CN108053101B - Budget control method and device for demander platform, storage medium and terminal - Google Patents
Budget control method and device for demander platform, storage medium and terminal Download PDFInfo
- Publication number
- CN108053101B CN108053101B CN201711215562.8A CN201711215562A CN108053101B CN 108053101 B CN108053101 B CN 108053101B CN 201711215562 A CN201711215562 A CN 201711215562A CN 108053101 B CN108053101 B CN 108053101B
- Authority
- CN
- China
- Prior art keywords
- budget
- difference
- period
- information
- spending
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/06—Resources, workflows, human or project management; Enterprise or organisation planning; Enterprise or organisation modelling
- G06Q10/063—Operations research, analysis or management
- G06Q10/0631—Resource planning, allocation, distributing or scheduling for enterprises or organisations
- G06Q10/06315—Needs-based resource requirements planning or analysis
Landscapes
- Business, Economics & Management (AREA)
- Human Resources & Organizations (AREA)
- Engineering & Computer Science (AREA)
- Strategic Management (AREA)
- Entrepreneurship & Innovation (AREA)
- Economics (AREA)
- Operations Research (AREA)
- Game Theory and Decision Science (AREA)
- Development Economics (AREA)
- Marketing (AREA)
- Educational Administration (AREA)
- Quality & Reliability (AREA)
- Tourism & Hospitality (AREA)
- Physics & Mathematics (AREA)
- General Business, Economics & Management (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
A budget control method and device, a storage medium and a terminal for a demand side platform are provided, wherein the method comprises the following steps: determining budget distribution proportions for the multiple equipment types according to respective average click consumption when information to be issued is issued on the multiple equipment types historically, wherein the budget distribution proportions corresponding to the equipment types with smaller average click consumption historically are larger; allocating a total budget to the plurality of device types according to the budget allocation proportion; and for each equipment type, adjusting the possibility that the information to be issued is issued to the equipment type according to the spending progress of the equipment type so as to enable the change rate of the spending progress along with the time to be less than or equal to a preset error range. The scheme of the invention can ensure that the budget allocation is more reasonable, track the actual cost condition of the budget in real time and ensure that the cost progress is in line with the expectation.
Description
Technical Field
The invention relates to the technical field of information release, in particular to a budget control method and device for a demander platform, a storage medium and a terminal.
Background
With the development of internet technology, people can surf the internet and browse the latest information and information all over the world without going out, and an information publisher can also publish information outwards by taking the internet as a platform.
A Demand Side Platform (DSP for short) is used as an agent of an information publisher, and needs to make a most reasonable information publishing policy for the information publisher, so that information to be published of the information publisher can be effectively published at a reasonable cost and is finally converted into a click amount.
However, the existing information distribution strategy still has limitations, and particularly in the aspect of budget control, the existing demand side platform cannot reasonably control budget allocation and spending progress, the former may result in a small amount of clicks which are finally obtained by investing great spending distribution information, and the latter may result in a situation that the spending progress is fast and slow and cannot meet the requirements of the information distributor, so that the information delivery effect of the information distributor is seriously affected.
Disclosure of Invention
The technical problem solved by the invention is how to reasonably control the budget distribution and the spending progress so as to improve the information distribution effect and ensure that the spending progress of the budget is in line with the expectation.
To solve the foregoing technical problem, an embodiment of the present invention provides a budget control method for a demand side platform, including: determining budget distribution proportions for the multiple equipment types according to respective average click consumption when information to be issued is issued on the multiple equipment types historically, wherein the budget distribution proportions corresponding to the equipment types with smaller average click consumption historically are larger; allocating a total budget to the plurality of device types according to the budget allocation proportion; and for each equipment type, adjusting the possibility that the information to be issued is issued to the equipment type according to the spending progress of the equipment type so as to enable the change rate of the spending progress along with the time to be less than or equal to a preset error range.
Optionally, the determining budget allocation proportions for the multiple device types according to the average click consumption of the multiple device types in the historical process of publishing information to be published includes: counting, for each device type, average click consumption of the device type historically according to a first period; respectively calculating the total average click consumption of each equipment type history; and obtaining the budget distribution proportion by adopting beta distribution simulation based on the historical total average click consumption of each equipment type.
Optionally, for each device type, adjusting the possibility of issuing the information to be issued to the device type according to the expense progress of the device type includes: and for each equipment type, adjusting the orientation threshold of the equipment type according to the expense progress of the equipment type, wherein the orientation threshold of the equipment type is used for judging whether to issue the information to be issued to the equipment type, issuing the information if the current condition of the equipment type meets the orientation threshold, and otherwise not issuing the information.
Optionally, for each device type, the adjusting the orientation threshold according to the spending progress of the device type includes: for each device type, partitioning the budget allocated to the device type by a second period; and adjusting the orientation threshold according to the spending progress of the equipment type in the second period by taking the second period as a unit.
Optionally, the adjusting the orientation threshold according to the spending progress of the device type in the second period by the unit of the second period includes: comparing the difference between the spending schedule in the second period and the time schedule, wherein the spending schedule in the second period is the proportion of the spending of the second period from beginning to the present in the budget allocated to the second period of the equipment type, and the time schedule in the second period is the proportion of the duration of the second period from beginning to the present in the total duration of the second period; and when the difference is larger than a preset error range, adjusting the orientation threshold.
Optionally, when the difference is greater than a preset error range, adjusting the orientation threshold includes: adjusting the orientation threshold based on a result of the PID adjustment to the difference.
Optionally, the orientation threshold is adjusted according to the following formula:
wherein T is the orientation threshold; the cur _ scale is the PID adjustment result of the pair of differences; the PID adjustment result for the difference is calculated according to the following formula: cur _ scale ═ cur _ scale _ h + (P × error + I × acc _ error + D × diff _ error), wherein the cur _ scale is the PID adjustment result; the cur _ scale _ h is the historical PID adjustment result of the difference; p, I and D are respectively preset PID parameters; the error is the difference; the acc _ error is integral difference, and the integral difference is the sum of historical accumulated difference and the difference; the diff _ error is a differential difference, which is the difference between the difference and the difference in the previous second cycle.
An embodiment of the present invention further provides a budget control device for a demander platform, including: the device comprises a determining module, a calculating module and a calculating module, wherein the determining module is used for determining budget distribution proportions aiming at a plurality of device types according to respective average click consumption when information to be issued is issued on the plurality of device types historically, and the budget distribution proportions corresponding to the device types with smaller average click consumption historically are larger; an allocation module for allocating a total budget to the plurality of device types according to the budget allocation proportion; and the adjusting module is used for adjusting the possibility that the information to be issued is issued to the equipment type according to the spending progress of the equipment type so that the change rate of the spending progress along with the time is smaller than or equal to a preset error range.
Optionally, the determining module includes: the statistics submodule is used for counting average click consumption of the equipment types in history according to a first period for each equipment type; the calculation submodule is used for calculating the total average click consumption of each equipment type history respectively; and the simulation submodule is used for obtaining the budget distribution proportion by adopting beta distribution simulation on the basis of the historical total average click consumption of each equipment type.
Optionally, the adjusting module includes: and the adjusting submodule adjusts the orientation threshold of each equipment type according to the expense progress of the equipment type, wherein the orientation threshold of the equipment type is used for judging whether to issue the information to be issued to the equipment type, if the current condition of the equipment type meets the orientation threshold, the information is issued, and if not, the information is not issued.
Optionally, the regulator sub-module includes: a partitioning unit that partitions, for each device type, a budget allocated to the device type according to a second period; a first adjusting unit that adjusts the orientation threshold according to a spending progress of the device type in the second period in units of the second period.
Optionally, the first adjusting unit includes: a comparing unit, configured to compare a difference between the spending schedule in the second period and a time schedule, where the spending schedule in the second period is a proportion of the spending of the second period starting up to now in the budget allocated to the second period of the device type, and the time schedule in the second period is a proportion of the duration of the second period starting up to now in the total duration of the second period; and the second adjusting unit is used for adjusting the orientation threshold value when the difference is larger than a preset error range.
Optionally, the second adjusting unit includes: a third adjusting unit that adjusts the orientation threshold based on a result of the PID adjustment of the difference.
Optionally, the third adjusting unit adjusts the orientation threshold according to the following formula:
wherein T is the orientation threshold; the cur _ scale is the PID adjustment result of the pair of differences; the PID adjustment result for the difference is calculated according to the following formula: cur _ scale ═ cur _ scale _ h + (P × error + I × acc _ error + D × diff _ error), wherein the cur _ scale is the PID adjustment result; the cur _ scale _ h is the historical PID adjustment result of the difference; p, I and D are respectively preset PID parameters; the error is the difference; the acc _ error is integral difference, and the integral difference is the sum of historical accumulated difference and the difference; the diff _ error is a differential difference, which is the difference between the difference and the difference in the previous second cycle.
The embodiment of the invention also provides a storage medium, wherein computer instructions are stored on the storage medium, and the computer instructions execute the steps of the method when running.
The embodiment of the present invention further provides a terminal, which includes a memory and a processor, where the memory stores computer instructions capable of running on the processor, and the processor executes the steps of the method when executing the computer instructions.
Compared with the prior art, the technical scheme of the embodiment of the invention has the following beneficial effects:
the technical scheme of the embodiment of the invention provides a budget control method for a demand side platform, which comprises the following steps: determining budget distribution proportions for the multiple equipment types according to respective average click consumption when information to be issued is issued on the multiple equipment types historically, wherein the budget distribution proportions corresponding to the equipment types with smaller average click consumption historically are larger; allocating a total budget to the plurality of device types according to the budget allocation proportion; and for each equipment type, adjusting the possibility that the information to be issued is issued to the equipment type according to the spending progress of the equipment type so as to enable the change rate of the spending progress along with the time to be less than or equal to a preset error range. Compared with the existing budget allocation scheme, the demand side platform adopting the scheme of the embodiment of the invention can allocate more budgets to a plurality of equipment types more reasonably based on the budget allocation principle that the equipment types with good information distribution effect allocate more budgets. Further, based on the solution of the embodiment of the present invention, the demand side platform can track the spending progress of the budget after being allocated to each device type in real time, so as to ensure that the budget can be smoothly spent, and the spending progress of the budget conforms to the expectation of a user (such as an information publisher). The technical personnel in the field understand that the scheme of the embodiment of the invention can ensure that the budget spending progress is in accordance with the user expectation while improving the information distribution effect.
Further, for each equipment type, adjusting an orientation threshold of the equipment type according to the expense progress of the equipment type, wherein the orientation threshold of the equipment type is used for judging whether to issue information to be issued to the equipment type, issuing the information if the current condition of the equipment type meets the orientation threshold, and otherwise not issuing the information. Those skilled in the art understand that by adjusting the orientation threshold, the possibility of the information to be published being published to the device type can be adjusted, and the budget spending progress can be regulated. For example, when the spending progress of the device type is too fast, the orientation threshold may be increased to reduce the possibility that the information to be published is published to the device type, thereby slowing the spending progress of the device type; for another example, when the spending progress of the device type is too slow, the targeting threshold may be adjusted down to increase the probability that the information to be published is published to the device type, thereby speeding up the spending progress of the device type.
Drawings
FIG. 1 is a flowchart of a budget control method for a demander platform according to a first embodiment of the present invention;
FIG. 2 is a revenue effect graph of a beta distribution simulation used in the first embodiment of the present invention;
FIG. 3 is a diagram of another revenue effect of the beta distribution simulation employed by the first embodiment of the present invention;
FIG. 4 is a diagram illustrating the effect of determining the budget allocation ratio by using beta distribution simulation according to the first embodiment of the present invention;
FIG. 5 is a diagram illustrating the effect of the first embodiment of the present invention in determining the budget allocation ratio using the historical ratio of average click consumptions of the respective device types;
FIG. 6 is a schematic representation of the variation of orientation threshold over time using the solution of the first embodiment of the invention;
FIG. 7 is a schematic representation of the change in cost over time of using the solution of the first embodiment of the invention;
FIG. 8 is a schematic structural diagram of a budget control device for a demand side platform according to a second embodiment of the present invention.
Detailed Description
As will be understood by those skilled in the art, as the background art, the existing Demand Side Platform (DSP) cannot provide a reasonable control scheme for budget, so that the information distribution effect is poor and the budget spending progress is uncontrollable.
In order to solve the above technical problem, a technical solution of an embodiment of the present invention provides a budget control method for a demand side platform, including: determining budget distribution proportions for the multiple equipment types according to respective average click consumption when information to be issued is issued on the multiple equipment types historically, wherein the budget distribution proportions corresponding to the equipment types with smaller average click consumption historically are larger; allocating a total budget to the plurality of device types according to the budget allocation proportion; and for each equipment type, adjusting the possibility that the information to be issued is issued to the equipment type according to the spending progress of the equipment type so as to enable the change rate of the spending progress along with the time to be less than or equal to a preset error range.
Those skilled in the art understand that the demand side platform adopting the scheme of the embodiment of the invention can allocate more budgets to a plurality of equipment types more reasonably based on the budget allocation principle that the equipment types with good information distribution effect allocate more budgets. Further, based on the solution of the embodiment of the present invention, the demand side platform can track the spending progress of the budget after being allocated to each device type in real time, so as to ensure that the budget can be smoothly spent, and the spending progress of the budget conforms to the expectation of a user (such as an information publisher). The technical personnel in the field understand that the scheme of the embodiment of the invention can ensure that the budget spending progress is in accordance with the user expectation while improving the information distribution effect.
In order to make the aforementioned objects, features and advantages of the present invention comprehensible, embodiments accompanied with figures are described in detail below.
Fig. 1 is a flowchart of a budget control method for a demander platform according to a first embodiment of the present invention. The budget can be used for the demander platform to release information; the main body of the information release can be single information to be released or any information to be released in the information group to be released; the information to be published included in the information publishing group to be published may be provided by the same information publisher, for example, for a first information publishing activity (campaigns) of the information publisher, the information publisher may provide a batch of information to be published, and the batch of information to be published may be used as the information group to be published.
Specifically, in this embodiment, the budget control method may include the following steps:
step S101, according to respective average click consumption when information to be issued is issued on a plurality of device types historically, determining budget distribution proportions for the plurality of device types, wherein the budget distribution proportions corresponding to the device types with smaller average click consumption historically are larger.
And step S102, distributing the total budget to the plurality of equipment types according to the budget distribution proportion.
Step S103, for each equipment type, adjusting the possibility of issuing the information to be issued to the equipment type according to the spending progress of the equipment type so as to enable the change rate of the spending progress along with the time to be less than or equal to a preset error range.
More specifically, the device type may be used to describe different device sides, such as a Personal Computer (PC) side and a mobile device side.
Further, the device type may also be used to describe devices using different systems, such as devices using an android System (e.g., mobile devices such as a mobile phone) and devices using an apple Operating System (IOS).
Further, for each device type, multiple devices may be associated in the internet, and the demander platform adopting the scheme of this embodiment may select a part of the devices from the multiple devices to publish the information to be published to the selected part of the devices. Accordingly, the cost of distributing information to this portion of devices uses the budget allocated to that device type.
Further, the Cost Per Click (CPC) may be used to measure an information publishing effect of the information to be published on the device type. As will be understood by those skilled in the art, since the average click consumption is the ratio of the total cost for delivering the information to be published to the device type and the number of clicks, a smaller value of the average click consumption means a higher number of clicks, i.e., a better information publishing effect on the device type, and the demand side platform can determine the budget allocation ratio for each device type subsequently.
As a non-limiting example, the step S101 may include the steps of: counting, for each device type, average click consumption of the device type historically according to a first period; respectively calculating the total average click consumption of each equipment type history; and obtaining the budget distribution proportion by adopting beta distribution simulation based on the historical total average click consumption of each equipment type.
Preferably, the first period may be 1 hour.
Taking the device types including an ios device type and an android device type as an example, average click consumption of the information to be published on the ios device type is counted once per hour from the time when the information to be published is released to the present, and an average click consumption list ios _ cpc _ list of the ios device type [ cpc1, cpc2, …, cpcN ] can be obtained, where cpcN is average click consumption of the information to be published on the ios device type in an nth first period (i.e., nth hour). Similarly, until the information to be published is released, counting average click consumption of the information to be published on the android device type once per hour, and obtaining an average click consumption list android _ cpc _ list of the android device type [ cpc1, cpc2, …, cpcN ], where cpcN is average click consumption of the information to be published on the android device type in the nth first period (i.e., nth hour).
And respectively calculating and obtaining total average click consumption sum (ios _ cpc _ list) and sum (android _ cpc _ list) in the history of the ios device type and the android device type according to the average click consumption list of the ios device type and the average click consumption list of the android device type. For example, the total average click consumption of the ios device type may be obtained by accumulating all average click consumptions in the average click consumption list of the ios device type.
The information release per hour is considered as an independent experiment, and the yield is the average click consumption per hour to perform beta distribution simulation, so that the simulation effect graphs shown in fig. 2 and 3 can be obtained. In particular, the beta distribution can be understood as the result of observation tests based on multiple tosses of a coin, the result being a probability distribution of the probability of the face of the coin. The number of the coin tossing observation tests is the number of the independent tests, namely the number of the intervals obtained by dividing according to the first period.
Further, the average click consumption proportion of the information to be published on different device types can be simulated through the beta distribution. Preferably, the average click consumption weight of the device type may be understood as a weight, i.e. a budget allocation proportion for the device type.
For example, the average click consumption weight of the android device type may be determined by a beta distribution simulation according to the following formula:
android_cpc_probability＝beta(sum(android_cpc_list),sum(ios_cpc_list))
wherein the android _ cpc _ performance is an average click consumption proportion of the android device type; the beta () is a beta distribution function.
The expected value of the average click consumption proportion of the android device type can be determined based on the formula sum (ios _ cpc _ list)/(sum (ios _ cpc _ list) + sum (android _ cpc _ list)), and at the initial stage of information distribution, a certain degree of variance may exist, and as the information distribution progresses, the data included in the average click consumption list of the android device type is more and more, and the variance gradually decreases.
Similarly, the average click consumption specific gravity of the ios device type may be determined by beta distribution simulation according to the following formula:
ios_cpc_probability＝beta(sum(ios_cpc_list),sum(android_cpc_list))
wherein the ios _ cpc _ capability is an average click consumption specific gravity of the ios device type.
Further, in step S102, according to the determined average click consumption specific gravity of the ios device type and the average click consumption specific gravity of the android device type, a total budget may be allocated to the android device type and the ios device type.
For example, the budget ios _ budget allocated to the ios device type may be determined by calculating a formula ios _ budget ═ budget × ios _ cpc _ probability. Wherein the budget is the total budget.
For another example, the budget android _ budget allocated to the android device type may be determined by a formula android _ budget × android _ cpc _ robustness calculation.
In practical applications, the device types may not be limited to the android device type and the ios device type, and are not described herein again.
As a variation, the budget allocation ratio may also be determined according to the ratio between the average click consumptions of the historical device types, so as to ensure that a smaller average click consumption can be allocated to a larger budget. However, the budget allocation effect exploration component finally obtained in this variation is relatively low compared to the scheme using the beta distribution simulation.
Referring to fig. 2, it is understood by those skilled in the art that the beta distribution simulation has the advantages that, in the early stage of information distribution (i.e. when the information to be distributed is just distributed), less data of average click consumption can be obtained, the confidence interval obtained by the beta distribution simulation is wider (see fig. 2), and since the final average click consumption proportion is randomly obtained from the values covered by the histogram of fig. 2 (but the probability that the values closer to the two ends are obtained is smaller), the value of the average click consumption proportion is wider, and the probability that the average click consumption proportion is deviated from the central point is larger, which is beneficial to search when the data is insufficient.
As the information distribution continues, the data of the average click consumption obtained gradually increases, the confidence interval obtained by adopting beta distribution simulation gradually narrows (see fig. 3), the finally obtained average click consumption proportion is basically limited in a small numerical range, no exploration effect exists, and the effect of the previous information distribution has a great influence on the subsequent budget distribution proportion.
Further, referring to fig. 4, when the budget allocation ratio is determined by using the foregoing beta distribution simulation, it can be seen from the figure that the fluctuation of the budget allocation ratio is large at the initial stage of information distribution, so that the demanding platform can perform deeper exploration of information distribution among different device types, thereby determining a more reasonable budget allocation ratio at the later stage of information distribution.
In contrast, referring to fig. 5, when the budget allocation ratio is determined using the ratio of the average click consumptions of the respective device types in the history, the average click consumption of the respective device types at the initial stage of information distribution highly affects the following budget allocation ratio, and the exploration component is small.
Those skilled in the art may select the beta distribution simulation or ratio according to actual needs, or determine the budget distribution ratio by other methods, which is not described herein.
It should be noted that the specific duration of the first period may also be changed according to actual needs; in addition to the average click consumption, the budget allocation ratio may also be obtained by using the beta distribution simulation based on other indexes that can also describe the information distribution effect of the information to be distributed on the device type, which is not described herein again.
Further, the predetermined error range may be ± 8% (0.5/60), and those skilled in the art may adjust the specific value of the predetermined error range according to actual needs, which is not described herein.
Further, the step S103 may include: and for each equipment type, adjusting the orientation threshold of the equipment type according to the expense progress of the equipment type, wherein the orientation threshold of the equipment type is used for judging whether to issue the information to be issued to the equipment type, issuing the information if the current condition of the equipment type meets the orientation threshold, and otherwise not issuing the information.
For example, when information is published, for a traffic of a specific device type (that is, a behavior of triggering an information publishing requirement on a device of the device type), the demander platform may direct the traffic, determine whether the traffic is a good-quality traffic according to a preset optimization rule, and publish the information to be published to a device that generates the traffic that is determined as the good-quality traffic.
Preferably, the content included in the current condition of the device type may be determined according to the preset optimization rule. For example, the current condition of the device type may include attribute information (e.g., age, gender, etc.) of a user associated with the device generating the traffic.
As a non-limiting example, it may be determined whether the traffic can be rated as premium traffic by setting the directional threshold.
For example, after the traffic is subjected to the directional operation, a score value for the traffic may be obtained, and the score value and the directional threshold may both be [0,1 ]. Further, if the score value of the flow is smaller than the orientation threshold, the orientation is unsuccessful, the flow is not a high-quality flow, and the platform of the demand party cannot execute information issuing operation aiming at the flow; if the score value of the flow is larger than the orientation threshold value, the orientation is successful, the flow is high-quality flow, and the demand side platform executes information issuing operation aiming at the flow so as to issue the information to be issued to the equipment generating the flow.
Those skilled in the art understand that by adjusting the orientation threshold, the possibility of the information to be published being published to the device type can be adjusted, and the budget spending progress can be regulated. For example, when the spending progress of the device type is too fast, the orientation threshold may be increased to reduce the possibility that the information to be published is published to the device type, thereby slowing the spending progress of the device type; for another example, when the spending progress of the device type is too slow, the targeting threshold may be adjusted down to increase the probability that the information to be published is published to the device type, thereby speeding up the spending progress of the device type.
Or, the evaluation criterion of the high-quality traffic can be adjusted by adjusting the optimization rule, so that the possibility that the information to be issued is issued to the device type is adjusted from the other side, and the effect of regulating and controlling the spending progress of the budget allocated to the device type is achieved.
As a non-limiting example, for each device type, the budget allocated to the device type may be split by a second period; and adjusting the orientation threshold according to the spending progress of the equipment type in the second period by taking the second period as a unit.
Further, the DSP platform may segment the budget allocated to the device type by taking the second period as a unit, track and monitor the spending progress of the device type in each second period, and ensure that the budget spending progress in each second period meets expectations by a refined tracking strategy, thereby avoiding occurrence of an excessive error.
Preferably, the duration of the second period may be the same as the duration of the first period, such as 1 hour; alternatively, the duration of the second period may also be different from the duration of the first period, and those skilled in the art may change more embodiments according to actual needs, which is not described herein.
Preferably, the budget allocated to the device type may be divided into equal parts according to the number of the second periods, so as to allocate the budget to each second period evenly.
Alternatively, the budget difference may be allocated to each second cycle according to different time periods in which the second cycles are located. For example, for a time period in which average clicks historically consume less, more budget may be allocated to the second period included in the time period; for periods of historically larger average click consumption, less budget may be allocated to the second period encompassed by the period.
Further, a difference between the spending schedule in the second period and the time schedule may be compared, the spending schedule in the second period is a proportion of the duration of the second period from the beginning to the end in the budget allocated to the second period of the device type, and the time schedule in the second period is a proportion of the duration of the second period from the beginning to the end in the total duration of the second period; and when the difference is larger than a preset error range, adjusting the orientation threshold.
In a preferred embodiment, the preset error range may be determined according to the preset error range.
For example, for a particular device type, it may be determined whether the orientation threshold needs to be adjusted according to the following formula:
wherein cur _ minute is the current time; the cost _ minimum is the cost of the second cycle starting up to now; the budget allocated to the second period is the budget of the budget; the T2 is the duration of the second period; the cost _ minimum/budget _ route is the spending progress in the second period; cur _ minimum/T is the time progress in the second cycle. Preferably, when the duration of the second period is 1 hour, the time T2 is 60 minutes.
Specifically, when the difference between the spending schedule and the time schedule in the second period is ± (0.5/T2), it may be determined that the spending schedule of the budget is as expected without adjusting the orientation threshold.
And when the difference between the spending schedule and the time schedule in the second period is larger than (0.5/T2) or smaller than- (0.5/T), the spending schedule of the budget can be determined to be unexpected, and the orientation threshold value needs to be adjusted.
Further, the orientation threshold may be adjusted based on PID (proportional, integral, derivative) adjustments to the difference.
As one non-limiting example, the orientation threshold may be adjusted according to the following Sigmoid (Sigmoid) function:
wherein T is the orientation threshold; the cur _ scale is the PID adjustment result of the pair of differences.
Further, the PID adjustment result for the difference may be calculated according to the following formula:
cur_scale＝cur_scale_h+(P×error+I×acc_error+D×diff_error)
wherein, the cur _ scale is the PID adjustment result; the cur _ scale _ h is the historical PID adjustment result of the difference; p, I and D are respectively preset PID parameters; the error is the difference; the acc _ error is integral difference, and the integral difference is the sum of historical accumulated difference and the difference; the diff _ error is the differential difference, which in the discrete case is the difference between the difference and the difference in the previous second cycle.
Preferably, the specific value of the preset PID parameter can be set and adjusted according to actual needs and the execution effect of the embodiment, which is not described herein again.
As understood by those skilled in the art, since the range of values of the PID adjustment result cur _ scale for the difference may be (— infinity, + ∞), in order to be able to adapt to the range of values of the scoring value [0,1], the PID adjustment result cur _ scale for the difference may be monotonically and smoothly mapped into the range of values of [0,1] by the above normalization formula. Besides the Sigmoid function, other functions capable of monotonically and smoothly mapping the PID adjustment result cur _ scale of the pair of differences to a numerical range of [0,1] may be used to perform the operation of adjusting the orientation threshold in this embodiment, which is not described herein again.
In a typical application scenario, taking an android device type as an example, a final effect of the requiring side platform adopting the scheme of the present embodiment may be shown in fig. 6 and 7, where fig. 6 shows a change situation of an orientation threshold adjusted by adopting the scheme of the present embodiment with time, and correspondingly, fig. 7 shows a change curve of a spending progress of the requiring side platform adopting the scheme of the present embodiment on the android device type with time. In the application scenario, the second period is 1 hour.
It can be seen that the law of change of the spending progress described in fig. 7 is influenced by the adjustment result of the orientation threshold described in fig. 6.
Specifically, referring to fig. 6, curve 3 is an actual variation curve of the orientation threshold, and curve 4 is an initial value of the orientation threshold at the beginning of the second period.
Specifically, referring to fig. 7, a curve 5 is an actual variation curve of the spending progress, and a curve 6 is an ideal variation curve of the spending progress (i.e., a time progress curve).
Further, in conjunction with fig. 6 and 7, when the expense progress of the android device type is too fast (i.e. the slope of the curve 5 is greater than the slope of the curve 6), the orientation threshold (i.e. the portion of the curve in the same second period where the value of the curve 3 is greater than the value of the curve 4) may be increased to reduce the possibility that the information to be published is published to the device type, thereby slowing the expense progress of the device type (i.e. reducing the slope of the curve 5 to be equal to the slope of the curve 6); for another example, when the spending progress of the device type is too slow (i.e. the slope of the curve 5 is smaller than the slope of the curve 6), the orientation threshold (i.e. the portion of the curve in the same second period where the value of the curve 3 is smaller than the value of the curve 4) may be adjusted downward to increase the probability that the information to be published is published to the device type, thereby speeding up the spending progress of the device type (i.e. increasing the slope of the curve 5 to be equal to the slope of the curve 6).
Thus, by adopting the scheme of the embodiment, the total budget can be more reasonably allocated to a plurality of equipment types based on the budget allocation principle that the equipment types with good information distribution effect allocate more budgets. Further, based on the solution of the embodiment of the present invention, the demand side platform can track the spending progress of the budget after being allocated to each device type in real time, so as to ensure that the budget can be smoothly spent, and the spending progress of the budget conforms to the expectation of a user (such as an information publisher). For example, with the solution of the present embodiment, it is ensured that the orientation threshold can be always maintained at the highest point just enabling the budget at that time to be spent, so that the curves 5 and 6 in fig. 7 can be as closely fitted as possible, thereby obtaining the optimal information distribution effect in the case that the budget allocated to the specific second period of the specific device type is spent. The technical personnel in the field understand that the scheme of the embodiment of the invention can ensure that the budget spending progress is in accordance with the user expectation while improving the information distribution effect.
FIG. 8 is a schematic structural diagram of a budget control device for a demand side platform according to a second embodiment of the present invention. Those skilled in the art will understand that the budget control device 7 for the platform on the demand side (hereinafter referred to as the budget control device 7) in this embodiment can be used to implement the method solutions described in the embodiments shown in fig. 1 to 7.
Specifically, in this embodiment, the budget control device 7 may include a determining module 71, configured to determine budget allocation proportions for a plurality of device types according to respective average click consumptions when historically publishing information to be published on the plurality of device types, where a device type with a smaller average click consumption historically corresponds to a larger budget allocation proportion; an allocation module 72 for allocating a total budget to the plurality of device types according to the budget allocation proportion; the adjusting module 73 adjusts, for each device type, the possibility that the information to be issued is issued to the device type according to the spending progress of the device type so that the change rate of the spending progress with time is smaller than or equal to a preset error range.
Further, the determination module 71 may include a statistics sub-module 711 for counting, for each device type, average click consumptions of the device types over the history according to a first period; a calculation submodule 712, configured to calculate total average click consumptions in histories of the respective device types respectively; and the simulation submodule 713 is configured to obtain the budget allocation proportion by using beta distribution simulation based on the historical total average click consumption of each device type.
Further, the adjusting module 73 may include an adjusting sub-module 731, which adjusts, for each device type, an orientation threshold of the device type according to the spending progress of the device type, where the orientation threshold of the device type is used to determine whether to publish information to be published to the device type, and if the current condition of the device type meets the orientation threshold, publish the information, otherwise, not publish the information.
Further, the adjusting sub-module 731 may include a partitioning unit 7311, for each device type, partitioning the budget allocated to the device type by a second period; a first adjusting unit 7312, taking the second period as a unit, adjusts the orientation threshold according to the spending progress of the device type in the second period.
Further, the first adjusting unit 7312 may include a comparing unit 73121, configured to compare a difference between a spending schedule in the second period and a time schedule, where the spending schedule in the second period is a proportion of the spending of the second period from beginning to end in the budget allocated to the second period of the device type, and the time schedule in the second period is a proportion of the duration of the second period from beginning to end in the total duration of the second period; a second adjusting unit 73122, adjusting the orientation threshold when the difference is larger than a preset error range.
Further, the second adjusting unit 73122 may comprise a third adjusting unit 731221, which adjusts the orientation threshold based on the PID adjustment result of the difference.
In a preferred example, the third adjusting unit 731221 may adjust the orientation threshold according to the following formula:
wherein T is the orientation threshold; the cur _ scale is the PID adjustment result of the pair of differences.
Preferably, the PID adjustment result for the difference may be calculated according to the following formula:
cur_scale＝cur_scale_h+(P×error+I×acc_error+D×diff_error)
wherein, the cur _ scale is the PID adjustment result; the cur _ scale _ h is the historical PID adjustment result of the difference; p, I and D are respectively preset PID parameters; the error is the difference; the acc _ error is integral difference, and the integral difference is the sum of historical accumulated difference and the difference; the diff _ error is a differential difference, which is the difference between the difference and the difference in the previous second cycle.
For more details of the operation principle and the operation mode of the budget control device 7, reference may be made to the description in fig. 1 to 7, which are not repeated herein.
Further, the embodiment of the present invention further discloses a storage medium, on which computer instructions are stored, and when the computer instructions are executed, the method technical solution described in the embodiments shown in fig. 1 to fig. 7 is executed. Preferably, the storage medium may include a computer-readable storage medium. The storage medium may include ROM, RAM, magnetic or optical disks, etc.
Further, an embodiment of the present invention further discloses a terminal, which includes a memory and a processor, where the memory stores a computer instruction capable of running on the processor, and the processor executes the method technical solution described in the embodiments shown in fig. 1 to 7 when running the computer instruction.
Although the present invention is disclosed above, the present invention is not limited thereto. Various changes and modifications may be effected therein by one skilled in the art without departing from the spirit and scope of the invention as defined in the appended claims.
Claims (14)
1. A budget control method for a demander platform, comprising:
determining budget distribution proportions for the multiple equipment types according to respective average click consumption when information to be issued is issued on the multiple equipment types historically, wherein the budget distribution proportions corresponding to the equipment types with smaller average click consumption historically are larger;
allocating a total budget to the plurality of device types according to the budget allocation proportion;
for each equipment type, adjusting the possibility that the information to be issued is issued to the equipment type according to the expense progress of the equipment type so that the change rate of the expense progress along with time is smaller than or equal to a preset error range;
wherein determining budget allocation proportions for a plurality of device types according to respective average click consumptions historically used when publishing information to be published on the plurality of device types comprises:
counting, for each device type, average click consumption of the device type historically according to a first period;
respectively calculating the total average click consumption of each equipment type history;
and obtaining the budget distribution proportion by adopting beta distribution simulation based on the historical total average click consumption of each equipment type.
2. The budget control method according to claim 1, wherein for each device type, adjusting the probability of the information to be published being published to the device type according to the spending progress of the device type comprises:
and for each equipment type, adjusting the orientation threshold of the equipment type according to the expense progress of the equipment type, wherein the orientation threshold of the equipment type is used for judging whether to issue the information to be issued to the equipment type, issuing the information if the current condition of the equipment type meets the orientation threshold, and otherwise not issuing the information.
3. The budget control method according to claim 2, wherein the adjusting, for each device type, the orientation threshold according to the spending progress of the device type comprises:
for each device type, partitioning the budget allocated to the device type by a second period;
and adjusting the orientation threshold according to the spending progress of the equipment type in the second period by taking the second period as a unit.
4. The budget control method according to claim 3, wherein the adjusting the directional threshold according to the spending progress of the device type in the second period in units of the second period comprises:
comparing the difference between the spending schedule in the second period and the time schedule, wherein the spending schedule in the second period is the proportion of the spending of the second period from beginning to the present in the budget allocated to the second period of the equipment type, and the time schedule in the second period is the proportion of the duration of the second period from beginning to the present in the total duration of the second period;
and when the difference is larger than a preset error range, adjusting the orientation threshold.
5. The budget control method according to claim 4, wherein the adjusting the orientation threshold when the difference is larger than a preset error range comprises:
adjusting the orientation threshold based on a result of the PID adjustment to the difference.
6. The budget control method according to claim 5, wherein the orientation threshold is adjusted according to the following formula:
wherein T is the orientation threshold; the cur _ scale is the PID adjustment result of the pair of differences;
the PID adjustment result for the difference is calculated according to the following formula:
cur_scale＝cur_scale_h+(P×error+I×acc_error+D×diff_error)
wherein, the cur _ scale is the PID adjustment result; the cur _ scale _ h is the historical PID adjustment result of the difference; p, I and D are respectively preset PID parameters; the error is the difference; the acc _ error is integral difference, and the integral difference is the sum of historical accumulated difference and the difference; the diff _ error is a differential difference, which is the difference between the difference and the difference in the previous second cycle.
7. A budget control device for a demander platform, comprising:
the device comprises a determining module, a calculating module and a calculating module, wherein the determining module is used for determining budget distribution proportions aiming at a plurality of device types according to respective average click consumption when information to be issued is issued on the plurality of device types historically, and the budget distribution proportions corresponding to the device types with smaller average click consumption historically are larger;
an allocation module for allocating a total budget to the plurality of device types according to the budget allocation proportion;
the adjusting module is used for adjusting the possibility that the information to be issued is issued to the equipment type according to the spending progress of the equipment type so that the change rate of the spending progress along with the time is smaller than or equal to a preset error range;
wherein the determining module comprises:
the statistics submodule is used for counting average click consumption of the equipment types in history according to a first period for each equipment type;
the calculation submodule is used for calculating the total average click consumption of each equipment type history respectively;
and the simulation submodule is used for obtaining the budget distribution proportion by adopting beta distribution simulation on the basis of the historical total average click consumption of each equipment type.
8. The budget control device of claim 7, wherein the adjusting module comprises:
and the adjusting submodule adjusts the orientation threshold of each equipment type according to the expense progress of the equipment type, wherein the orientation threshold of the equipment type is used for judging whether to issue the information to be issued to the equipment type, if the current condition of the equipment type meets the orientation threshold, the information is issued, and if not, the information is not issued.
9. The budget control device of claim 8, wherein the adjustment sub-module comprises:
a partitioning unit that partitions, for each device type, a budget allocated to the device type according to a second period;
a first adjusting unit that adjusts the orientation threshold according to a spending progress of the device type in the second period in units of the second period.
10. The budget control device according to claim 9, wherein the first adjusting unit comprises:
a comparing unit, configured to compare a difference between the spending schedule in the second period and a time schedule, where the spending schedule in the second period is a proportion of the spending of the second period starting up to now in the budget allocated to the second period of the device type, and the time schedule in the second period is a proportion of the duration of the second period starting up to now in the total duration of the second period;
and the second adjusting unit is used for adjusting the orientation threshold value when the difference is larger than a preset error range.
11. The budget control device according to claim 10, wherein the second adjusting unit comprises:
a third adjusting unit that adjusts the orientation threshold based on a result of the PID adjustment of the difference.
12. The budget control device according to claim 11, wherein the third adjusting unit adjusts the orientation threshold according to the following formula:
wherein T is the orientation threshold; the cur _ scale is the PID adjustment result of the pair of differences;
the PID adjustment result for the difference is calculated according to the following formula:
cur_scale＝cur_scale_h+(P×error+I×acc_error+D×diff_error)
wherein, the cur _ scale is the PID adjustment result; the cur _ scale _ h is the historical PID adjustment result of the difference; p, I and D are respectively preset PID parameters; the error is the difference; the acc _ error is integral difference, and the integral difference is the sum of historical accumulated difference and the difference; the diff _ error is a differential difference, which is the difference between the difference and the difference in the previous second cycle.
13. A storage medium having stored thereon computer instructions, wherein said computer instructions when executed perform the steps of the method of any of claims 1 to 6.
14. A terminal comprising a memory and a processor, the memory having stored thereon computer instructions executable on the processor, wherein the processor, when executing the computer instructions, performs the steps of the method of any one of claims 1 to 6.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201711215562.8A CN108053101B (en) | 2017-11-28 | 2017-11-28 | Budget control method and device for demander platform, storage medium and terminal |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201711215562.8A CN108053101B (en) | 2017-11-28 | 2017-11-28 | Budget control method and device for demander platform, storage medium and terminal |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN108053101A CN108053101A (en) | 2018-05-18 |
| CN108053101B true CN108053101B (en) | 2021-01-19 |
Family
ID=62120825
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201711215562.8A Active CN108053101B (en) | 2017-11-28 | 2017-11-28 | Budget control method and device for demander platform, storage medium and terminal |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN108053101B (en) |
Families Citing this family (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109087206A (en) * | 2018-08-01 | 2018-12-25 | 深圳前海桔子信息技术有限公司 | A kind of dissemination method of financial product, device, equipment and storage medium |
| CN109886736B (en) * | 2019-01-28 | 2021-08-10 | 上海晶赞融宣科技有限公司 | Budget data processing method and device for DSP platform, storage medium and terminal |
| CN110545450B (en) * | 2019-09-09 | 2021-12-03 | 深圳市网心科技有限公司 | Node distribution method, system, electronic equipment and storage medium |
Citations (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104574160A (en) * | 2015-01-29 | 2015-04-29 | 深圳市英威诺科技有限公司 | Smooth advertisement traffic control method |
Family Cites Families (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103136695B (en) * | 2013-03-26 | 2017-06-13 | 明日互动（北京）广告传媒有限公司 | Media deliver Value Engineering Method and its system |
| CN104299153A (en) * | 2014-10-27 | 2015-01-21 | 北京金和软件股份有限公司 | Advertisement promoting and statistic method based on network platform |
| CN104851023B (en) * | 2015-05-07 | 2018-11-23 | 容一飞 | A kind of real time bid online feedback control method and system |
| US20170098236A1 (en) * | 2015-10-02 | 2017-04-06 | Yahoo! Inc. | Exploration of real-time advertising decisions |
| CN106504012A (en) * | 2016-09-18 | 2017-03-15 | 北京三快在线科技有限公司 | A kind of method of adjustment of characteristic parameter, device and electronic equipment |
| CN106447396A (en) * | 2016-09-22 | 2017-02-22 | 晶赞广告（上海）有限公司 | Method for allocating online advertising budget |
| CN106803194A (en) * | 2017-01-20 | 2017-06-06 | 晶赞广告（上海）有限公司 | Online competitive price probabilistic model generation method |
-
2017
- 2017-11-28 CN CN201711215562.8A patent/CN108053101B/en active Active
Patent Citations (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104574160A (en) * | 2015-01-29 | 2015-04-29 | 深圳市英威诺科技有限公司 | Smooth advertisement traffic control method |
Non-Patent Citations (1)
| Title |
|---|
| 在线广告中实时竞价机制研究与算法实现;朱丽辉;《中国优秀硕士学位论文全文数据库 信息科技辑》;20151015;正文第1-6页 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN108053101A (en) | 2018-05-18 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Xu et al. | Smart pacing for effective online ad campaign optimization | |
| CN108053101B (en) | Budget control method and device for demander platform, storage medium and terminal | |
| US9310785B2 (en) | Apparatus and a method for controlling power supply and demand | |
| Agmon Ben-Yehuda et al. | Ginseng: Market-driven memory allocation | |
| US9031941B2 (en) | Information processing apparatus, terminal apparatus, information presentation system, calculation method of evaluation scores, and program | |
| US20100262497A1 (en) | Systems and methods for controlling bidding for online advertising campaigns | |
| US20180300750A1 (en) | Method and apparatus for delivering media files | |
| CN107045455A (en) | A kind of Docker Swarm cluster resource method for optimizing scheduling based on load estimation | |
| WO2012134765A1 (en) | Selective delivery of content items | |
| CN103336722B (en) | A kind of CPU resources of virtual machine monitoring and dynamic allocation method | |
| CN109886736B (en) | Budget data processing method and device for DSP platform, storage medium and terminal | |
| Wang et al. | Frequency regulation service provision in data center with computational flexibility | |
| JP2010512604A (en) | Bid optimization in search engine marketing | |
| CN109978583A (en) | A kind of control method and device that advertisement is launched | |
| US20150112795A1 (en) | Efficient budget pacing | |
| US8849695B2 (en) | Energy efficiency improvements in cloud-based environments | |
| CN108932627B (en) | Estimation method and device for consumption speed of CPA (continuous advertisement access) advertisement and readable storage medium | |
| Lang et al. | Handling forecast errors while bidding for display advertising | |
| Mehrtash et al. | Reserve and energy scarcity pricing in United States power markets: A comparative review of principles and practices | |
| CN113450211A (en) | User credit granting method, device, electronic equipment and storage medium | |
| CN113222720A (en) | Reputation-based privacy protection incentive mechanism method, device and storage medium | |
| Chen et al. | Data center optimal regulation service reserve provision with explicit modeling of quality of service dynamics | |
| CN111724181A (en) | Information pushing method, device, medium and electronic equipment | |
| CN104899462B (en) | Determine the method and device of the action value of information releasing position | |
| CN111210262A (en) | Spontaneous edge application deployment and pricing method based on incentive mechanism |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |