Flow control method, device, equipment, storage medium and program product

Abstract

The present application relates to a flow control method, device, equipment, storage medium and program product. The method includes: obtaining target products and target flows of each of the target products; obtaining the current flow of the target products; adjusting the current flow according to the target flow so that the current flow of the target products is adapted to the target flow, and the adjustment process includes at least one of proportional adjustment, integral adjustment and / or derivative adjustment. By using this method, the flow can be smoothly controlled.

Description

Technical Field

[0001] The present application relates to the field of computer technologies, and particularly to a traffic control method, apparatus, device, storage medium, and program product. Background Art

[0002] With the development of computer technologies, product recommendation technologies have emerged. Programs will recommend products to customers corresponding to those products according to the product recommendation technologies, and corresponding traffic will be generated during the process of recommending products to users.

[0003] Most of the existing product recommendation technologies are applicable to e-commerce scenarios. When applying such product recommendation technologies to live users, due to online factors, smooth control of traffic cannot be achieved. Summary of the Invention

[0004] Based on this, in view of the above technical problems, it is necessary to provide a traffic control method, apparatus, computer device, computer-readable storage medium, and computer program product that can smoothly control traffic.

[0005] In a first aspect, the present application provides a traffic control method, and the method includes:

[0006] Obtain target products and the target traffic of each of the target products;

[0007] Obtain the current traffic of the target products;

[0008] Adjust the current traffic according to the target traffic so that the current traffic of the target products is adapted to the target traffic, and the adjustment process includes at least one of proportional adjustment, integral adjustment, and / or derivative adjustment.

[0009] In one embodiment, it further includes: obtaining adjustment parameters obtained by performing adjustment processing on the target products, where the adjustment parameters include at least one of a proportional adjustment parameter, an integral adjustment parameter, and a derivative adjustment parameter; calculating a first score according to the adjustment parameters; calculating a second score for the target products according to user preferences; calculating a score of the target products according to the first score and the second score of the target products; sorting the target products according to the score, and adjusting the traffic of the target products according to the sorting result.

[0010] In one embodiment, the obtaining of the target products includes: obtaining products to be processed; selecting target products from the products to be processed.

[0011] In one embodiment, obtaining the target traffic of the target product includes: determining an initial target traffic based on the historical traffic of the target product; obtaining the group to which the target product belongs; determining the position of the target product within the group; and adjusting the initial target traffic according to the determined position to obtain the target traffic.

[0012] In one embodiment, obtaining the target traffic of the target product includes: obtaining the configured traffic in the configuration data of the target product as the target traffic.

[0013] In one embodiment, after obtaining the current traffic of the target product, it further includes: when the algorithm type is the target algorithm, then continue to adjust the current traffic according to the target traffic so that the current traffic of the target product is adapted to the target traffic; when the algorithm type is the global algorithm, then adjust the current traffic according to the target traffic through the global algorithm.

[0014] In one embodiment, adjusting the current traffic according to the target traffic through the global algorithm includes: obtaining the target product and the target traffic, current traffic, and previous regulation parameter of each target product; updating the regulation parameter according to the target traffic, current traffic, and the previous regulation parameter; obtaining the sorting score corresponding to the target product; and adjusting the current traffic according to the updated regulation parameter and the sorting score of the target product so that the current traffic of the target product is adapted to the target traffic.

[0015] In a second aspect, the present application provides a method for evaluating the effect of traffic control, and the method includes:

[0016] Grouping the target products to obtain an experimental group and a control group;

[0017] Obtaining the standard observation index and standard core index of the target product in the experimental group within the first period, and the standard observation index and standard core index of the target product in the control group;

[0018] Obtaining the experimental observation index, experimental core index of the target product in the experimental group under the above traffic control method within the second period, and the control observation index and control core index of the target product in the control group;

[0019] Calculating a first difference between the experimental observation index of the experimental group and the standard observation index, and a second difference between the control observation index of the control group and the standard observation index;

[0020] If the absolute mean error between the first difference and the second difference is within a preset range, and the relationships between the experimental core indicators and standard core indicators of the experimental group and the control core indicators and standard core indicators of the control group satisfy a preset relationship, it is determined that the flow control algorithm meets the preset requirements.

[0021] In a third aspect, the present application provides a configuration method for flow control, the method comprising:

[0022] Receiving a target product selection instruction;

[0023] Determining a target product according to the target product selection instruction;

[0024] Receiving a configuration strategy configuration instruction for the target product, and configuring the flow control algorithm corresponding to the target product according to the configuration strategy configuration instruction, where the flow control algorithm at least includes the above-mentioned flow control method.

[0025] In one embodiment, the configuring the flow control algorithm corresponding to the target product according to the configuration strategy configuration instruction includes: configuring a corresponding algorithm type for the flow control algorithm corresponding to the target product according to the configuration strategy configuration instruction.

[0026] In one embodiment, the method further includes: configuring at least one of an execution time, a regulation scenario, and a target flow corresponding to the target product.

[0027] In a fourth aspect, the present application further provides a flow control device, the device comprising:

[0028] An acquisition module, configured to acquire a target product and a target flow of each of the target products;

[0029] A current flow acquisition module, configured to acquire the current flow of the target product;

[0030] A flow regulation module, configured to perform regulation processing on the current flow according to the target flow so that the current flow of the target product adapts to the target flow, where the regulation processing includes at least one of proportional regulation, integral regulation, and / or derivative regulation.

[0031] In a fifth aspect, the present application further provides an effect evaluation device for flow control, the method comprising:

[0032] A grouping module, configured to group target products to obtain an experimental group and a control group;

[0033] A standard index acquisition module, configured to acquire standard observation indicators and standard core indicators of the target products in the experimental group within a first period, and standard observation indicators and standard core indicators of the target products in the control group;

[0034] An experimental control index acquisition module, which is used in the second cycle, and the flow control algorithm of the experimental group at least includes the above-mentioned flow control method, and acquires the experimental observation index and experimental core index of the target product of the experimental group and the control observation index and control core index of the target product of the control group within the second cycle;

[0035] A first calculation module, which is used to calculate the first difference between the experimental observation index of the experimental group and the standard observation index, and the second difference between the control observation index of the control group and the standard observation index;

[0036] A determination module, which is used to determine that the flow control algorithm meets the preset requirements if the absolute average error between the first difference and the second difference is within the preset range, and the relationship between the experimental core index and standard core index of the experimental group and the control core index and standard core index of the control group meets the preset relationship.

[0037] In a sixth aspect, the present application further provides a flow control configuration device, and the device includes:

[0038] A receiving module, which is used to receive a target product selection instruction;

[0039] A target product determination module, which is used to determine the target product according to the target product selection instruction;

[0040] A configuration module, which is used to receive a configuration strategy configuration instruction for the target product, and configure the flow control algorithm corresponding to the target product according to the configuration strategy configuration instruction, and the flow control algorithm at least includes the above-mentioned flow control method.

[0041] In a seventh aspect, the present application further provides a computer device. The computer device includes a memory and a processor, the memory stores a computer program, and when the processor executes the computer program, it implements the steps of the method according to any one of claims 1 to 7, 8 or 9.

[0042] In an eighth aspect, the present application further provides a computer-readable storage medium. The computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, it implements the steps of the method according to any one of claims 1 to 7, 8 or 9.

[0043] In a ninth aspect, the present application further provides a computer program product. The computer program product includes a computer program, and when the computer program is executed by a processor, it implements the steps of the method according to any one of claims 1 to 7, 8 or 9.

[0044] The above-mentioned flow control method, device, equipment, storage medium and program product obtain the target product, the target flow rate and the current flow rate of each target product, and perform at least one of proportional adjustment, integral adjustment and / or derivative adjustment on the current flow rate according to the target flow rate, so that the current flow rate of the target product is adapted to the target flow rate. Through proportional adjustment, the difference between the current flow rate and the target flow rate of the target product gradually decreases. When the above difference decreases to a certain extent, the current flow rate may fluctuate above and below the target flow rate, that is, a constant deviation is generated. At this time, the integral adjustment is used to eliminate the deviation of the current flow rate fluctuating above and below the target flow rate, so that the value of the current flow rate no longer fluctuates, that is, a fixed value. At this time, the current flow rate may have a fixed difference from the target flow rate. At this time, the derivative adjustment is used to eliminate the fixed difference between the current flow rate and the target flow rate. The server jointly adjusts the current flow rate according to the adjustment parameters of the proportional adjustment, the adjustment parameters of the integral adjustment and the adjustment parameters of the derivative adjustment, so that the process of flow rate adjustment is smooth and controllable. Description of the Drawings

[0045] Figure 1 It is an application environment diagram of the flow control method in an embodiment;

[0046] Figure 2 It is a schematic flowchart of the flow control method in an embodiment;

[0047] Figure 3 It is a schematic architecture diagram of the flow control method in an embodiment;

[0048] Figure 4 It is a schematic diagram of the steps of the flow control method in another embodiment;

[0049] Figure 5 It is a schematic diagram of the adjustment process of the flow control method in an embodiment;

[0050] Figure 6 It is a schematic diagram of the global algorithm of the flow control method in an embodiment;

[0051] Figure 7 It is a structural block diagram of the flow control device in an embodiment;

[0052] Figure 8 It is a structural block diagram of the effect evaluation device for flow control in an embodiment;

[0053] Figure 9 It is a structural block diagram of the configuration device for flow control in an embodiment;

[0054] Figure 10 It is an internal structure diagram of a computer device in an embodiment. Detailed Implementation Modes

[0055] To make the objectives, technical solutions and advantages of the present application more clear and understandable, the present application will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and are not used to limit the present application.

[0056] The traffic control method provided by the embodiment of the present application can be applied to an application environment as Figure 1 shown. Among them, the terminal 102 communicates with the server 104 through the network. The data storage system can store the data that the server 104 needs to process. The data storage system can be integrated on the server 104, or can be placed in the cloud or other network servers. The user provides the terminal 102 to browse the product, generating traffic. The server 104 obtains the target product, the target traffic of each target product, and the current traffic of the target product; and adjusts the current traffic according to the target traffic so that the current traffic of the target product is adapted to the target traffic. The adjustment process includes at least one of proportional adjustment, integral adjustment, and / or derivative adjustment. Among them, the terminal 102 can be, but is not limited to, various personal computers, laptop computers, smart phones, tablet computers, Internet of Things devices, and portable wearable devices. The Internet of Things devices can be smart speakers, smart TVs, smart air conditioners, smart in-vehicle devices, etc. The portable wearable devices can be smart watches, smart bracelets, head-mounted devices, etc. The server 104 can be implemented by an independent server or a server cluster composed of multiple servers.

[0057] In one embodiment, as Figure 2 shown, a traffic control method is provided. Taking the method applied to the Figure 1 server 104 in it as an example for description, it includes the following steps:

[0058] Step 202, obtain the target product and the target traffic of each target product.

[0059] Among them, the target product is the object that needs to be traffic-regulated. The target product can be different objects such as commodities, live streamers, etc. The present embodiment does not limit the type of objects that need to be traffic-regulated. The target traffic is the exposure volume that the target product wants to achieve. This exposure volume is formulated according to the target product. The formulation method can be to learn and estimate the historical traffic of the target product according to some algorithms, or set by the user according to experience, and no specific limitation is made here.

[0060] Specifically, the server obtains at least one target product and the target traffic corresponding to each target product from the products.

[0061] Step 204, obtain the current traffic of the target product.

[0062] Among them, the current traffic is the exposure volume of the target product at the current moment. Taking the host as an example, this exposure volume can refer to the number of people in the host's live broadcast room at the current moment.

[0063] Step 206, adjust the current traffic according to the target traffic so that the current traffic of the target product is adapted to the target traffic. The adjustment process includes at least one of proportional adjustment, integral adjustment, and / or derivative adjustment.

[0064] Among them, proportional adjustment is to control the current traffic according to the error ratio based on the error between the current traffic and the target traffic of the target product. Integral adjustment is to calculate the sum of the errors between the current traffic and the target traffic and the corresponding errors during the traffic control within a period of time, and control the current traffic according to the sum of the errors. Derivative adjustment is to take the first derivative of the error between the predicted next traffic after adjusting the current traffic and the target traffic, and control the current traffic according to the result of the above first derivative. Among them, the predicted target traffic is predicted based on the regression model according to the historical time-series exposure characteristics of the target product and the own attribute characteristics of the target characteristics.

[0065] Specifically, the server calculates the error between the target traffic and the current traffic, and calculates the adjustment term of the proportional adjustment according to the above error. The server also calculates the adjustment term of the integral adjustment according to the sum of the differences between the current traffic and the target traffic corresponding to each adjustment process within a period of time. The server takes the first derivative of the error between the predicted next traffic after adjusting the current traffic and the target traffic, and calculates the adjustment term of the derivative adjustment according to the result of the above first derivative. The server jointly adjusts the current traffic according to the adjustment term of the proportional adjustment, the adjustment term of the integral adjustment, and the adjustment term of the derivative adjustment. Whenever the current traffic of the target product changes, step 202 is executed to make the current traffic of the target product adapted to the target traffic. Among them, the adjustment term is a specific adjustment value.

[0066] Specifically, the server calculates the error between the target flow and the current flow, and multiplies the above error by a positive constant, i.e., the proportional parameter Kp (representing proportionality), to calculate the adjustment term for proportional adjustment. The server also sums the differences between the current flow and the target flow corresponding to each adjustment process carried out within a period of time, and multiplies the sum of the above differences by a positive constant, i.e., the integral parameter Ip, to calculate the adjustment term for integral adjustment. The server takes the first derivative of the error between the next flow after adjusting the estimated current flow and the target flow, and multiplies the result of the above first derivative by a positive constant, i.e., the derivative parameter Dp, to calculate the adjustment term for derivative adjustment. The server adjusts the current flow jointly according to the adjustment term for proportional adjustment, the adjustment term for integral adjustment, and the adjustment term for derivative adjustment. Whenever the current flow of the target product changes, step 202 is executed to make the current flow of the target product match the target flow. It should be noted that the proportional parameter Kp in this embodiment is the core main force among the three parameters. Proportional control is an immediate control. As long as there is a deviation between the current flow and the target flow, the control quantity is immediately output. Increasing Kp can accelerate the system response and reduce the error between the current flow and the target flow, but the overshoot of the system will increase and the stability will deteriorate. The integral parameter Ip is the general main force among the three parameters. Integral control is a repair control. As long as there is a deviation between the current flow and the target flow, it will gradually control in the direction of eliminating the deviation. Decreasing Ip can reduce the overshoot and enhance the stability of the system. The derivative parameter Dp is the reserve personnel among the three parameters. Derivative control is a pre-control, which is based on the rate of change of the deviation for control and is generally less used. Increasing Kd can accelerate the system response. The determination of the three parameters is based on historical data for simulation. First, the proportional parameter is adjusted, then the integral parameter is adjusted, and finally the derivative parameter is adjusted. Observe the standard deviation of the error between the current flow and the target flow of the adjusted target product and the number of adjustments to reach the target flow to determine the reasonable range of the parameters.

[0067] In the above flow control method, the target product and the target flow and the current flow of each target product are obtained, and at least one of proportional adjustment, integral adjustment, and / or derivative adjustment is performed on the current flow according to the target flow to make the current flow of the target product match the target flow. Through proportional adjustment, the difference between the current flow and the target flow of the target product is gradually reduced. When the above difference is reduced to a certain extent, the current flow may fluctuate above and below the target flow, that is, a constant deviation is generated. At this time, the integral adjustment is used to eliminate the deviation of the current flow fluctuating above and below the target flow, so that the value of the current flow no longer fluctuates, that is, a fixed value. At this time, the current flow may have a fixed difference from the target flow. At this time, the derivative adjustment is used to eliminate the fixed difference between the current flow and the target flow. The server jointly adjusts the current flow according to the adjustment term for proportional adjustment, the adjustment term for integral adjustment, and the adjustment term for derivative adjustment, so that the process of flow adjustment is smooth and controllable.

[0068] In one embodiment, it further includes: obtaining adjustment parameters obtained by performing adjustment processing on a target product, where the adjustment parameters include at least one of a proportional adjustment parameter, an integral adjustment parameter, and a derivative adjustment parameter, and calculating a first score according to the adjustment parameters.

[0069] Among them, the proportional adjustment parameter, the integral adjustment parameter, and the derivative adjustment parameter are all obtained by simulating based on historical data.

[0070] Specifically, when the server obtains the proportional adjustment, integral adjustment, and / or derivative adjustment of the current flow of the target product, according to the error (Pterm) between the target flow and the current flow in the adjustment term of the proportional adjustment, the sum (Iterm) of the differences between the current flow and the target flow corresponding to each adjustment process within a period of time in the adjustment term of the integral adjustment, and the result of taking the first derivative of the error between the estimated next flow after the current flow adjustment and the target flow in the adjustment term of the derivative adjustment, the corresponding proportional adjustment parameter (Kp), integral adjustment parameter (Ip), and derivative adjustment parameter (Dp) are obtained. Among them, the parameter acquisition method is to obtain by simulating based on historical data. According to the error (Pterm) in the adjustment term of the above-mentioned proportional adjustment and the corresponding proportional adjustment parameter (Kp), the sum (Iterm) of the differences in the adjustment term of the integral adjustment and the corresponding integral adjustment parameter (Ip), the result of the first derivative (Dterm) in the adjustment term of the derivative adjustment and the corresponding derivative parameter (Dp), the first parameter (P1) of the first score P is calculated. Specifically, the calculation formula of the first parameter is as follows:

[0071] P1 = Kp * PTerm + Ip * ITerm + Dp * DTerm (1)

[0072] The server obtains the time T1 when the flow of the target product is first adjusted, and the time T2 when the flow of the target product is currently adjusted. The calculation formula of P obtained according to P1, T1, and T2 is as follows:

[0073] P = P1 * EXP(-(300 - (T1 - T2)) / K) (2)

[0074] Among them, EXP(-(300 - (T1 - T2)) / K) is the attenuation processing of the current timestamp - the first exposure adjustment timestamp within 5 minutes, and K is the attenuation adjustment parameter.

[0075] Calculate a second score for the target product according to user preferences; calculate the score of the target product according to the first score and the second score of the target product; sort the target products according to the scores, and adjust the flow of the target product according to the sorting result.

[0076] Specifically, the server obtains the preferences of the users who log in to the page of the target product, and calculates the second score (score) of the target product according to the user preferences. The server determines the score of the target product based on the first score and the second score of the target product, and sorts the target products according to the scores of the target products. The target products are displayed on the user's interface in turn according to the sorting result, so as to realize the traffic regulation of the target products. Among them, the calculation formula for the score F of the target product is:

[0077] F = r * P + score (3)

[0078] Among them, r is the adjustment coefficient.

[0079] In the above traffic control method, the target products are sorted according to the scores, and the traffic of the target products is adjusted in turn according to the sorting rules, so that the realization of the traffic regulation is smoother, more controllable and less abrupt.

[0080] In one embodiment, the server obtains the products to be processed; selects the target products from the products to be processed.

[0081] Among them, the products to be processed are all products or live streamers.

[0082] Specifically, the server obtains the products to be processed, and the vector recall model selects the target products from the products to be processed.

[0083] In the specific implementation process, the server obtains the products to be processed and the processing principles for processing the products to be processed, and synchronizes the above products to be processed and the processing principles to the ES index offline in batches. The server screens the products to be processed therein through at least one of the processing principles in the ES index to determine the initial products. The above processing principles include product selection rules (item embedding), product tags and filtering options. The server can determine the target products according to the business rules. Optionally, the server can also further screen the initial products according to the scores to determine the target products. Among them, the formula for determining the number of target products is:

[0084] The number of target products = the number of initial products / k (4)

[0085] For example, the server uses the item embeddings in the ES index to select products to be processed. The server obtains user preferences (user embeddings) online through model inference. The server then combines the user embeddings with the item embeddings in the server's ES index to select the initial products. Based on the number of initial products, the server uses a formula to determine the number of target products to obtain the number of target products. The server then further selects the initial products according to the corresponding scores based on the number of target products to determine the target products.

[0086] In the above flow control method, a target product is selected from the products to be processed, and flow control can be performed according to the characteristics of the target product in a targeted manner.

[0087] In one embodiment, obtaining a target flow rate of a target product includes: determining an initial target flow rate based on historical flow rates of the target product; obtaining a group to which the target product belongs; determining a position of the target product within the group; and adjusting the initial target flow rate based on the determined position to obtain the target flow rate.

[0088] Specifically, the server predicts the target product's future traffic, or initial target traffic, based on its historical traffic. The server groups the target products according to their scores and determines their positions within the groups based on their tags. The server obtains the target product's tags and corresponding scores, determines the target product's group and position within the group based on these scores and tags, and adjusts the target product's initial target traffic according to the corresponding adjustment rules for the group and position within the group to obtain the corresponding target traffic.

[0089] In the specific implementation process, the server labels newly added anchors or newly listed products as "new products", labels the products of merchants who are about to leave the platform or anchors who are about to leave the platform as "pre-churn", labels the products of merchants who have violated regulations or anchors who have violated regulations as "violation", classifies target products labeled with "new products" and "pre-churn" into support positions, classifies target products labeled with "violation" into suppression positions, and divides target products into 50 segments according to scores. The server predicts the future traffic of the target product, i.e., the initial target traffic, based on the historical traffic of the target product. The server obtains the label of target product 1 as "new product" and the score as X. Based on X, the server determines the specific segment of the target product 1 among the 50 segments, determines the target product as a support position based on "new product", and adjusts the initial traffic according to the adjustment rules corresponding to the support position information of the specific segment to obtain the corresponding target traffic.

[0090] Among them, the adjustment rule corresponding to the support position is: obtain the conversion rate X of the traffic orders of the target product in the most recent 7 days and the average value G of the conversion rates of the traffic orders of the corresponding group. If X > G, then increase the initial traffic. The increase amplitude is (X - G) / G, and the increase amplitude does not exceed the configured adjustment value (15%). Among them, G is determined according to the segment where the target product is located in 50 segments. The initial traffic after the increase is the target traffic.

[0091] Among them, the adjustment rule corresponding to the suppression position is: obtain the conversion rate X of the traffic orders of the target product in the most recent 7 days and the average value G of the conversion rates of the traffic orders of the corresponding group. If X < G, then decrease the initial traffic. The decrease amplitude is (X - G) / G, and the decrease amplitude does not exceed the configured adjustment value (15%). Among them, G is determined according to the segment where the target product is located in 50 segments. The initial traffic after the decrease is the target traffic. It can be understood that the configured adjustment value can be changed, not limited to 15%, and can be adjusted according to the actual situation.

[0092] In the above traffic control method, the initial target traffic is adjusted according to the position to obtain the target traffic, so that the adjustment granularity of the target traffic is fine and targeted.

[0093] In one embodiment, obtaining the target traffic of the target product includes: obtaining the configured traffic in the configuration data of the target product as the target traffic.

[0094] Among them, the configuration data is the data in the configuration policy. The server controls the traffic of the target product according to the configuration policy. The configuration policy can be automatically generated by the server according to the algorithm, or a corresponding configuration policy can be customized for the target product according to the corresponding input.

[0095] Specifically, the server obtains the configuration data of the target product and takes the configured traffic as the target traffic. The above configured traffic can be a value customized according to business requirements.

[0096] In the above traffic control method, taking the configured traffic in the configuration data as the target traffic makes the determination method of the target traffic diverse and can meet the determination methods of the target traffic in different business scenarios.

[0097] After obtaining the current traffic of the target product, one embodiment further includes: when the algorithm type is the target algorithm, then continue to adjust the current traffic according to the target traffic so that the current traffic of the target product adapts to the target traffic; when the algorithm type is the global algorithm, then adjust the current traffic according to the target traffic through the global algorithm.

[0098] Specifically, after the server obtains the current traffic of the target product, it obtains the configuration information of the target product. The configuration information includes the type of traffic control algorithm of the target product. When the algorithm type is the target algorithm, the current traffic is continuously adjusted according to the target traffic to make the current traffic of the target product match the target traffic. When the algorithm type is the global algorithm, the current traffic is adjusted according to the target traffic through the global algorithm. Among them, the target algorithm is to adjust the current traffic according to the target traffic to make the current traffic of the target product match the target traffic, and the adjustment process includes at least one of proportional adjustment, integral adjustment, and / or derivative adjustment.

[0099] The target algorithm in the above traffic control method can achieve smooth control of the traffic of the target product, and the global algorithm can achieve control of the global optimal traffic in the overall sorting stage. Different algorithm types can meet different traffic control requirements.

[0100] The adjustment of the current traffic according to the target traffic through the global algorithm in the above embodiment includes: obtaining the target product and the target traffic, current traffic, and previous regulation parameter of each target product; updating the regulation parameter according to the target traffic, current traffic, and previous regulation parameter; obtaining the sorting score corresponding to the target product; and adjusting the current traffic according to the updated regulation parameter and sorting score of the target product to make the current traffic of the target product match the target traffic.

[0101] Specifically, the server obtains the target product in the same way as in the above embodiment. After the server obtains the target product, it sorts the obtained target product through the sorting service to obtain a list of target products with sorting parameters. The server generates a corresponding traffic control strategy according to the target products in the target product list. The above traffic control strategy includes the planning of the target traffic of the target product. The server obtains the target traffic in the traffic control strategy of the target product in the target product list, the real-time exposure volume of the target product, that is, the current traffic, and the previous regulation parameter, and updates the current regulation parameter in real time according to the target traffic, current traffic, and previous regulation parameter of the target product. The specific calculation formula of the current regulation parameter is:

[0102] Current regulation parameter = previous regulation parameter - adjustment factor * (target traffic - current traffic) (5)

[0103] Among them, the adjustment factor is obtained according to experimental data and can be adjusted manually. The server obtains the sorting parameters corresponding to each target product in the list according to the list of target products with sorting parameters. Optionally, the sorting parameter is a sorting score, that is, the server obtains the sorting scores corresponding to each target product in the list according to the list of target products with sorting scores. The server calculates the allocation probability according to the updated regulation parameters and sorting scores of the target products, and adjusts the current traffic according to the allocation probability of the target products so that the current traffic of the target products is adapted to the target traffic. The specific calculation formula of the allocation probability is:

[0104] Allocation probability = Exp[(sorting score - current regulation parameter) / regularization parameter] (6)

[0105] Among them, the regularization parameter is obtained according to experimental data and can be adjusted manually.

[0106] The above traffic control method can achieve the control of the global optimal traffic in the overall sorting stage through a global algorithm.

[0107] After regulating the traffic of the target products through the above traffic control method, the effect of the traffic control can also be evaluated. The traffic control effect evaluation method in this embodiment includes: grouping the target products to obtain an experimental group and a control group; obtaining the standard observation index and standard core index of the target products in the experimental group within the first cycle, and the standard observation index and standard core index of the target products in the control group; obtaining the experimental observation index, experimental core index generated by the target products in the experimental group under the traffic control method in any of the above embodiments within the second cycle, and the control observation index and control core index of the target products in the control group; calculating the first difference between the experimental observation index of the experimental group and the standard observation index, and the second difference between the control observation index of the control group and the standard observation index; if the absolute average error between the first difference and the second difference is within a preset range, and the relationship between the experimental core index and standard core index of the experimental group and the control core index and standard core index of the control group meets a preset relationship, it is determined that the traffic control algorithm meets the preset requirements.

[0108] Specifically, the server performs double sharding (hash AB sharding) on the target product and the users of the target product to obtain an experimental group and a control group. Among them, in the experimental group, the target product of group A is recommended to the users of group A, and in the control group, the target product of group B is recommended to the users of group B. In the first cycle, the server uses the first traffic control algorithm to adjust the current traffic of the target products in the experimental group and the control group, and obtains the standard observation indicators and standard core indicators corresponding to the experimental group and the control group. Among them, the standard observation indicator is the absolute average error between the actual traffic on the same day and the target traffic, which is the regulation observation indicator for the target product. The standard core indicators include: traffic revenue (usually measured by sales), the number of target products, the number of clicks on the target products, the number of completed target products, and the traffic Gini coefficient. Among them, the traffic Gini coefficient is to queue all products in ascending order of the traffic they obtain, divide them into n groups with equal numbers, and the proportion of the cumulative traffic obtained by the products from the first group to the i-th group in the total traffic obtained by all products is Wi. The formula for the traffic Gini coefficient (G expo ) is:

[0109]

[0110] In the second cycle, the server uses the target algorithm to adjust the current traffic of the target products in the experimental group, and obtains the experimental observation indicators and experimental core indicators generated by the experimental group; for the control group, the server continues to use the first traffic control algorithm to adjust the current traffic of the target products in the control group, and obtains the control observation indicators and control core indicators of the target products in the control group.

[0111] The server calculates the first difference between the experimental observation indicator of the experimental group and the standard observation indicator, and the second difference between the control observation indicator of the control group and the standard observation indicator. The server judges whether the absolute average error between the first difference and the second difference is within the preset range, and also needs to judge whether the relationship between the experimental core indicators of the experimental group and the standard core indicators and the control core indicators of the control group and the standard core indicators meets the preset relationship. If the absolute average error between the first difference and the second difference is within the preset range, and the relationship between the experimental core indicators of the experimental group and the standard core indicators and the control core indicators of the control group and the standard core indicators meets the preset relationship, it is determined that the traffic control algorithm meets the preset requirements. Among them, the preset relationship is that the traffic revenue does not decrease, the number of target products / the number of clicks on the target products increases, and the traffic Gini coefficient decreases. It can be understood that the preset range and the preset relationship are obtained based on historical experience and can be adjusted as the historical data changes. If the absolute average error between the first difference and the second difference is not within the preset range, or the relationship between the experimental core indicators of the experimental group and the standard core indicators and the control core indicators of the control group and the standard core indicators does not meet the preset relationship, the parameters in the traffic control algorithm are adjusted, and the next effect evaluation is carried out.

[0112] The method for evaluating the effect of traffic control determines the influence of the traffic control algorithm applied in the experimental group through the experimental group and the control group, or feedbacks the experimental results for corresponding parameter adjustment, increasing the reliability and stability of the traffic control algorithm.

[0113] Before traffic control of the target product, it also includes a configuration method for traffic control. The method includes: receiving a target product selection instruction; determining the target product according to the target product selection instruction; receiving a configuration policy configuration instruction for the target product, and configuring the traffic control algorithm corresponding to the target product according to the configuration policy configuration instruction, where the traffic control algorithm is any one of the traffic control methods in the above embodiments.

[0114] Specifically, the server receives the target product selection instruction input by the user and determines the target product from the products according to the target product selection instruction. Among them, the method for determining the target product is the same as the determination method in the above embodiments and will not be elaborated here. After the server determines the target product, it receives the traffic control algorithm input by the user for the target product for configuration, and the server adjusts the current traffic of the target product according to the configured traffic control algorithm so that the current traffic of the target product is adapted to the target traffic.

[0115] In the above traffic control configuration method, the server receives the configuration policy configuration instruction for the target product, making the configuration policy variable, and the configuration policy can be changed according to different requirements.

[0116] Configuring the traffic control algorithm corresponding to the target product according to the configuration policy configuration instruction in the above embodiments includes: configuring the corresponding algorithm type for the traffic control algorithm corresponding to the target product according to the configuration policy configuration instruction.

[0117] Specifically, the server configures the corresponding algorithm type for the traffic control algorithm corresponding to the target product as the target algorithm or the global algorithm according to the configuration policy configuration instruction.

[0118] In the above traffic control configuration method, the server receives the configuration policy configuration instruction for the target product, configures the type of the traffic control algorithm, and configures different algorithm types according to different requirements.

[0119] In one embodiment, it further includes: configuring at least one of the execution time, regulation scenario, and target traffic corresponding to the target product.

[0120] Specifically, the server also configures at least one of the execution time, regulation scenario, and target traffic for the target product to execute the corresponding traffic control algorithm according to the configuration policy configuration instruction.

[0121] The above configuration method for traffic control. The server receives a configuration policy configuration instruction for a target product, enabling the configuration policy to be variable, and the configuration policy can be changed according to different requirements.

[0122] In one embodiment, a traffic control method is applied to a live streaming platform. Through Figure 3 and Figure 4 、 Figure 5 it is described as follows.

[0123] The overall architecture of the traffic control algorithm is as shown in Figure 3 . The server determines target streamers who need traffic control from the streamers on the live streaming platform through business rules or intelligent algorithms. The server configures a traffic control policy for the target streamers through a configuration policy configuration instruction. The traffic control configuration policy includes a regulation scenario, an execution time for traffic regulation, a target traffic volume, and an algorithm type. If the label of the target streamer is a potentially lost streamer, the formal description of the configuration policy of the configured target product can be: regulation scenario (category list page), execution time for traffic regulation (20210904 - 20210930), the target traffic volume which is the 5 - minute exposure volume of a single streamer (20), algorithm type (target algorithm). The server controls the traffic of the target product according to the algorithm type in the configuration policy corresponding to the target product, and evaluates the effect of the traffic control algorithm.

[0124] Among them, the method by which the server determines target streamers who need traffic control from the streamers on the live streaming platform through a recall matching algorithm is as shown in Figure 4 . The server synchronizes streamers (items), god emebedding (the selection rule for initial target streamers), and target population tag labels (labels of streamers) offline in batches to the ES index. The server obtains user embedding (users interested in content) through the online inference method of the model (model). The server filters the streamers in the ES based on the combination of godembedding and user embedding to obtain the top N items, which are the initial target streamers. The server obtains the number of target streamers according to the formula (4) for determining the number of target streamers based on the number of initial target streamers, and further filters the initial target streamers according to the corresponding scores based on the number of target products to determine the target streamers.

[0125] Among them, the method for determining the target traffic in the traffic control configuration strategy: First, the server obtains the historical time-series data of the target anchor, and estimates the target traffic of the target anchor based on the historical time-series data. The average traffic of the target anchor on the (T + 7)th day is predicted through a model, and this average traffic is used as the initial target traffic. Next, the server adjusts the initial target traffic according to the corresponding adjustment rules based on the tags of the target anchor and obtains the target traffic. Adjustment rule 1: Obtain the score of the target anchor, divide the target anchor into 50 groups according to the score of the target anchor, and each group has a corresponding average order conversion rate G; obtain the tags of the target anchor, and determine whether the position of the target anchor belongs to a suppression position or a support position according to the tags of the target anchor; Adjustment steps for the suppression position: Obtain the exposure order conversion rate X of the target anchor in the recent 7 days, judge the relationship between X and G. If X < G, then the initial target traffic of the target product is lowered, and the lowering value is (X - G) / G, but this lowering value does not exceed the configured adjustment value (15%); Adjustment steps for the support position: Obtain the exposure order conversion rate X of the target anchor in the recent 7 days, judge the relationship between X and G. If X > G, then the initial target traffic of the target product is increased, and the increasing value is (X - G) / G, but this increasing value does not exceed the configured adjustment value (15%). Adjustment rule 2: Adjust the initial target traffic of the target product according to the configuration instruction.

[0126] Among them, the server controls the traffic of the target product according to the algorithm type in the corresponding configuration strategy of the target product, and takes the target algorithm as an example for description. Based on each traffic control request, the server obtains the current traffic real_pv and the target traffic des_pv of each anchor in the target anchor List. If the current traffic of the target anchor is greater than the target traffic within 5 minutes, truncation processing is performed. If the current traffic of the target anchor is less than the target traffic, the server calculates the error term Pterm, the integral term Iterm, and the differential term DTerm of each target anchor respectively, and outputs P1 = Kp * PTerm + Ip * ITerm + Dp * DTerm, and normalizes it from 0 to 1; Kp is the error term parameter, Ip is the integral term parameter, and Dp is the differential term parameter. The server obtains the timestamp T1 when the first traffic adjustment is recorded, and performs attenuation processing on the current timestamp - the first traffic adjustment timestamp within 5 minutes, P = P1 * EXP(-(300 - (current timestamp - first traffic adjustment timestamp)) / K), where K is the attenuation adjustment parameter. The server calculates the score F of the target product (F = r * P + score), and sorts and displays the target anchors through the score. r is the adjustment coefficient. As the number of cycles of the server adjusting the current traffic increases, the variance of the error between the current traffic and the target traffic of each target anchor gradually shrinks, and the closer the current traffic is to the target traffic, the slower the traffic adjustment speed, as Figure 5 shown.

[0127] Among them, the server controls the traffic of the target product according to the algorithm type in the configuration policy corresponding to the target product. Taking the algorithm type as the global algorithm as an example, as Figure 6 shown. The method for the server to obtain the target anchor and the configuration policy of the target anchor is the same as the above-mentioned obtaining method, which will not be elaborated here. The server obtains the target traffic redis cache of the target anchor, statistically calculates the current traffic of the target anchor in real time, and updates the regulation parameters in real time according to the actual traffic, that is, the current traffic and the target traffic, and caches them in redis. The calculation formula of the current regulation parameter is formula (5). The server obtains the sorting score of each target anchor through the sorting model according to the recalled target anchor List, recalculates the allocation probability according to the sorting score and the current regulation parameter according to formula (6), and then outputs the target anchor List according to the allocation probability. Among them, the smaller the regulation factor in formula (6), the closer the order of the target anchor List output after regulation is to the order of the target anchor List output by the original sorting model.

[0128] Among them, the method for evaluating the effect of the algorithm for traffic control is as follows: The server adopts a dual-traffic-splitting method for users and target anchors. The specific steps are as follows:

[0129] 1. The server performs hash AB splitting on users and target anchors. Group A users are recommended to group A target anchors, and group B users are recommended to group B target anchors.

[0130] 2. The server monitors the stability of the core indicators a and core indicators b of the experimental group and the control group in the first cycle of 1-2 weeks. The core indicator a is the regulation observation indicator for the target anchor, which is expressed as the absolute average error between the actual traffic and the target traffic on the current day. The core indicator b is the observation core indicator for the optimization of the sorting belt exposure constraint, which is expressed as the traffic revenue (usually measured by sales), the number of target anchors, the number of people clicking on the target anchor, the number of target anchors with completed orders, and the traffic Gini coefficient. Among them, for the traffic Gini coefficient, all target anchors are queued in ascending order of the traffic they obtain, divided into n groups with equal numbers of people. The proportion of the cumulative traffic obtained by the target anchors from the first group to the i-th group in the total traffic obtained by all target anchors is Wi. Then the calculation formula of the Gini coefficient is formula (7).

[0131] 3. The server adds a regulation strategy to the A experimental group in the second cycle of 2 weeks.

[0132] 4. Compare the core metrics a and b before and after to draw conclusions on the regulation experiment. If the absolute average error between the actual traffic and the target traffic is within the acceptable range of the business, and when the traffic revenue does not decrease, the number of target streamers / clicks on target streamers increases, and the traffic Gini coefficient decreases, it indicates that the regulation strategy is effective; otherwise, adjust the parameters in the regulation strategy. The above traffic control method obtains the target products, the target traffic and the current traffic of each target product, and performs at least one of proportional regulation, integral regulation, and / or differential regulation on the current traffic according to the target traffic, so that the current traffic of the target product adapts to the target traffic. Through proportional regulation, the difference between the current traffic and the target traffic of the target product gradually decreases. When the difference decreases to a certain extent, the current traffic may fluctuate above and below the target traffic, that is, a constant deviation is generated. At this time, the integral regulation is used to eliminate the deviation of the current traffic fluctuating above and below the target traffic, so that the value of the current traffic no longer fluctuates, that is, a fixed value. At this time, there may be a fixed difference between the current traffic and the target traffic. At this time, the differential regulation is used to eliminate the fixed difference between the current traffic and the target traffic. The server adjusts the current traffic according to the regulation parameters of the proportional regulation, the regulation parameters of the integral regulation, and the regulation parameters of the differential regulation together, so that the process of traffic regulation is smooth and controllable.

[0133] It should be understood that although the steps in the flowcharts involved in the above-described embodiments are shown in sequence according to the arrows, these steps are not necessarily executed in the order indicated by the arrows. Unless there is a clear indication in this article, there is no strict order restriction for the execution of these steps, and these steps can be executed in other orders. Moreover, at least some of the steps in the flowcharts involved in the above-described embodiments may include multiple steps or multiple stages. These steps or stages are not necessarily executed at the same time, but can be executed at different times. The execution order of these steps or stages is not necessarily sequential, but can be executed alternately or alternately with at least some of the steps or stages in other steps or other steps.

[0134] Based on the same inventive concept, the embodiments of the present application also provide a traffic control device for implementing the above-mentioned traffic control method. The solution provided by this device to solve the problem is similar to the solution described in the above method. Therefore, the specific limitations in one or more embodiments of the traffic control device provided below can refer to the limitations on the traffic control method in the above text, and will not be repeated here.

[0135] In one embodiment, as Figure 7As shown, a flow control device is provided, including: an acquisition module 100, a current flow acquisition module 200, and a flow regulation module 300, where: The acquisition module 100 is used to acquire the target product and the target flow of each target product. The current flow acquisition module 200 is used to acquire the current flow of the target product. The flow regulation module 300 is used to adjust the current flow according to the target flow so that the current flow of the target product is adapted to the target flow, and the adjustment process includes at least one of proportional adjustment, integral adjustment, and / or differential adjustment.

[0136] In one embodiment, it further includes: a first score calculation module, which is used to acquire the adjustment parameters obtained by adjusting the target product, and the adjustment parameters include at least one of a proportional adjustment parameter, an integral adjustment parameter, and a differential adjustment parameter, and calculate a first score according to the adjustment parameters; a second score calculation module, which is used to calculate a second score for the target product according to user preferences; a score calculation module, which is used to calculate the score of the target product according to the first score and the second score of the target product; a sorting adjustment module, which is used to sort the target products according to the scores and adjust the flow of the target products according to the sorting results.

[0137] In one embodiment, the acquisition module 100 includes: a first acquisition module, which is used to acquire the product to be processed. A selection module, which is used to select the target product from the products to be processed.

[0138] In one embodiment, the acquisition module 100 includes: a first determination module, which is used to determine the initial target flow according to the historical flow of the target product. A grouped acquisition module, which is used to acquire the group to which the target product belongs. A position determination module, which is used to determine the position of the target product within the group. A target flow adjustment module, which is used to adjust the initial target flow according to the determined position to obtain the target flow.

[0139] In one embodiment, the acquisition module 100 includes: a second determination module, which is used to acquire the configured flow in the configuration data of the target product as the target flow.

[0140] In one embodiment, it further includes: a first adjustment module, which is used to continue to adjust the current flow according to the target flow when the algorithm type is the target algorithm so that the current flow of the target product is adapted to the target flow; a second adjustment module, which is used to adjust the current flow according to the target flow through the global algorithm when the algorithm type is the global algorithm.

[0141] In one embodiment, the second adjustment module includes: a data acquisition module, configured to acquire a target product, the target traffic, the current traffic, and the previous regulation parameter of each target product; an update module, configured to update the regulation parameter according to the target traffic, the current traffic, and the previous regulation parameter; a branch acquisition module, configured to acquire a sorting score corresponding to the target product; and a second adjustment execution module, configured to adjust the current traffic according to the updated regulation parameter and the sorting score of the target product, so that the current traffic of the target product is adapted to the target traffic.

[0142] In one embodiment, an effect evaluation device for traffic control is provided, including: a grouping module 400, a standard index acquisition module 500, an experimental control index acquisition module 600, a first calculation module 700, and a determination module 800; the grouping module 400 is configured to group target products to obtain an experimental group and a control group; the standard index acquisition module 500 is configured to acquire the standard observation index and the standard core index of the target products in the experimental group in the first period, and the standard observation index and the standard core index of the target products in the control group; the experimental control index acquisition module 600 is configured to, in the second period, the traffic control algorithm of the experimental group at least includes the traffic control method in any one of the above method embodiments, and acquire the experimental observation index and the experimental core index of the target products in the experimental group and the control observation index and the control core index of the target products in the control group in the second period; the first calculation module 700 is configured to calculate a first difference between the experimental observation index and the standard observation index of the experimental group, and a second difference between the control observation index and the standard observation index of the control group; the determination module 800 is configured to determine that the traffic control algorithm meets the preset requirements if the absolute average error between the first difference and the second difference is within the preset range, and the relationship between the experimental core index and the standard core index of the experimental group and the control core index and the standard core index of the control group meets the preset relationship.

[0143] In one embodiment, a configuration device for traffic control is provided, including: a receiving module 900, a target product determination module 1000, and a configuration module 1100; the receiving module 900 is configured to receive a target product selection instruction; the target product determination module 1000 is configured to determine a target product according to the target product selection instruction; the configuration module 1100 is configured to receive a configuration strategy configuration instruction for the target product, and configure a traffic control algorithm corresponding to the target product according to the configuration strategy configuration instruction, and the traffic control algorithm at least includes the traffic control method in any one of the above method embodiments.

[0144] In one embodiment, the configuration module includes:

[0145] An algorithm type configuration module, configured to configure a corresponding algorithm type for the traffic control algorithm corresponding to the target product according to the configuration strategy configuration instruction.

[0146] In one embodiment, it further includes: a first configuration module, configured to configure at least one of an execution time, a regulation scenario, and a target flow rate corresponding to a target product.

[0147] Each module in the above flow control device can be implemented in whole or in part by software, hardware, and their combination. Each of the above modules can be embedded in or independent of a processor in a computer device in hardware form, or stored in a memory in a computer device in software form, so as to facilitate the processor to call and execute the operations corresponding to each of the above modules.

[0148] In one embodiment, a computer device is provided. The computer device can be a terminal, and its internal structure diagram can be as Figure 10 shown. The computer device includes a processor, a memory, a communication interface, a display screen, and an input device connected through a system bus. Among them, the processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system and a computer program. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The communication interface of the computer device is used to communicate with an external terminal in a wired or wireless manner, and the wireless manner can be implemented through WIFI, a mobile cellular network, NFC (Near Field Communication), or other technologies. When the computer program is executed by the processor, it implements a flow control method. The display screen of the computer device can be a liquid crystal display screen or an electronic ink display screen, and the input device of the computer device can be a touch layer covered on the display screen, or a button, a trackball, or a touchpad provided on the shell of the computer device, or an external keyboard, touchpad, or mouse, etc.

[0149] Those skilled in the art can understand that Figure 10 the structure shown in

[0150] is only a block diagram of a part of the structure related to the solution of this application, and does not constitute a limitation on the computer device to which the solution of this application is applied. The specific computer device may include more or fewer components than those shown in the figure, or combine certain components, or have different component arrangements.

[0151] In one embodiment, when the processor executes the computer program, the following steps are further implemented: obtaining the adjustment parameters obtained by adjusting the target product, where the adjustment parameters include at least one of a proportional adjustment parameter, an integral adjustment parameter, and a derivative adjustment parameter, calculating a first score according to the adjustment parameters; calculating a second score for the target product according to user preferences; calculating the score of the target product according to the first score and the second score of the target product; sorting the target products according to the scores, and adjusting the flow rate of the target product according to the sorting result.

[0152] In one embodiment, the obtaining of the target product implemented when the processor executes the computer program includes: obtaining the product to be processed; selecting the target product from the products to be processed.

[0153] In one embodiment, the obtaining of the target flow rate of the target product implemented when the processor executes the computer program includes: determining an initial target flow rate according to the historical flow rate of the target product; obtaining the group to which the target product belongs; determining the position of the target product within the group; adjusting the initial target flow rate according to the determined position to obtain the target flow rate.

[0154] In one embodiment, the obtaining of the target flow rate of the target product implemented when the processor executes the computer program includes: obtaining the configured flow rate in the configuration data of the target product as the target flow rate.

[0155] In one embodiment, after obtaining the current flow rate of the target product implemented when the processor executes the computer program, it further includes: when the algorithm type is the target algorithm, then continue to adjust the current flow rate according to the target flow rate so that the current flow rate of the target product matches the target flow rate; when the algorithm type is the global algorithm, then adjust the current flow rate according to the target flow rate through the global algorithm.

[0156] In one embodiment, the adjusting the current flow rate according to the target flow rate through the global algorithm implemented when the processor executes the computer program includes: obtaining the target product and the target flow rate, the current flow rate, and the previous regulation parameter of each target product; updating the regulation parameter according to the target flow rate, the current flow rate, and the previous regulation parameter; obtaining the sorting score corresponding to the target product; adjusting the current flow rate according to the updated regulation parameter and the sorting score of the target product so that the current flow rate of the target product matches the target flow rate.

[0157] In one embodiment, a method for evaluating the effect of traffic control is provided. When the processor executes a computer program, the following steps are further implemented: grouping the target products to obtain an experimental group and a control group; obtaining the standard observation indexes and standard core indexes of the target products in the experimental group within the first period, and the standard observation indexes and standard core indexes of the target products in the control group; obtaining the experimental observation indexes, experimental core indexes of the target products in the experimental group under any one of the traffic control methods in the above embodiments within the second period, and the control observation indexes and control core indexes of the target products in the control group; calculating the first difference between the experimental observation indexes and the standard observation indexes of the experimental group, and the second difference between the control observation indexes and the standard observation indexes of the control group; if the absolute average error between the first difference and the second difference is within a preset range, and the relationship between the experimental core indexes and standard core indexes of the experimental group and the control core indexes and standard core indexes of the control group meets a preset relationship, it is determined that the traffic control algorithm meets the preset requirements.

[0158] In one embodiment, a method for configuring traffic control is provided. When the processor executes a computer program, the following steps are further implemented: receiving a target product selection instruction; determining the target product according to the target product selection instruction; receiving a configuration strategy configuration instruction for the target product, and configuring the traffic control algorithm corresponding to the target product according to the configuration strategy configuration instruction. The traffic control algorithm at least includes any one of the traffic control methods in the above embodiments.

[0159] In one embodiment, configuring the traffic control algorithm corresponding to the target product according to the configuration strategy configuration instruction implemented when the processor executes the computer program includes: configuring the corresponding algorithm type for the traffic control algorithm corresponding to the target product according to the configuration strategy configuration instruction.

[0160] In one embodiment, when the processor executes the computer program, the following steps are further implemented: configuring at least one of the execution time, regulation scenario, and target traffic corresponding to the target product.

[0161] In one embodiment, a computer-readable storage medium is provided, on which a computer program is stored. When the computer program is executed by a processor, the following steps are implemented: obtaining the target products and the target traffic of each target product; obtaining the current traffic of the target product; adjusting the current traffic according to the target traffic so that the current traffic of the target product is adapted to the target traffic. The adjustment process includes at least one of proportional adjustment, integral adjustment, and / or differential adjustment.

[0162] In one embodiment, when the computer program is executed by a processor, the following steps are further implemented: obtaining adjustment parameters obtained by performing adjustment processing on a target product, where the adjustment parameters include at least one of a proportional adjustment parameter, an integral adjustment parameter, and a derivative adjustment parameter, and calculating a first score based on the adjustment parameters; calculating a second score for the target product according to user preferences; calculating a score of the target product based on the first score and the second score of the target product; sorting the target products according to the scores, and adjusting the flow rate of the target product according to the sorting result.

[0163] In one embodiment, obtaining a target product when the computer program is executed by a processor includes: obtaining a product to be processed; and selecting a target product from the products to be processed.

[0164] In one embodiment, obtaining the target flow rate of a target product when the computer program is executed by a processor includes: determining an initial target flow rate according to the historical flow rate of the target product; obtaining the group to which the target product belongs; determining the position of the target product within the group; and adjusting the initial target flow rate according to the determined position to obtain the target flow rate.

[0165] In one embodiment, obtaining the target flow rate of a target product when the computer program is executed by a processor includes: obtaining the configured flow rate in the configuration data of the target product as the target flow rate.

[0166] In one embodiment, after obtaining the current flow rate of the target product when the computer program is executed by a processor, the following is further included: when the algorithm type is the target algorithm, then continue to perform adjustment processing on the current flow rate according to the target flow rate so that the current flow rate of the target product is adapted to the target flow rate; when the algorithm type is the global algorithm, then perform adjustment processing on the current flow rate according to the target flow rate through the global algorithm.

[0167] In one embodiment, performing adjustment processing on the current flow rate according to the target flow rate through the global algorithm when the computer program is executed by a processor includes: obtaining the target product and the target flow rate, the current flow rate, and the previous regulation parameter of each target product; updating the regulation parameter according to the target flow rate, the current flow rate, and the previous regulation parameter; obtaining the sorting score corresponding to the target product; and performing adjustment processing on the current flow rate according to the updated regulation parameter and the sorting score of the target product so that the current flow rate of the target product is adapted to the target flow rate.

[0168] In one embodiment, an effect evaluation method for traffic control is provided. When the computer program is executed by a processor, the following steps are further implemented: grouping target products to obtain an experimental group and a control group; obtaining the standard observation indexes and standard core indexes of the target products in the experimental group within the first period, and the standard observation indexes and standard core indexes of the target products in the control group; obtaining the experimental observation indexes, experimental core indexes of the target products in the experimental group under the traffic control method of any one of the above embodiments within the second period, and the control observation indexes and control core indexes of the target products in the control group; calculating a first difference between the experimental observation indexes of the experimental group and the standard observation indexes, and a second difference between the control observation indexes of the control group and the standard observation indexes; if the absolute average error between the first difference and the second difference is within a preset range, and the relationship between the experimental core indexes and standard core indexes of the experimental group and the control core indexes and standard core indexes of the control group meets a preset relationship, it is determined that the traffic control algorithm meets the preset requirements.

[0169] In one embodiment, a configuration method for traffic control is provided. When the computer program is executed by a processor, the following steps are further implemented: receiving a target product selection instruction; determining the target product according to the target product selection instruction; receiving a configuration strategy configuration instruction for the target product, and configuring the traffic control algorithm corresponding to the target product according to the configuration strategy configuration instruction. The traffic control algorithm at least includes the traffic control method of any one of the above embodiments.

[0170] In one embodiment, configuring the traffic control algorithm corresponding to the target product according to the configuration strategy configuration instruction when the computer program is executed by a processor includes: configuring the corresponding algorithm type for the traffic control algorithm corresponding to the target product according to the configuration strategy configuration instruction.

[0171] In one embodiment, when the computer program is executed by a processor, the following steps are further implemented: configuring at least one of the execution time, regulation scenario, and target traffic corresponding to the target product.

[0172] In one embodiment, a computer program product is provided, including a computer program. When the computer program is executed by a processor, the following steps are implemented: obtaining the target product and the target traffic of each target product; obtaining the current traffic of the target product; adjusting the current traffic according to the target traffic to make the current traffic of the target product match the target traffic. The adjustment process includes at least one of proportional adjustment, integral adjustment, and / or differential adjustment.

[0173] In one embodiment, when the computer program is executed by a processor, the following steps are further implemented: obtaining adjustment parameters obtained by performing adjustment processing on a target product, where the adjustment parameters include at least one of a proportional adjustment parameter, an integral adjustment parameter, and a derivative adjustment parameter, calculating a first score based on the adjustment parameters; calculating a second score for the target product according to user preferences; calculating a score of the target product based on the first score and the second score of the target product; sorting the target products according to the scores, and adjusting the flow rate of the target product according to the sorting result.

[0174] In one embodiment, the obtaining of the target product implemented when the computer program is executed by a processor includes: obtaining a product to be processed; and selecting a target product from the product to be processed.

[0175] In one embodiment, the obtaining of the target flow rate of the target product implemented when the computer program is executed by a processor includes: determining an initial target flow rate according to the historical flow rate of the target product; obtaining the group to which the target product belongs; determining the position of the target product within the group; and adjusting the initial target flow rate according to the determined position to obtain the target flow rate.

[0176] In one embodiment, the obtaining of the target flow rate of the target product implemented when the computer program is executed by a processor includes: obtaining the configured flow rate in the configuration data of the target product as the target flow rate.

[0177] In one embodiment, after obtaining the current flow rate of the target product implemented when the computer program is executed by a processor, the following is further included: when the algorithm type is the target algorithm, then continue to adjust the current flow rate according to the target flow rate so that the current flow rate of the target product matches the target flow rate; when the algorithm type is the global algorithm, then adjust the current flow rate according to the target flow rate through the global algorithm.

[0178] In one embodiment, the adjusting the current flow rate according to the target flow rate through the global algorithm implemented when the computer program is executed by a processor includes: obtaining the target product and the target flow rate, the current flow rate, and the previous regulation parameter of each target product; updating the regulation parameter according to the target flow rate, the current flow rate, and the previous regulation parameter; obtaining the sorting score corresponding to the target product; and adjusting the current flow rate according to the updated regulation parameter and the sorting score of the target product so that the current flow rate of the target product matches the target flow rate.

[0179] In one embodiment, a method for evaluating the effect of traffic control is provided. When the computer program is executed by a processor, the following steps are further implemented: grouping the target products to obtain an experimental group and a control group; obtaining the standard observation indexes and standard core indexes of the target products in the experimental group within the first period, and the standard observation indexes and standard core indexes of the target products in the control group; obtaining the experimental observation indexes, experimental core indexes of the target products in the experimental group under any one of the traffic control methods in the above embodiments within the second period, and the control observation indexes and control core indexes of the target products in the control group; calculating a first difference between the experimental observation indexes and the standard observation indexes of the experimental group, and a second difference between the control observation indexes and the standard observation indexes of the control group; if the absolute average error between the first difference and the second difference is within a preset range, and the relationship between the experimental core indexes and standard core indexes of the experimental group and the control core indexes and standard core indexes of the control group meets a preset relationship, it is determined that the traffic control algorithm meets the preset requirements.

[0180] In one embodiment, a method for configuring traffic control is provided. When the computer program is executed by a processor, the following steps are further implemented: receiving a target product selection instruction; determining the target product according to the target product selection instruction; receiving a configuration policy configuration instruction for the target product, and configuring the traffic control algorithm corresponding to the target product according to the configuration policy configuration instruction, where the traffic control algorithm is any one of the traffic control methods in the above embodiments.

[0181] In one embodiment, configuring the traffic control algorithm corresponding to the target product according to the configuration policy configuration instruction when the computer program is executed by a processor includes: configuring the corresponding algorithm type for the traffic control algorithm corresponding to the target product according to the configuration policy configuration instruction.

[0182] In one embodiment, when the computer program is executed by a processor, the following steps are further implemented: configuring at least one of the execution time, regulation scenario, and target traffic corresponding to the target product.

[0183] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data for analysis, stored data, displayed data, etc.) involved in this application are all information and data that have been authorized by the user or fully authorized by all parties.

[0184] Those of ordinary skill in the art can understand that all or part of the processes in the methods of the above embodiments can be completed by instructing relevant hardware through a computer program. The computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above methods. Among them, any reference to a memory, database, or other medium used in the embodiments provided in the present application can include at least one of non-volatile and volatile memories. Non-volatile memory can include Read-Only Memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetoresistive random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. By way of illustration and not limitation, RAM can be in various forms, such as static random access memory (SRAM) or dynamic random access memory (DRAM), etc. The databases involved in the embodiments provided in the present application can include at least one of relational databases and non-relational databases. Non-relational databases can include distributed databases based on blockchain, etc., without limitation. The processors involved in the embodiments provided in the present application can be general-purpose processors, central processing units, graphics processing units, digital signal processors, programmable logic devices, data processing logics based on quantum computing, etc., without limitation.

[0185] The technical features of the above embodiments can be combined arbitrarily. For the sake of brevity of description, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, it should be considered as the scope described in this specification.

[0186] The above-described embodiments only represent several implementation manners of the present application. The description is relatively specific and detailed, but it should not be construed as a limitation on the patent scope of the present application. It should be noted that for those of ordinary skill in the art, without departing from the concept of the present application, several modifications and improvements can still be made, and these all belong to the protection scope of the present application. Therefore, the protection scope of the present application should be subject to the appended claims.

Claims

1. A flow control method, characterized in that, The method includes: Obtaining the target product and the target traffic of each target product; the target traffic is the exposure volume to be achieved; Obtaining the current traffic of the target product; the current traffic is the exposure volume at the current moment; the exposure volume refers to the number of people in the live streamer's live room; Adjusting the current traffic according to the target traffic so that the current traffic of the target product is adapted to the target traffic, and the adjustment process includes at least one of proportional adjustment, integral adjustment, and / or derivative adjustment; The method further includes: Obtaining the adjustment parameters obtained by adjusting the target product, where the adjustment parameters include at least one of a proportional adjustment parameter, an integral adjustment parameter, and a derivative adjustment parameter, and calculating the first parameter of the first score through the error in the adjustment term of proportional adjustment and the corresponding proportional adjustment parameter, the sum of the differences in the adjustment term of integral adjustment, and the result of the first derivative in the adjustment term of derivative adjustment and the corresponding derivative adjustment parameter; Among them, the calculation formula of the first parameter: P1 = Kp * PTerm + Ip * ITerm + Dp * DTerm, where P1 represents the first parameter; Kp represents the proportional adjustment parameter; PTerm represents the error in the adjustment term of proportional adjustment; Ip represents the integral adjustment parameter; ITerm represents the sum of the differences in the adjustment term of integral adjustment; Dp represents the derivative adjustment parameter; DTerm represents the result of the first derivative in the adjustment term of derivative adjustment; Obtaining the time T1 when the traffic of the target product is adjusted for the first time, and the time T2 when the traffic of the target product is adjusted currently; Based on the time T1 when the traffic of the target product is adjusted for the first time, the time T2 when the traffic of the target product is adjusted currently, and the first parameter, substituting into P = P1 * EXP(-(300 - (T2 - T1)) / K) to obtain the first score; Among them, P is the first score; P1 is the first parameter; EXP(-(300 - (T2 - T1)) / K) is the attenuation processing of the current timestamp - the first exposure adjustment timestamp within 5 minutes, and K is the attenuation adjustment parameter; Calculating a second score for the target product according to user preferences; Calculating the score of the target product according to the first score and the second score of the target product; Sorting the target products according to the scores, and adjusting the traffic of the target products according to the sorting results.

2. The method according to claim 1, wherein The obtaining of the target product includes: Obtaining the product to be processed; Selecting the target product from the products to be processed.

3. The method according to claim 1, characterized in that The obtaining of the target traffic of the target product includes: Determining the initial target traffic according to the historical traffic of the target product; Obtaining the group to which the target product belongs; Determining the position of the target product within the group; Adjusting the initial target traffic according to the determined position to obtain the target traffic.

4. The method according to claim 1, characterized in that The obtaining of the target traffic of the target product includes: Obtaining the configured traffic in the configuration data of the target product as the target traffic.

5. The method according to claim 1, wherein After obtaining the current traffic of the target product, it further includes: When the algorithm type is the target algorithm, continue to adjust the current traffic according to the target traffic so that the current traffic of the target product is adapted to the target traffic; When the algorithm type is the global algorithm, adjust the current traffic according to the target traffic through the global algorithm.

6. The method according to claim 5, characterized in that The adjusting the current traffic according to the target traffic through the global algorithm includes: Obtain the target product and the target traffic, current traffic, and previous regulation parameter of each target product; Update the regulation parameter according to the target traffic, current traffic, and the previous regulation parameter; Obtain the sorting score corresponding to the target product; Adjust the current traffic according to the updated regulation parameter of the target product and the sorting score so that the current traffic of the target product is adapted to the target traffic.

7. A method for evaluating the effect of flow control, characterized in that The method includes: Group the target products to obtain an experimental group and a control group; Obtain the standard observation index and standard core index of the target products in the experimental group within the first period, and the standard observation index and standard core index of the target products in the control group; Obtain the experimental observation index, experimental core index of the target products in the experimental group under the traffic control method described in any one of claims 1 to 6 within the second period, and the control observation index and control core index of the target products in the control group; Calculate the first difference between the experimental observation index of the experimental group and the standard observation index, and the second difference between the control observation index of the control group and the standard observation index; If the absolute average error between the first difference and the second difference is within a preset range, and the relationship between the experimental core index and standard core index of the experimental group and the control core index and standard core index of the control group satisfies a preset relationship, determine that the traffic control algorithm meets the preset requirements.

8. A configuration method for flow control, characterized in that, The method includes: Receive a target product selection instruction; Determine the target product according to the target product selection instruction; Receive a configuration strategy configuration instruction for the target product, and configure the traffic control algorithm corresponding to the target product according to the configuration strategy configuration instruction. The traffic control algorithm at least includes the traffic control method described in any one of claims 1 to 6.

9. The method according to claim 8, wherein The configuring the traffic control algorithm corresponding to the target product according to the configuration strategy configuration instruction includes: Configure the corresponding algorithm type for the traffic control algorithm corresponding to the target product according to the configuration strategy configuration instruction.

10. The method according to claim 8, wherein The method further includes: Configure at least one of the execution time, regulation scenario, and target traffic corresponding to the target product.

11. A flow control device, characterized in that, The device includes: An acquisition module, configured to acquire the target product and the target traffic of each target product; the target traffic is the exposure amount to be reached; A current traffic acquisition module, configured to acquire the current traffic of the target product; the current traffic is the exposure amount at the current moment; the exposure amount refers to the number of people in the host's live broadcast room; A flow regulation module, which is used to adjust the current flow according to the target flow so that the current flow of the target product is adapted to the target flow, and the adjustment process includes at least one of proportional adjustment, integral adjustment, and / or derivative adjustment; A first score calculation module, which is used to obtain the adjustment parameters obtained by adjusting the target product. The adjustment parameters include at least one of a proportional adjustment parameter, an integral adjustment parameter, and a derivative adjustment parameter, and calculate the first parameter of the first score through the error in the adjustment term of proportional adjustment and the corresponding proportional adjustment parameter, the sum of the differences in the adjustment term of integral adjustment and the corresponding integral adjustment parameter, and the result of the first derivative in the adjustment term of derivative adjustment and the corresponding derivative adjustment parameter; Among them, the calculation formula of the first parameter: P1 = Kp * PTerm + Ip * ITerm + Dp * DTerm, where P1 represents the first parameter; Kp represents the proportional adjustment parameter; PTerm represents the error in the adjustment term of proportional adjustment; Ip represents the integral adjustment parameter; ITerm represents the sum of the differences in the adjustment term of integral adjustment; Dp represents the derivative adjustment parameter; DTerm represents the result of the first derivative in the adjustment term of derivative adjustment; Obtain the time T1 when the flow of the target product is adjusted for the first time, and the time T2 when the flow of the target product is adjusted currently; based on the time T1 when the flow of the target product is adjusted for the first time, the time T2 when the flow of the target product is adjusted currently, and the first parameter, substitute them into P = P1 * EXP(-(300 - (T2 - T1)) / K) to obtain the first score; where P is the first score; P1 is the first parameter; EXP(-(300 - (T2 - T1)) / K) is the decay process of the current timestamp - the first exposure adjustment timestamp within 5 minutes, and K is the decay adjustment parameter; A second score calculation module, which is used to calculate the second score for the target product according to user preferences; a score calculation module, which is used to calculate the score of the target product according to the first score and the second score of the target product; a sorting adjustment module, which is used to sort the target products according to the scores and adjust the flow of the target products according to the sorting results.

12. An apparatus for evaluating the effect of flow control, characterized in that, The method includes: A grouping module, which is used to group the target products to obtain an experimental group and a control group; A standard index acquisition module, which is used to acquire the standard observation index and the standard core index of the target products in the experimental group during the first period, and the standard observation index and the standard core index of the target products in the control group; An experimental control index acquisition module, which is used to, in the second period, the flow control algorithm of the experimental group at least includes the flow control method described in any one of claims 1 to 6, and acquire the experimental observation index and the experimental core index of the target products in the experimental group during the second period, and the control observation index and the control core index of the target products in the control group; A first calculation module, configured to calculate a first difference between the experimental observation indexes of the experimental group and the standard observation indexes, and a second difference between the control observation indexes of the control group and the standard observation indexes; A determination module, configured to determine that the flow control algorithm meets the preset requirements if the absolute average error between the first difference and the second difference is within a preset range, and the relationships between the experimental core indexes and the standard core indexes of the experimental group and the control core indexes and the standard core indexes of the control group meet the preset relationships.

13. A configuration device for flow control, characterized in that, The device includes: A receiving module, configured to receive a target product selection instruction; A target product determination module, configured to determine a target product according to the target product selection instruction; A configuration module, configured to receive a configuration strategy configuration instruction for the target product, and configure the flow control algorithm corresponding to the target product according to the configuration strategy configuration instruction, where the flow control algorithm at least includes the flow control method described in any one of claims 1 to 6.

14. A computer device, comprising a memory and a processor, the memory storing a computer program, characterized in that, When the processor executes the computer program, the steps of the method described in any one of claims 1 to 6, 7 or 8 are implemented.

15. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, the steps of the method described in any one of claims 1 to 6, 7 or 8 are implemented.

16. A computer program product, comprising a computer program, characterized in that, When the computer program is executed by the processor, the steps of the method described in any one of claims 1 to 6, 7 or 8 are implemented.

Images