broadcast room i at time t, The total number of people in live broadcast room i at time t+1 has been calculated, so it can be Calculate the F after each iteration t+1 (That is, after each iteration, a parameter W, a parameter β, and a parameter λ are calculated, the calculated parameters W, β, and λ are substituted into F t+1 Calculate until the iteration termination condition is reached, then select F t+1 The parameters W, β and λ when the maximum value is the most optimal solution), that is, the optimal parameters W, β and λ can be obtained, so that F t+1 maximum:

[0123]

[0124]

[0125]

[0126] After the optimal solutions of the parameters W, β and λ are obtained, the parameters W, β and λ can be

[0127] The parameters W and β of the live broadcast platform at time t+1 and the state vector of the live broadcast room i at time t can be Substituting the following formula, we can calculate the click-through conversion rate of live broadcast platform i at time t+1:

[0128]

[0129] Therefore, the live broadcast room i can be ranked based on the click-through conversion rate of the live broadcast room i at time t+1 obtained by the calculation. The larger the click-through conversion rate calculated, the higher the ranking, and the higher the ranking, the higher the exposure pit position obtained by the live broadcast room. In summary, it can be seen that in the embodiment provided by the present invention, the pit position ranking of the live broadcast room i at the next time t+1 can be determined by the state vector of the live broadcast room i at time t, and the click-through conversion rate of the live broadcast room i at time t+1 can be obtained by pit position ranking, so that when ranking, the ranking of the live broadcast room is more accurate, which improves the enthusiasm of the anchor.

[0130] The above describes the exposure sorting method for the live broadcast room in the embodiment of the present invention. The following describes the exposure sorting device for the live broadcast room in the embodiment of the present invention.

[0131] See also Figure 2 , a virtual structural diagram of an exposure ranking device for a live broadcast room in an embodiment of the present invention, the exposure ranking device for a live broadcast room is applied to a live broadcast platform, and the exposure ranking device for a live broadcast room includes:

[0132] A first determining unit 201 is used to determine a state vector of a target live broadcast room at a first moment, wherein the state vector includes the real number of viewers of the target live broadcast room, the fake number of viewers of the target live broadcast room, the proportion of the fake number of viewers of the target live broadcast room, the total number of viewers exposed to the target live broadcast room, the logarithm of the average viewing time of users in the target live broadcast room, and the first click-through conversion rate of the target live broadcast room;

[0133] An estimating unit 202, configured to estimate the number of false viewers of the target live broadcast room at a second moment according to the state vector, where the second moment is a moment after the first moment, and an interval between the second moment and the first moment is a preset duration;

[0134] The first calculation unit 203 is used to calculate the logarithm of the average viewing time of users in the target live broadcast room at the second moment according to the logarithm of the average viewing time of users in the target live broadcast room;

[0135] A second calculation unit 204 is used to calculate the total number of people in the target live broadcast room at the second moment according to the number of false viewers in the target live broadcast room and the real number of viewers in the target live broadcast room;

[0136] The third calculation unit 205 is used to iteratively calculate the first utility parameter of the live broadcast platform at the second moment according to the logarithm of the average viewing time of users of the target live broadcast room at the second moment, the total number of people in the target live broadcast room at the second moment, and the state vector;

[0137] A fourth calculation unit 206, configured to calculate the utility of the live broadcast platform at the second moment according to the first utility parameter, the logarithm of the average viewing time of users at the second moment, the total number of people in the target live broadcast room at the second moment, and the state vector;

[0138] A second determining unit 207, configured to determine the first utility parameter corresponding to the maximum utility of the live broadcast platform at the second moment as a second utility parameter;

[0139] A third determining unit 208 is configured to determine a second click-through conversion rate of the target live broadcast room at the second moment according to the second utility parameter and the state vector;

[0140] The sorting unit 209 is used to sort the exposure of the target live broadcast room at the second moment based on the second click-through conversion rate.

[0141] Optionally, the estimating unit 202 is specifically configured to:

[0142] The number of fake viewers of the target live broadcast room at the second moment is estimated by the following formula:

[0143]

[0144] Wherein, t is the first moment, t+1 is the second moment, and f i t+1 is the number of false viewers of the target live broadcast room i at the second moment, is the state vector, Δ is the decision function, Obtained through the following formula:

[0145]

[0146] Among them, u is a k*1 vector, k is the dimension of the state vector of the target live broadcast room i, and α is a constant.

[0147] Optionally, the first calculating unit 203 is specifically configured to:

[0148] The logarithm of the average viewing time of users in the target live broadcast room at the second moment is obtained by the following formula:

[0149]

[0150] Wherein, t is the first moment, t+1 is the second moment, is a normal distribution with a mean of τ is the standard deviation, is the logarithm of the average viewing time of users of the target live broadcast room at the second moment, is the logarithm of the average viewing time of users in the target live broadcast room at the first moment.

[0151] Optionally, the first utility parameter includes a parameter W t+1 And the parameter β t+1 , the third calculation unit 205 is specifically used for:

[0152] The parameter W of the first utility parameter of the live broadcast platform at the second moment is obtained by iteration through the following formula: t +1 :

[0153]

[0154] Among them, W t+1 (k+1) is the parameter W in the first utility parameter t+1 The value at the kth iteration is, is the total number of people in the target live broadcast room i at the first moment, is the state vector, n is the number of live broadcast rooms in the live broadcast platform, τ is the standard deviation, is the logarithm of the average viewing time of users in the target live broadcast room at the first moment, and ξ is the learning rate;

[0155] The parameter β of the first utility parameter of the live broadcast platform at the second moment is obtained by iteratively obtaining the following formula: t+1 :

[0156]

[0157] Among them, β t+1 (k+1) is the parameter β in the second utility parameter t+1 The value at the kth iteration, τ is the standard deviation, is the logarithm of the average viewing time of users in the target live broadcast room at the first moment.

[0158] Optionally, the fourth calculation unit 206 is specifically configured to:

[0159] The utility of the live broadcast platform at the second moment is obtained by the following formula:

[0160]

[0161] Where n is the number of live broadcast rooms in the live broadcast platform, i is the target live broadcast room, t is the first moment, t+1 is the second moment, F t+1 is the utility of the live broadcast platform at the second moment, is the logarithm of the average viewing time of users of the target live broadcast room at the second moment, is the state vector, λ is a constant used to adjust the degree of negative impact, W t+1 and β t+1 is the first utility parameter.

[0162] Optionally, the third determining unit 208 is specifically configured to:

[0163] The second click conversion rate is obtained by the following formula:

[0164]

[0165] Among them, ctr i t+1is the second click conversion rate, i is the target live broadcast room, t+1 is the second time, is the state vector, as well as is the second utility parameter.

[0166] above Figure 2 The exposure sorting device for a live broadcast room in an embodiment of the present invention is described from the perspective of modular functional entities. The exposure sorting device for a live broadcast room in an embodiment of the present invention is described in detail from the perspective of hardware processing. Figure 3 An embodiment of the exposure sorting device 300 of the live broadcast room in the embodiment of the present invention includes:

[0167] An input device 301, an output device 302, a processor 303 and a memory 304 (the number of the processor 303 can be one or more, Figure 3 In some embodiments of the present invention, the input device 301, the output device 502, the processor 303 and the memory 304 may be connected via a bus or other means, wherein: Figure 3 The example of connecting through bus is taken in the following.

[0168] Wherein, by calling the operation instruction stored in the memory 304, the processor 303 is used to perform the following steps:

[0169] Determine a state vector of a target live broadcast room at a first moment, the state vector including the number of real viewers of the target live broadcast room, the number of fake viewers of the target live broadcast room, the ratio of the number of fake viewers of the target live broadcast room, the total number of people exposed to the target live broadcast room, the logarithm of the average viewing time of users in the target live broadcast room, and the first click conversion rate of the target live broadcast room;

[0170] Estimate the number of false viewers of the target live broadcast room at a second moment according to the state vector, where the second moment is a moment after the first moment, and the interval between the second moment and the first moment is a preset duration;

[0171] Calculate the logarithm of the average viewing time of users in the target live broadcast room at the second moment according to the logarithm of the average viewing time of users in the target live broadcast room;

[0172] Calculate the total number of people in the target live broadcast room at the second moment according to the number of false viewers in the target live broadcast room and the real number of viewers in the target live broadcast room;

[0173] Iteratively calculating a first utility parameter of the live broadcast platform at the second moment according to the logarithm of the average viewing time of users of the target live broadcast room at the second moment, the total number of people in the target live broadcast room at the second moment, and the state vector;

[0174] Calculate the utility of the live broadcast platform at the second moment according to the first utility parameter, the logarithm of the average viewing time of users at the second moment, the total number of people in the target live broadcast room at the second moment, and the state vector;

[0175] Determine the first utility parameter corresponding to the maximum utility of the live broadcast platform at the second moment as the second utility parameter;

[0176] Determine a second click-through conversion rate of the target live broadcast room at the second moment according to the second utility parameter and the state vector;

[0177] Based on the second click-through conversion rate, the target live broadcast room is ranked for exposure at the second moment.

[0178] By calling the operation instructions stored in the memory 304, the processor 303 is also used to execute Figure 1 Any method in the corresponding embodiment.

[0179] See also Figure 4 , Figure 4 A schematic diagram of an electronic device according to an embodiment of the present invention.

[0180] like Figure 4 As shown, an embodiment of the present invention provides an electronic device, including a memory 410, a processor 420, and a computer program 411 stored in the memory 420 and executable on the processor 420. When the processor 420 executes the computer program 411, the following steps are implemented:

[0181] Determine a state vector of a target live broadcast room at a first moment, the state vector including the number of real viewers of the target live broadcast room, the number of fake viewers of the target live broadcast room, the ratio of the number of fake viewers of the target live broadcast room, the total number of people exposed to the target live broadcast room, the logarithm of the average viewing time of users in the target live broadcast room, and the first click conversion rate of the target live broadcast room;

[0182] Estimate the number of false viewers of the target live broadcast room at a second moment according to the state vector, where the second moment is a moment after the first moment, and the interval between the second moment and the first moment is a preset duration;

[0183] Calculate the logarithm of the average viewing time of users in the target live broadcast room at the second moment according to the logarithm of the average viewing time of users in the target live broadcast room;

[0184] Calculate the total number of people in the target live broadcast room at the second moment according to the number of false viewers in the target live broadcast room and the real number of viewers in the target live broadcast room;

[0185] Iteratively calculating a first utility parameter of the live broadcast platform at the second moment according to the logarithm of the average viewing time of users of the target live broadcast room at the second moment, the total number of people in the target live broadcast room at the second moment, and the state vector;

[0186] Calculate the utility of the live broadcast platform at the second moment according to the first utility parameter, the logarithm of the average viewing time of users at the second moment, the total number of people in the target live broadcast room at the second moment, and the state vector;

[0187] Determine the first utility parameter corresponding to the maximum utility of the live broadcast platform at the second moment as the second utility parameter;

[0188] Determine a second click-through conversion rate of the target live broadcast room at the second moment according to the second utility parameter and the state vector;

[0189] Based on the second click-through conversion rate, the target live broadcast room is ranked for exposure at the second moment.

[0190] In the specific implementation process, when the processor 420 executes the computer program 411, it can achieve Figure 1 Any implementation manner in the corresponding embodiments.

[0191] Since the electronic device introduced in this embodiment is a device used to implement an exposure sorting device for a live broadcast room in an embodiment of the present invention, based on the method introduced in the embodiment of the present invention, technical personnel in this field can understand the specific implementation mode of the electronic device of this embodiment and its various variations. Therefore, how the electronic device implements the method in the embodiment of the present invention will not be described in detail here. As long as the equipment used by technical personnel in this field to implement the method in the embodiment of the present invention belongs to the scope of protection of the present invention.

[0192] See also Figure 5 , Figure 5 A schematic diagram of an embodiment of a computer-readable storage medium provided in an embodiment of the present invention.

[0193] like Figure 5 As shown, this embodiment provides a computer-readable storage medium 500, on which a computer program 511 is stored. When the computer program 511 is executed by a processor, the following steps are implemented:

[0194] Determine a state vector of a target live broadcast room at a first moment, the state vector including the number of real viewers of the target live broadcast room, the number of fake viewers of the target live broadcast room, the ratio of the number of fake viewers of the target live broadcast room, the total number of people exposed to the target live broadcast room, the logarithm of the average viewing time of users in the target live broadcast room, and the first click conversion rate of the target live broadcast room;

[0195] Estimate the number of false viewers of the target live broadcast room at a second moment according to the state vector, where the second moment is a moment after the first moment, and the interval between the second moment and the first moment is a preset duration;

[0196] Calculate the logarithm of the average viewing time of users in the target live broadcast room at the second moment according to the logarithm of the average viewing time of users in the target live broadcast room;

[0197] Calculate the total number of people in the target live broadcast room at the second moment according to the number of false viewers in the target live broadcast room and the real number of viewers in the target live broadcast room;

[0198] Iteratively calculating a first utility parameter of the live broadcast platform at the second moment according to the logarithm of the average viewing time of users of the target live broadcast room at the second moment, the total number of people in the target live broadcast room at the second moment, and the state vector;

[0199] Calculate the utility of the live broadcast platform at the second moment according to the first utility parameter, the logarithm of the average viewing time of users at the second moment, the total number of people in the target live broadcast room at the second moment, and the state vector;

[0200] Determine the first utility parameter corresponding to the maximum utility of the live broadcast platform at the second moment as the second utility parameter;

[0201] Determine a second click-through conversion rate of the target live broadcast room at the second moment according to the second utility parameter and the state vector;

[0202] Based on the second click-through conversion rate, the target live broadcast room is ranked for exposure at the second moment.

[0203] In the specific implementation process, when the computer program 511 is executed by the processor, it can achieve Figure 1 Any implementation manner in the corresponding embodiments.

[0204] It should be noted that in the above embodiments, the description of each embodiment has its own emphasis, and for parts that are not described in detail in a certain embodiment, reference can be made to the relevant descriptions of other embodiments.

[0205] Those skilled in the art will appreciate that embodiments of the present invention may be provided as methods, systems, or computer program products. Therefore, the present invention may take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware. Moreover, the present invention may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0206] The present invention is described with reference to flowcharts and / or block diagrams of methods, devices (systems), and computer program products according to embodiments of the present invention. It should be understood that each process and / or block in the flowchart and / or block diagram, as well as the combination of processes and / or blocks in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded computer, or other programmable data processing device to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0207] These computer program instructions may also be stored in a computer readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer readable memory produce an article of manufacture including an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0208] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.

[0209] The embodiment of the present invention also provides a computer program product, which includes computer software instructions. When the computer software instructions are executed on a processing device, the processing device executes the following Figure 1 The process in the method for designing a digital platform for a wind farm in the corresponding embodiment.

[0210] The computer program product includes one or more computer instructions. When the computer program instructions are loaded and executed on a computer, the process or function described in the embodiment of the present invention is generated in whole or in part. The computer may be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions may be stored in a computer-readable storage medium, or transmitted from one computer-readable storage medium to another computer-readable storage medium. For example, the computer instructions may be transmitted from one website site, computer, server or data center to another website site, computer, server or data center by wired (e.g., coaxial cable, optical fiber, digital subscriber line (DSL)) or wireless (e.g., infrared, wireless, microwave, etc.) means. The computer-readable storage medium may be any available medium that a computer can store or a data storage device such as a server or data center that includes one or more available media integrated. The available medium may be a magnetic medium (e.g., a floppy disk, a hard disk, a tape), an optical medium (e.g., a DVD), or a semiconductor medium (e.g., a solid state disk (SSD)), etc.

Claims

1. A method for ranking exposure in live broadcast rooms, applied to a live broadcast platform, characterized in that: include: Determine a state vector of a target live broadcast room at a first moment, the state vector including the number of real viewers of the target live broadcast room, the number of fake viewers of the target live broadcast room, the ratio of the number of fake viewers of the target live broadcast room, the total number of people exposed to the target live broadcast room, the logarithm of the average viewing time of users in the target live broadcast room, and the first click conversion rate of the target live broadcast room; Estimate the number of false viewers of the target live broadcast room at a second moment according to the state vector, where the second moment is a moment after the first moment, and the interval between the second moment and the first moment is a preset duration; Calculate the logarithm of the average viewing time of users in the target live broadcast room at the second moment according to the logarithm of the average viewing time of users in the target live broadcast room; Calculate the total number of people in the target live broadcast room at the second moment according to the number of false viewers in the target live broadcast room and the real number of viewers in the target live broadcast room; Iteratively calculating a first utility parameter of the live broadcast platform at the second moment according to the logarithm of the average viewing time of users of the target live broadcast room at the second moment, the total number of people in the target live broadcast room at the second moment, and the state vector; Calculate the utility of the live broadcast platform at the second moment according to the first utility parameter, the logarithm of the average viewing time of users at the second moment, the total number of people in the target live broadcast room at the second moment, and the state vector; Determine the first utility parameter corresponding to the maximum utility of the live broadcast platform at the second moment as the second utility parameter; Determine a second click-through conversion rate of the target live broadcast room at the second moment according to the second utility parameter and the state vector; Based on the second click-through conversion rate, the target live broadcast room is ranked for exposure at the second moment.

2. The method according to claim 1, characterized in that The estimating the number of false viewers of the target live broadcast room at the second moment according to the state vector comprises: The number of fake viewers of the target live broadcast room at the second moment is estimated by the following formula: Wherein, t is the first moment, t+1 is the second moment, and f i t+1 is the number of false viewers of the target live broadcast room i at the second moment, is the state vector, △ is the decision function, Obtained through the following formula: Among them, u is a k*1 vector, k is the dimension of the state vector of the target live broadcast room i, and α is a constant.

3. The method according to claim 1, characterized in that Calculating the logarithm of the average viewing time of users in the target live broadcast room at the second moment according to the logarithm of the average viewing time of users in the target live broadcast room comprises: The logarithm of the average viewing time of users in the target live broadcast room at the second moment is obtained by the following formula: Wherein, t is the first moment, t+1 is the second moment, is a normal distribution with a mean of τ is the standard deviation, is the logarithm of the average viewing time of users of the target live broadcast room at the second moment, is the logarithm of the average viewing time of users in the target live broadcast room at the first moment.

4. The method according to claim 1, characterized in that The first utility parameter includes parameter W t+1 And the parameter β t+1 , the iterative calculation of the first utility parameter of the live broadcast platform at the second moment according to the logarithm of the average viewing time of users of the target live broadcast room at the second moment, the total number of people in the target live broadcast room at the second moment, and the state vector includes: The parameter W of the first utility parameter of the live broadcast platform at the second moment is obtained by iteration through the following formula: t+1 : Among them, W t+1 (k+1) is the parameter W in the first utility parameter t+1 The value at the kth iteration is, is the total number of people in the target live broadcast room i at the first moment, is the state vector, n is the number of live broadcast rooms in the live broadcast platform, τ is the standard deviation, is the logarithm of the average viewing time of users in the target live broadcast room at the first moment, and ξ is the learning rate; The parameter β of the first utility parameter of the live broadcast platform at the second moment is obtained by iteratively obtaining the following formula: t+1 : Among them, β t+1 (k+1) is the parameter β in the second utility parameter t+1 The value at the kth iteration, τ is the standard deviation, is the logarithm of the average viewing time of users in the target live broadcast room at the first moment.

5. The method according to claim 4, characterized in that The calculating the utility of the live broadcast platform at the second moment according to the first utility parameter, the logarithm of the average viewing time of users at the second moment, the total number of people in the target live broadcast room at the second moment, and the state vector comprises: The utility of the live broadcast platform at the second moment is obtained by the following formula: Where n is the number of live broadcast rooms in the live broadcast platform, i is the target live broadcast room, t is the first moment, t+1 is the second moment, F t+1 is the utility of the live broadcast platform at the second moment, is the logarithm of the average viewing time of users of the target live broadcast room at the second moment, is the state vector, λ is a constant used to adjust the degree of negative impact, W t+1 and β t+1 is the first utility parameter.

6. The method according to any one of claims 1 to 5, characterized in that Determining the second click conversion rate of the target live broadcast room at the second moment according to the second utility parameter and the state vector includes: The second click conversion rate is obtained by the following formula: Among them, ctr i t+1 is the second click conversion rate, i is the target live broadcast room, t+1 is the second time, is the state vector, as well as is the second utility parameter.

7. A device for ranking exposure in a live broadcast room, applied to a live broadcast platform, characterized in that: include: A first determining unit is used to determine a state vector of a target live broadcast room at a first moment, wherein the state vector includes the real number of viewers of the target live broadcast room, the fake number of viewers of the target live broadcast room, the proportion of the fake number of viewers of the target live broadcast room, the total number of people exposed to the target live broadcast room, the logarithm of the average viewing time of users in the target live broadcast room, and the first click-through conversion rate of the target live broadcast room; an estimating unit, configured to estimate the number of false viewers of the target live broadcast room at a second moment according to the state vector, wherein the second moment is a moment after the first moment, and an interval between the second moment and the first moment is a preset duration; A first calculation unit, configured to calculate the logarithm of the average viewing time of users in the target live broadcast room at the second moment according to the logarithm of the average viewing time of users in the target live broadcast room; A second calculation unit is used to calculate the total number of people in the target live broadcast room at the second moment according to the false number of viewers in the target live broadcast room and the real number of viewers in the target live broadcast room; A third calculation unit is used to iteratively calculate a first utility parameter of the live broadcast platform at the second moment according to the logarithm of the average viewing time of users of the target live broadcast room at the second moment, the total number of people in the target live broadcast room at the second moment, and the state vector; a fourth calculation unit, configured to calculate the utility of the live broadcast platform at the second moment according to the first utility parameter, the logarithm of the average viewing time of users at the second moment, the total number of people in the target live broadcast room at the second moment, and the state vector; A second determining unit, configured to determine the first utility parameter corresponding to the maximum utility of the live broadcast platform at the second moment as a second utility parameter; A third determining unit, configured to determine a second click-through conversion rate of the target live broadcast room at the second moment according to the second utility parameter and the state vector; A sorting unit is used to sort the exposure of the target live broadcast room at the second moment based on the second click-through conversion rate.

8. The device according to claim 7, characterized in that The first utility parameter includes parameter W t+1 And the parameter β t+1 , the third computing unit is specifically used for: The parameter W of the first utility parameter of the live broadcast platform at the second moment is obtained by iteration through the following formula: t+1 : Among them, W t+1 (k+1) is the parameter W in the first utility parameter t+1 The value at the kth iteration is, is the total number of people in the target live broadcast room i at the first moment, is the state vector, n is the number of live broadcast rooms in the live broadcast platform, τ is the standard deviation, is the logarithm of the average viewing time of users in the target live broadcast room at the first moment, and ξ is the learning rate; The parameter β of the first utility parameter of the live broadcast platform at the second moment is obtained by iteratively obtaining the following formula: t+1 : Among them, β t+1 (k+1) is the parameter β in the second utility parameter t+1 The value at the kth iteration, τ is the standard deviation, is the logarithm of the average viewing time of users in the target live broadcast room at the first moment.

9. An electronic device, comprising a memory and a processor, characterized in that: The processor is used to implement the steps of the exposure sorting method for the live broadcast room as described in any one of claims 1 to 6 when executing the computer management program stored in the memory.

10. A computer-readable storage medium having a computer management program stored thereon, characterized in that: When the computer management program is executed by the processor, the steps of the exposure sorting method for the live broadcast room as described in any one of claims 1 to 6 are implemented.

Images