Traffic Control Methods

Abstract

The embodiment of the present invention discloses a flow control method. The flow control method includes: obtaining the first accumulated actual flow of a specific material among multiple materials in the current period of the target date up to the current moment; obtaining the estimated flow in the current period of the target date; when the first accumulated actual flow is less than the estimated flow, obtaining the actual expectation of the flow of the specific material at the current moment; obtaining the estimated expectation of the flow of the specific material at the current moment in the current period of the target date; and when the actual expectation of the flow of the specific material is less than the estimated expectation of the flow of the specific material, based on the difference between the estimated expectation and the actual expectation of the flow of the specific material, adjusting the position of the material to be controlled in the specific material. The present invention accurately meets the multi-material flow control needs in the mixed recommendation scenario.

Description

Technical Field

[0001] The embodiment of the present invention specifically relates to a flow control method. Background Art

[0002] In some comprehensive recommendation scenarios for different material types, it is necessary to recommend and distribute unused materials. The functions of different material types are sometimes different. In order to meet some business needs, some special materials need to be supported with a certain degree of traffic, while at the same time minimizing the loss of the optimization indicators of the scenario (such as click-through rate / number of clicks). At present, traffic control is mainly solved by control theory and constrained optimization modeling methods. The demand for traffic control is added to the constraints, and the optimal value of the overall traffic distribution is solved through linear programming. When performing traffic control online, PID (proportional-integral-differential controller) is used for online traffic control. Existing solutions are unable to model and adjust different dimensions of different materials. Summary of the invention

[0003] The purpose of the present invention is to provide a flow control method which can accurately meet the flow control requirements of multiple materials in a mixed recommendation scenario.

[0004] In order to achieve the above-mentioned purpose, the present invention provides a flow control method on the one hand, which includes: obtaining a first cumulative actual flow of a specific material among multiple materials in the current time period of a target date up to the current moment; obtaining an estimated flow in the current time period of the target date based on the target flow of the specific material and the flow of the specific material in the current time period of multiple preset dates before the target date; in a case where the first cumulative actual flow is less than the estimated flow, obtaining the actual expected flow of the specific material at the current moment in the current time period of the target date; based on the target flow of the specific material and the flow of the specific material in each time period of the multiple preset dates, obtaining the estimated expected flow of the specific material at the current moment in the current time period of the target date; and in a case where the actual expected flow of the specific material is less than the estimated expected flow of the specific material, adjusting the position of the material to be controlled among the specific materials based on the difference between the estimated expected flow and the actual expected flow of the specific material.

[0005] In at least one embodiment, obtaining the estimated flow rate in the current time period in the target date includes: determining the estimated flow rate of the specific material in the current time period based on the target flow rate of the specific material in the target date and the flow rate of the specific material in the current time period in the multiple preset dates, wherein the target flow rate of the specific material is associated with the flow rate of the specific material in the multiple preset dates; updating the estimated flow rate in the current time period based on the expected flow rate in the current time period determined at the previous moment before the current moment and the first accumulated actual flow rate, so as to obtain the estimated flow rate in the current time period corresponding to the current moment.

[0006] In at least one embodiment, obtaining the actual expectation of the flow of the specific material at the current moment in the current time period in the target date includes: obtaining the expectation of the flow of the specific material based on the flow of the specific material at each display location at the current moment and the average value of the flow of the multiple materials at each display location in each preset date of the multiple preset dates; and determining the actual expectation of the flow of the specific material based on the user's preference for the specific material and the expectation of the flow of the specific material.

[0007] In at least one embodiment, obtaining the estimated expectation of the flow of the specific material at the current moment in the current time period in the target date includes: obtaining the estimated probability of the flow of the specific material in the current time period in the target date based on the target flow of the specific material and the flow of the specific material in the current time period in the multiple preset dates; obtaining the second cumulative actual flow of the specific material in the target date up to the current moment in the current time period; obtaining the second cumulative estimated flow of the specific material in the target date up to the current moment in the current time period; and determining the estimated expectation of the flow of the specific material at the current moment in the current time period based on the estimated probability of the flow of the specific material, the second cumulative actual flow, the second cumulative estimated flow and the global preference of the specific material.

[0008] In at least one embodiment, obtaining the second cumulative estimated flow rate of the specific material in the target date up to the current moment in the current period includes: determining the estimated flow rate of the specific material in each period based on the target flow rate of the specific material in the target date and the flow rate of the specific material in each period in the multiple preset dates before the target date, wherein the target flow rate of the specific material is associated with the flow rate of the specific material in the multiple preset dates; updating the estimated flow rate of the specific material in each period before the current period and the current period based on the third cumulative actual flow rate and the third cumulative estimated flow rate of the specific material in each period before the current period up to each moment and the fourth cumulative actual flow rate and the fourth cumulative estimated flow rate of the specific material in the current period up to each moment before the current moment, so as to obtain the updated estimated flow rate of the specific material in each period before the current period and the current period; and determining the second cumulative estimated flow rate of the specific material in the target date up to the current moment in the current period based on the updated estimated flow rate of the specific material in each period before the current period and the current period.

[0009] In at least one embodiment, determining the estimated expectation of the flow of the specific material at the current moment in the current time period includes: determining the regulation expectation of the flow of the specific material at the current moment in the current time period based on the estimated probability of the flow of the specific material, the second cumulative actual flow and the second cumulative estimated flow; and determining the estimated expectation of the specific material at the current moment in the current time period based on the global preference of the specific material and the regulation expectation of the flow of the specific material at the current moment in the current time period.

[0010] In at least one embodiment, the specific material includes at least one of the following: a new house, a second-hand house, a rental house, and a live broadcast.

[0011] In at least one embodiment, the position adjustment of the material to be regulated among the specific materials includes: determining the material to be regulated based on the difference between the estimated expectation and the actual expectation of the flow rate of the specific material; and adjusting the display position of the material to be regulated.

[0012] In at least one embodiment, the position adjustment of the material to be regulated among the specific materials also includes: determining the expected flow of the multiple materials after adjustment and counting the number of adjustments; determining the expected loss of the flow of the multiple materials based on the actual expectation of the flow of each material among the multiple materials and the expected flow of the multiple materials after adjustment; and continuing to adjust the position of the material to be regulated among the specific materials when the counted number of adjustments is less than the preset number and the expected loss of the flow of the multiple materials is less than the loss threshold.

[0013] Through the above technical scheme, the present invention creatively obtains the actual expectation of the flow rate of the specific material at the current moment in the current period when the first cumulative actual flow rate of a specific material among multiple materials in the current period of the target date up to the current moment is less than the estimated flow rate in the current period; then, based on the target flow rate of the specific material and the flow rate of the specific material in each period of the multiple preset dates, obtains the estimated expectation of the flow rate of the specific material at the current moment in the current period of the target date; then, when the actual expectation of the flow rate of the specific material is less than the estimated expectation of the flow rate of the specific material, based on the difference between the estimated expectation and the actual expectation of the flow rate of the specific material, adjusts the position of the material to be regulated among the specific materials. Therefore, the present invention can accurately meet the needs of multi-material flow control in mixed recommendation scenarios.

[0014] A second aspect of the present invention provides a computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, the above-mentioned flow control method is implemented.

[0015] A third aspect of the present invention provides an electronic device, comprising: a processor; a memory for storing executable instructions of the processor; the processor is used to read the executable instructions from the memory and execute the instructions to implement the above-mentioned traffic control method.

[0016] A fourth aspect of the present invention provides a computer program product, the computer program product comprising a computer program, and the computer program implements the above-mentioned traffic control method when executed by a processor.

[0017] Other features and advantages of the present invention will be described in detail in the following detailed description. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] The accompanying drawings are used to provide a further understanding of the embodiments of the present invention and constitute a part of the specification. Together with the following specific implementations, they are used to explain the embodiments of the present invention, but do not constitute a limitation on the embodiments of the present invention. In the accompanying drawings:

[0019] Figure 1 is a flow chart of a flow control method provided by an embodiment of the present invention;

[0020] Figure 2 An embodiment of the present invention provides obtaining an estimated expectation of the flow rate of each material at the current time within the current period of the target date; and

[0021] Figure 3 is a flow chart of flow control provided by an embodiment of the present invention; and

[0022] Figure 4 This is a schematic diagram of data involved in traffic control provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0023] The specific implementation of the present invention is described in detail below in conjunction with the accompanying drawings. It should be understood that the specific implementation described here is only used to illustrate and explain the present invention, and is not used to limit the present invention.

[0024] Figure 1 FIG. 1 is a flow chart of a flow control method provided by an embodiment of the present invention. Figure 1 As shown, the flow control method may include the following steps S101-S105.

[0025] Step S101, obtaining a first accumulated actual flow rate of a specific material among a plurality of materials within a current period of a target date up to the current moment.

[0026] The multiple materials may include new houses, second-hand houses, rental houses, and live broadcasts. Accordingly, the specific materials may be reasonably set according to actual needs. For example, the specific materials include at least one of the following: new houses, second-hand houses, rental houses, and live broadcasts. In the case where the specific materials are new houses and live broadcasts, the traffic corresponding to the new houses may be the number of clicks on the new houses; and the traffic corresponding to the live broadcasts may be the number of exposures of the live broadcasts.

[0027] For new houses, a real-time data list based on system feedback (such as Figure 4 The real-time data of new houses and the real-time data of live broadcasts shown in the figure add up the clicks on new houses at each moment in the current time period up to the current moment; for live broadcasts, based on the real-time data list fed back by the system, add up the exposures of the live broadcasts at each moment in the current time period up to the current moment.

[0028] Step S102, based on the target flow of the specific material and the flow of the specific material in the current period of multiple preset dates before the target date, obtain the estimated flow in the current period of the target date.

[0029] The estimation of regulated materials refers to estimating the traffic that needs to be maintained on the target date (e.g., T+1 day) of the material according to the regulated materials and the regulated dimensions, the traffic at the hourly level (i.e., time period) (and then estimating the traffic at each moment. Unless otherwise specified, the hourly traffic in each embodiment of the present invention is evenly distributed at each moment), etc. Taking the example of ensuring the number of clicks on new houses and the number of exposures on live broadcasts, the process of determining the estimated number of clicks on new houses and the estimated number of exposures on live broadcasts in each corresponding time period is described respectively.

[0030] Specifically, for new houses, firstly, according to the estimated click volume of new houses within n+1 days before the target date (i.e., day T+1), the estimated click volume of new houses on the target date (i.e., the target click volume of new houses, such as Figure 4 The new house click estimation data shown in the figure) where k1 is a value greater than 1 (for example, k1=1.03, which can be reasonably set according to actual needs):

[0031]

[0032] Next, the number of clicks on the target date (i.e., day T+1) is distributed to each time period according to the following formula (1), thereby obtaining the estimated number of clicks on the new house in the current time period j on the target date (i.e., day T+1):

[0033]

[0034] For live broadcast, firstly, according to the live broadcast exposure in n+1 days before the target date (i.e., day T+1), the estimated exposure of the live broadcast on the target date (i.e., the target exposure of the live broadcast, such as Figure 4 The live broadcast exposure estimation data shown in FIG. 1 ), where k2 is a value greater than 1 (for example, k2=1.03, which can be reasonably set according to actual needs):

[0035]

[0036] Then, the exposure of the live broadcast on the target date (i.e., day T+1) is distributed to each time period according to the following formula (2), thereby obtaining the estimated exposure of the live broadcast in the current time period j on the target date (i.e., day T+1):

[0037]

[0038] The estimated click volume of the new house in period j and the estimated exposure volume of the live broadcast in period j are determined using the above formulas (1) and (2) respectively.

[0039] In another embodiment, for the flow estimation on the target date (i.e., day T+1), although the flow estimation is performed on each material the day before the target date, the flow estimation may have a relatively large deviation due to objective and uncontrollable factors. If the estimated value that is significantly different from the actual flow is used for flow control, it is difficult to ensure the control target and efficiency. Based on this, it is necessary to correct the estimated value online in real time according to the actual flow.

[0040] Among them, obtaining the estimated flow rate in the current time period in the target date may include: determining the estimated flow rate of the specific material in the current time period based on the target flow rate of the specific material in the target date and the flow rate of the specific material in the current time period in the multiple preset dates, wherein the target flow rate of the specific material is associated with the flow rate of the specific material in the multiple preset dates; and updating the estimated flow rate in the current time period based on the expected flow rate in the current time period determined at the previous moment of the current moment and the first accumulated actual flow rate to obtain the estimated flow rate in the current time period corresponding to the current moment.

[0041] Taking the new house material as an example, according to the accumulated actual click volume a(t) of the new house in the current period up to the current moment and the estimated traffic c(t-1) in the current period determined at the previous moment, the estimated click volume in the current period is corrected to obtain the updated estimated click volume c(t) of the new house in the current period corresponding to the current moment (e.g. Figure 4 (see revised estimate data for the table below).

[0042] Step 1: Monitor the current time t. If the accumulated actual click volume a(t) ≥ the estimated click volume c(t-1), proceed to step 2 (no need to perform traffic control); otherwise, proceed to step 3.

[0043] Step 2: Update the estimated click volume of new houses in the current period based on the following formula to obtain the updated estimated click volume c(t);

[0044]

[0045] Step 3: Execute the following steps S103-S105 to control the user's traffic flow.

[0046] Specifically, for the new house material, if the current time t in the current time period of the target date is 6:00 on the T+1 day, the estimated number of clicks on the new house in the time period of 6:00-7:00 is determined by the above formula (1) as c(0).

[0047] If the current time t within the current period of the target date is 6:01 on the (T + 1)-th day, count the actual click volume a(1) of new house clicks between 6:00 and 6:01 on the (T + 1)-th day; determine the estimated click volume c(0) within this period at 6:01 on the (T + 1)-th day. If a(1) ≥ c(0), then

[0048] If the current time t is updated to 6:02 on the (T + 1)-th day, count the actual click volume a(2) of new house clicks between 6:00 and 6:02 on the (T + 1)-th day; determine the estimated click volume c(1) within this period at 6:02 on the (T + 1)-th day. If a(2) ≥ c(1), then

[0049] If the current time t is updated to 6:03 on the (T + 1)-th day, count the actual click volume a(3) of new house clicks between 6:00 and 6:03 on the (T + 1)-th day; if a(3) < c(2), then perform traffic regulation for the user through the following steps S103 - S105.

[0050] As t is updated, continuously update the estimated traffic within the current period through a similar process as above, and take the estimated traffic within the current period of the last update as the estimated traffic within this current period.

[0051] For the material of live broadcast, the estimated exposure volume of the live broadcast within the current period can be determined in a similar manner as above.

[0052] Step S103, in the case where the first cumulative actual traffic is less than the first cumulative estimated traffic, obtain the actual expectation of the traffic of the specific material at the current time within the current period of the target date.

[0053] For step S103, the obtaining of the actual expectation of the traffic of the specific material at the current time within the current period of the target date may include: based on the traffic of the specific material at each display position at the current time and the average value of the traffic of the multiple materials at each display position within each preset date among the multiple preset dates, obtain the expectation of the traffic of the specific material; and based on the user's preference for the specific material and the expectation of the traffic of the specific material, determine the actual expectation of the traffic of the specific material.

[0054] Specifically, the following takes the new house and live broadcast materials as examples for illustration respectively. Based on the result list returned by the comprehensive scenario of multi-material traffic regulation, calculate the actual expectation of the traffic of each material (such as Figure 4 the returned result expectation data shown).

[0055] For the new house material, first, based on the exposure of the new house at display position i and the overall exposure at display position i (if the target date is the T+1 day, the average daily exposure at display position i in the previous n+1 days), the exposure probability of the new house at display position i is determined according to the following formula: expose i :

[0056]

[0057] Next, according to the number of clicks on new homes at placement i and the total number of clicks at placement i (if the target date is day T+1, then the average number of clicks per day at placement i in the previous n+1 days), the click probability of new homes at placement i is determined according to the following formula: i :

[0058]

[0059] Furthermore, according to the exposure probability of the new house at the display position i determined above, i and the probability of clicking on a new house i , the expected number of clicks on the new house is determined according to the following formula (i.e., the product of the exposure probability of the new house at each display position i and the click probability of the new house is summed up):

[0060]

[0061] Finally, the user's preference for new houses can be determined comprehensively through the user's historical data and preference model, such as Figure 4 The personalized business line preference data shown (which can be determined according to existing methods) If the user is a new user, the global business line preference can be used as the user's business line preference.

[0062] Based on the product of the user's preference for the new house determined above and the expected number of clicks on the new house determined above, the actual expected number of clicks on the new house is determined. In this embodiment, if the user prefers the new house, the regulation range of the new house's listing and the new house's live broadcast is increased accordingly; if the preference is weak, the regulation range of the corresponding business line is reduced. In order to achieve personalized regulation, it is necessary to estimate the user's business preferences, and the user's preference estimate data for the specific material is shown in the above formula (it represents the user's preference for a specific material). Therefore, the introduction of user business line preferences is to ensure that different users have different regulation ranges to improve the utilization efficiency of traffic.

[0063] For the live broadcast material, first, according to the exposure probability of the new house at the display position j, expose j , the expected exposure of the live broadcast is determined according to the following formula (i.e., the exposure probability of the live broadcast at each display position j is summed):

[0064]

[0065] Then, the user's preference for live broadcast can be comprehensively determined through the user's historical data and preference model.

[0066] Based on the product of the user's preference for live broadcast determined above and the expected exposure rate of live broadcast determined above, the actual expected exposure rate of live broadcast is determined. This embodiment improves the utilization efficiency of traffic, implements a personalized traffic regulation strategy, and increases the traffic distribution strength for users with regulated material preferences by using business line preference estimation data, thereby improving the utilization efficiency of traffic on the basis of completing regulation.

[0067] Step S104, obtaining the estimated expected flow rate of each material at the current moment in the current period of the target date.

[0068] For step S104, the step of obtaining the estimated expected flow rate of each material at the current time within the current period of the target date may include the following steps S201-S204, such as Figure 2 shown.

[0069] Step S201, based on the target flow of the specific material and the flow of the specific material in the current period of the plurality of preset dates, obtain an estimated probability of the flow of the specific material in the current period of the target date.

[0070] Specifically, for new houses, firstly, according to the estimated exposure of new houses within n+1 days before the target date (i.e., day T+1), the estimated exposure of new houses on the target date is determined according to the following formula:

[0071]

[0072] Next, the estimated exposure of new houses on the target date (i.e., day T+1) is allocated to each time period according to the following formula (5), thereby obtaining the estimated exposure of new houses in the current time period j on the target date (i.e., day T+1):

[0073]

[0074] Based on the estimated exposure volume of the new house in the current period j in the target date (i.e., the T+1 day) and the estimated click volume of the new house (i.e., the formula (1)), the estimated click probability of the new house in the current period j in the target date (i.e., the T+1 day) is determined according to the following formula:

[0075]

[0076] For live broadcast, the estimated exposure request volume of the live broadcast on the target date is determined according to the following formula based on the estimated exposure request volume of the live broadcast within n+1 days before the target date (i.e., day T+1):

[0077]

[0078] Next, the exposure request volume of the live broadcast on the target date (i.e., day T+1) is distributed to each time period according to the following formula, thereby obtaining the estimated exposure request volume of the live broadcast in the current time period j on the target date (i.e., day T+1):

[0079]

[0080] According to the estimated exposure request volume (see formula (7)) and the estimated exposure volume (see formula (2)) of the live broadcast in the current time period j in the target date (i.e., the T+1 day) determined above, the estimated exposure probability of the live broadcast in the current time period j in the target date (i.e., the T+1 day) is determined according to the following formula (8):

[0081]

[0082] Therefore, for new homes and live broadcasts, it is necessary to estimate the number of clicks on new homes and the amount of exposure of live broadcasts that need to be achieved by traffic control on the target date (i.e., day T+1), and distribute the amount that needs to be maintained in a day to the amount that needs to be maintained in each hourly period.

[0083] Step S202, obtaining a second accumulated actual flow of the specific material from the target date to the current moment in the current time period.

[0084] If the current time in the current time period of the target date is 6:20 on the T+1 day, for the new house material, the actual clicks on the new house from 0:00 to 6:20 on the T+1 day are counted; for the live broadcast material, the actual exposure of the live broadcast from 0:00 to 6:20 on the T+1 day is counted.

[0085] In order to achieve the control requirements more accurately, it is necessary to dynamically adjust the control range of the estimated expected clicks on new houses or live broadcast exposures during the entire control process. When the actual flow (i.e., the second cumulative actual flow) and the target flow (i.e., the second cumulative estimated flow) are far apart, the control range of the estimated click probability of new houses in time period j is increased; when the actual flow (i.e., the second cumulative actual flow) and the target flow (i.e., the second cumulative estimated flow) are close, the control range is reduced. The following introduces PID control (proportional-integral-differential controller) to adjust the control range in real time.

[0086] Step S203, obtaining a second cumulative estimated flow rate of the specific material from the target date to the current moment in the current period.

[0087] For step S203, the obtaining of the second cumulative estimated flow rate of the specific material from the target date to the current moment in the current period includes: determining the estimated flow rate of the specific material in each period based on the target flow rate of the specific material on the target date and the flow rate of the specific material in each period of the multiple preset dates before the target date, wherein the target flow rate of the specific material is associated with the flow rate of the specific material in the multiple preset dates; determining the estimated flow rate of the specific material in each period based on the third cumulative actual flow rate of the specific material in each period before the current period and the third cumulative actual flow rate of the specific material in each moment. The fourth cumulative actual flow and the fourth cumulative estimated flow of the specific material in the current time period up to the current moment are calculated, and the estimated flow of the specific material in each time period before the current time period and the current time period is updated to obtain the updated estimated flow of the specific material in each time period before the current time period and the current time period; and based on the updated estimated flow of the specific material in each time period before the current time period and the current time period, the second cumulative estimated flow of the specific material in the target date up to the current moment in the current time period is determined.

[0088] Specifically, first, for a new house, the estimated number of clicks on the new house in time period j can be determined by the above formula (1).

[0089] For the new house material, if the current time t in the current time period of the target date is 6:20 on the T+1 day, for each time period in the 0:00-6:00 time period, the estimated click volume of the new house in each time period can be updated in real time in combination with the accumulated actual click volume in each time period up to each time point. The estimated click volume of the last update in each time period is used as the estimated click volume in each time period. Finally, the estimated click volume of each time period in the 0:00-6:00 time period is accumulated to obtain the accumulated estimated click volume in the 0:00-6:00 time period.

[0090] For the 20 minutes in the time period of 6:00-7:00, the estimated click volume of the last update in the first 20 minutes of the time period is used as the estimated click volume in the time period. Then the estimated click volume in the time period is multiplied by 1 / 3 (the proportion of the current time in the current time period) to obtain the estimated click volume in 6:00-6:20. Finally, the cumulative estimated click volume in the time period of 0:00-6:00 is accumulated with the estimated click volume in 6:00-6:20 to obtain the cumulative estimated click volume of new houses between 0:00-6:20.

[0091] For live broadcast, the estimated exposure of the live broadcast in time period j can be determined by the above formula (2). And the cumulative estimated exposure of the live broadcast between 0:00-6:20 can be determined according to the above similar process.

[0092] Step S204, determining the expected estimated flow rate of the specific material at the current moment in the current time period based on the estimated probability of the flow rate of the specific material, the second accumulated actual flow rate, the second accumulated estimated flow rate and the global preference of the specific material.

[0093] For step S204, determining the estimated expectation of the flow of the specific material at the current moment in the current time period may include: determining the regulation expectation of the flow of the specific material at the current moment in the current time period based on the estimated probability of the flow of the specific material, the second cumulative actual flow and the second cumulative estimated flow; and determining the estimated expectation of the specific material at the current moment in the current time period based on the global preference of the specific material and the regulation expectation of the flow of the specific material at the current moment in the current time period.

[0094] For the new house material, based on the estimated click probability of the new house in period j determined by the above formula (6), the second cumulative actual click volume and the second cumulative estimated click volume, the new house click PID expectation at the current moment in period j is determined according to the following formula (9):

[0095]

[0096] For the live broadcast material, based on the estimated exposure probability of the live broadcast in period j determined by the above formula (8), the second accumulated actual exposure amount and the second accumulated estimated exposure amount, the live broadcast exposure PID expectation at the current moment in period j is determined according to the following formula (10):

[0097]

[0098] According to the above PID control formula (9) or (10), it can be concluded that if the difference between the second cumulative actual click volume (or exposure volume) and the second cumulative estimated click volume (or exposure volume) is large (for example, or ), the expected output will become larger, and the control range (i.e., or Similarly, if the second cumulative actual click volume (or exposure) is close to the second cumulative estimated click volume (or exposure), the adjustment range will be reduced.

[0099] Wherein, the global preference of the specific material (such as Figure 4 The global business line preference data shown) can be determined by the following formula:

[0100]

[0101] The average number of clicks on the new house indicates the average number of clicks on the new house on each of the multiple preset dates; and the global average number of clicks indicates the average number of clicks on all houses on each date.

[0102]

[0103] The average click volume of the live broadcast represents the average exposure of the live broadcast on each of the multiple preset dates; and the global average click volume represents the average click volume of all listings on each date.

[0104] Finally, the product of the global preference of the new house and the expected live exposure PID at the current moment in the j period is taken as the estimated expectation of the specific material at the current moment in the j period (e.g. Figure 4 Therefore, this embodiment uses a PID controller to adjust the control amplitude in real time, and designs an hourly target correction strategy, so that the control target can be achieved when the flow fluctuates.

[0105] Step S105, when the actual expected flow rate of a specific material among the multiple materials is less than the estimated expected flow rate of the specific material, based on the difference between the estimated expected flow rate of the specific material and the actual expected flow rate, adjust the position of the material to be regulated among the specific materials.

[0106] For step S105, the position adjustment of the material to be regulated among the specific materials may include: determining the material to be regulated based on the difference between the estimated expectation and the actual expectation of the flow rate of the specific material; and adjusting the display position of the material to be regulated.

[0107] Specifically, the difference between the estimated expectation and the actual expectation of each material is calculated, and the material corresponding to the maximum difference is found; the display position of the material is adjusted. If the material corresponding to the maximum difference is a new house, the front display position of the new house is adjusted to the Nth display position (for example, the 4th position).

[0108] Taking the real estate sector as an example, the existing real estate display positions are determined according to user preferences, which is reasonable to a certain extent. However, user preferences for different materials are determined according to different models, so the display positions between different materials cannot accurately represent user preferences. Therefore, after adjusting the display position of a specific material, a certain loss will be introduced.

[0109] For step S105, the position adjustment of the materials to be regulated among the specific materials may also include: determining the expected flow of the multiple materials after adjustment and counting the number of adjustments; determining the expected loss of the flow of the multiple materials based on the actual expectation of the flow of each material among the multiple materials and the expected flow of the multiple materials after adjustment; and continuing to adjust the position of the materials to be regulated among the specific materials when the counted number of adjustments is less than the preset number and the expected loss of the flow of the multiple materials is less than the loss threshold.

[0110] Specifically, according to the above (3)-(4), the expected flow rate of the multiple materials after adjustment can be determined, and at the same time, the number of adjustments is counted; the actual expectation of the flow rate of each material is summed to obtain the overall expectation, that is, the expected flow rate of the multiple materials after adjustment is subtracted from the overall expectation to obtain the expected loss of the flow rate of the multiple materials; it is determined whether the number of adjustments is less than the preset number and whether the expected loss is less than the loss threshold. If so, continue to adjust the position of the material to be regulated (such as the new house) among the specific materials; otherwise, restore to the display position before this adjustment.

[0111] If it is determined that the actual expectation of the flow rate of a specific material among the multiple materials is greater than or equal to the estimated expectation of the flow rate of the specific material, no regulation is required.

[0112] This embodiment can perform traffic control in different dimensions for multiple materials. After going online, it completes the traffic control in the click dimension of new houses and the exposure dimension of live broadcast cards. If new materials that need to be regulated are added, the online service part only needs to modify the configuration.

[0113] The above-mentioned embodiments of the present invention are applicable to the case where the specific material is a second-hand house, a rental house, etc. That is, by simply replacing the relevant description of the material of the new house, the relevant implementation methods of the materials such as the second-hand house and the rental house can be obtained, and they will not be described in detail here.

[0114] Specifically, now Figure 3 The flow shown briefly describes the flow control process.

[0115] like Figure 3 As shown, the flow control process may include the following steps S301-S305.

[0116] Step S301, determine whether to guarantee the quantity, if so, execute step S302, otherwise, end.

[0117] Step S302, read the quantity guarantee configuration.

[0118] Read the volume guarantee configuration to obtain the maximum number of adjustments and the maximum loss expectation threshold.

[0119] Step S303, reading estimated data and feedback data.

[0120] Read Figure 4 Estimated data on new homes / live broadcasts and other materials on display, business line preference data, and real-time feedback data.

[0121] Step S304, determine whether the user dimension is effective, if so, execute step S305, otherwise, end.

[0122] If the user dimension is effective, the quantity control strategy is applied to the user, that is, step S305 is executed.

[0123] At the same time, the users for whom the volume control function has been effective will be recorded in real time, so that the volume control function will no longer be effective for a period of time in the future, thereby improving traffic utilization.

[0124] Step S305: quantity maintenance control.

[0125] Combined with the maximum adjustment times and the maximum loss expectation threshold read in step S302 and the estimated data and feedback data read in step S303, corresponding quantity control is performed. For the quantity control process, please refer to the description of the above flow control process, which will not be repeated here. In addition, log records are made and the results are returned.

[0126] In summary, the present invention creatively obtains the actual expected flow rate of the specific material at the current moment in the current period when the first accumulated actual flow rate of a specific material among multiple materials in the current period in the target date up to the current moment is less than the estimated flow rate in the current period; then obtains the estimated expected flow rate of the specific material at the current moment in the current period in the target date based on the target flow rate of the specific material and the flow rate of the specific material in each period in the multiple preset dates; then, when the actual expected flow rate of the specific material is less than the estimated expected flow rate of the specific material, the position of the material to be regulated among the specific materials is adjusted based on the difference between the estimated expected flow rate of the specific material and the actual expected flow rate. Therefore, the present invention can accurately meet the needs of multi-material flow control in mixed recommendation scenarios.

[0127] An embodiment of the present invention also provides a flow control system. The flow control system includes: a first acquisition device for acquiring a first accumulated actual flow of a specific material among multiple materials in the current period of the target date up to the current moment; a second acquisition device for acquiring an estimated flow in the current period of the target date based on the target flow of the specific material and the flow of the specific material in the current period of multiple preset dates before the target date; a third acquisition device for acquiring an actual expectation of the flow of the specific material at the current moment in the current period of the target date when the first accumulated actual flow is less than the estimated flow; a fourth acquisition device for acquiring an estimated expectation of the flow of the specific material at the current moment in the current period of the target date based on the target flow of the specific material and the flow of the specific material in each period of the multiple preset dates; and an adjustment device for adjusting the position of the material to be controlled among the specific materials based on the difference between the estimated expectation and the actual expectation of the flow of the specific material when the actual expectation of the flow of the specific material is less than the estimated expectation of the flow of the specific material.

[0128] For specific details and benefits of the flow control system provided by the present invention, please refer to the above description of the flow control method, which will not be repeated here.

[0129] An embodiment of the present invention further provides a computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, the above-mentioned traffic control method is implemented.

[0130] An embodiment of the present invention further provides an electronic device, comprising: a processor; a memory for storing a computer program; the processor is configured to read the computer program from the memory and execute the computer program to implement the above-mentioned traffic control method.

[0131] An embodiment of the present invention further provides a computer program product, which includes a computer program. When the computer program is executed by a processor, the above-mentioned traffic control method is implemented.

[0132] The optional implementation modes of the embodiments of the present invention are described in detail above in conjunction with the accompanying drawings. However, the embodiments of the present invention are not limited to the specific details in the above implementation modes. Within the technical concept of the embodiments of the present invention, various simple modifications can be made to the technical scheme of the embodiments of the present invention, and these simple modifications all belong to the protection scope of the embodiments of the present invention.

[0133] It should also be noted that the various specific technical features described in the above specific embodiments can be combined in any suitable manner without contradiction. To avoid unnecessary repetition, the embodiments of the present invention will not further describe various possible combinations.

[0134] Those skilled in the art can understand that all or part of the steps in the above-mentioned embodiment method can be completed by instructing the relevant hardware through a program, and the program is stored in a storage medium, including a number of instructions to enable a single-chip microcomputer, a chip or a processor to execute all or part of the steps of the method described in each embodiment of the present application. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), disk or optical disk and other media that can store program codes.

[0135] In addition, various implementation modes of the embodiments of the present invention may be arbitrarily combined, and as long as they do not violate the concept of the embodiments of the present invention, they should also be regarded as the contents disclosed by the embodiments of the present invention.

Claims

1. A flow control method, characterized in that: The flow control method comprises: Obtaining a first accumulated actual flow rate of a specific material among a plurality of materials within a current period of a target date up to a current time; Based on the target flow rate of the specific material and the flow rate of the specific material in the current period of multiple preset dates before the target date, obtaining the estimated flow rate in the current period of the target date; When the first accumulated actual flow rate is less than the estimated flow rate, obtaining the actual expected flow rate of the specific material at the current time within the current period of the target date; Based on the target flow rate of the specific material and the flow rate of the specific material in each time period of the plurality of preset dates, obtaining an estimated expectation of the flow rate of the specific material at the current moment in the current time period of the target date; and When the actual expectation of the flow rate of the specific material is less than the estimated expectation of the flow rate of the specific material, based on the difference between the estimated expectation and the actual expectation of the flow rate of the specific material, the position of the material to be regulated in the specific material is adjusted, The step of obtaining the actual expected flow rate of the specific material at the current time within the current period of the target date includes: Obtaining an expected flow rate of the specific material based on the flow rate of the specific material at each display location at the current moment and an average value of the flow rates of the multiple materials at each display location on each of the multiple preset dates; and Based on the user's preference for the specific material and the expectation of the flow rate of the specific material, determine the actual expectation of the flow rate of the specific material, The step of obtaining the estimated expected flow rate of the specific material at the current time within the current period of the target date includes: Based on the target flow rate of the specific material and the flow rate of the specific material in the current period of the plurality of preset dates, obtaining an estimated probability of the flow rate of the specific material in the current period of the target date; Acquire a second accumulated actual flow rate of the specific material from the target date to the current time within the current period; Obtaining a second cumulative estimated flow rate of the specific material from the target date to the current time within the current period; and Based on the estimated probability of the flow of the specific material, the second accumulated actual flow, the second accumulated estimated flow and the global preference of the specific material, the estimated expectation of the flow of the specific material at the current moment in the current time period is determined.

2. The flow control method according to claim 1, characterized in that: The obtaining of the estimated traffic volume in the current period of the target date includes: Determining an estimated flow rate of the specific material in the current period based on the target flow rate of the specific material in the target date and the flow rate of the specific material in the current period among the plurality of preset dates, wherein the target flow rate of the specific material is associated with the flow rate of the specific material in the plurality of preset dates; and Based on the predicted flow in the current time period determined at a moment before the current moment and the first accumulated actual flow, the estimated flow in the current time period is updated to obtain the estimated flow in the current time period corresponding to the current moment.

3. The flow control method according to claim 1, characterized in that: The step of obtaining the second cumulative estimated flow rate of the specific material on the target date up to the current time in the current period includes: Determining an estimated flow rate of the specific material in each time period based on the target flow rate of the specific material on the target date and the flow rate of the specific material in each time period of the plurality of preset dates before the target date, wherein the target flow rate of the specific material is associated with the flow rate of the specific material in the plurality of preset dates; Based on the third accumulated actual flow and the third accumulated estimated flow of the specific material up to each moment in each time period before the current time period and the fourth accumulated actual flow and the fourth accumulated estimated flow of the specific material up to each moment in the current time period before the current moment, the estimated flow of the specific material in each time period before the current time period and the current time period is updated to obtain updated estimated flow of the specific material in each time period before the current time period and the current time period; and The second accumulated estimated flow rate of the specific material up to the current moment in the current period in the target date is determined based on the updated estimated flow rates of the specific material in each period before the current period and in the current period.

4. The flow control method according to claim 1, characterized in that: The determining of the estimated expectation of the flow rate of the specific material at the current moment in the current time period includes: Determining the control expectation of the flow rate of the specific material at the current moment in the current time period based on the estimated probability of the flow rate of the specific material, the second accumulated actual flow rate, and the second accumulated estimated flow rate; and Based on the global preference of the specific material and the expectation for regulating the flow of the specific material at the current moment in the current time period, the estimated expectation for the specific material at the current moment in the current time period is determined.

5. The flow control method according to claim 1, characterized in that: The specific materials include at least one of the following: new houses, second-hand houses, rental houses and live broadcasts.

6. The flow control method according to claim 1, characterized in that: The position adjustment of the material to be regulated in the specific material includes: Determining the material to be regulated based on the difference between the estimated expectation and the actual expectation of the flow rate of the specific material; and Adjust the display position of the material to be regulated.

7. The flow control method according to claim 6, characterized in that: The position adjustment of the material to be regulated in the specific material further includes: Determining the expected flow rates of the plurality of materials after adjustment and counting the number of adjustments; Determining an expected loss of the flow rate of the plurality of materials based on an actual expected flow rate of each of the plurality of materials and an adjusted expected flow rate of the plurality of materials; and When the counted number of adjustments is less than the preset number and the expected loss of the flow of the multiple materials is less than the loss threshold, continue to adjust the position of the material to be regulated among the specific materials.

8. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a computer program, which, when executed by a processor, implements the traffic control method described in any one of claims 1 to 7.

9. An electronic device, characterized in that: The electronic device comprises: processor; a memory for storing instructions executable by the processor; The processor is used to read the executable instructions from the memory and execute the instructions to implement the traffic control method described in any one of claims 1 to 7.

10. A computer program product, comprising a computer program, characterized in that: When the computer program is executed by a processor, the flow control method described in any one of claims 1 to 7 is implemented.

Images