CN109829753B - Method and equipment for distributing flow to advertisement orders - Google Patents
Method and equipment for distributing flow to advertisement orders Download PDFInfo
- Publication number
- CN109829753B CN109829753B CN201910015119.9A CN201910015119A CN109829753B CN 109829753 B CN109829753 B CN 109829753B CN 201910015119 A CN201910015119 A CN 201910015119A CN 109829753 B CN109829753 B CN 109829753B
- Authority
- CN
- China
- Prior art keywords
- flow
- order
- orders
- advertisement
- optimization
- 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
- 238000000034 method Methods 0.000 title claims abstract description 29
- 238000009826 distribution Methods 0.000 claims abstract description 143
- 238000005457 optimization Methods 0.000 claims abstract description 119
- 238000004422 calculation algorithm Methods 0.000 claims abstract description 50
- XLYOFNOQVPJJNP-UHFFFAOYSA-N water Substances O XLYOFNOQVPJJNP-UHFFFAOYSA-N 0.000 claims abstract description 13
- 238000013178 mathematical model Methods 0.000 claims description 5
- 238000004590 computer program Methods 0.000 description 10
- 238000010586 diagram Methods 0.000 description 8
- 230000000694 effects Effects 0.000 description 5
- 238000003860 storage Methods 0.000 description 5
- 230000006870 function Effects 0.000 description 4
- 238000012986 modification Methods 0.000 description 3
- 230000004048 modification Effects 0.000 description 3
- 238000004364 calculation method Methods 0.000 description 2
- 101100234002 Drosophila melanogaster Shal gene Proteins 0.000 description 1
- 235000015076 Shorea robusta Nutrition 0.000 description 1
- 244000166071 Shorea robusta Species 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 238000004519 manufacturing process Methods 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 230000008569 process Effects 0.000 description 1
- 230000008685 targeting Effects 0.000 description 1
- 230000001960 triggered effect Effects 0.000 description 1
Images
Landscapes
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention discloses a method and equipment for distributing flow to advertisement orders, which are used for optimizing an initial flow distribution scheme obtained by an HWM algorithm, and continuously distributing the rest advertisement orders by adjusting the proportion of various flows distributed to each order, so that the distribution scheme can be obtained in a short time, and the compensation loss is reduced. The method comprises the following steps: obtaining an advertisement order, and determining an initial flow distribution scheme of the advertisement order by using a high-water-level HWM algorithm; and performing at least one time of optimization based on the initial flow distribution scheme, wherein the flow distribution scheme obtained by completing the optimization each time is the flow distribution scheme before the next optimization, and when performing the optimization each time, distributing at least part of flow required by meeting the order flow demand for the rest advertisement orders according to the flow distribution scheme before optimizing and the flow required by meeting the order flow demand for the rest advertisement orders which do not meet the order flow demand.
Description
Technical Field
The present invention relates to advertisement distribution, and in particular, to a method and apparatus for distributing traffic to an advertisement order.
Background
On-line advertising, which is sold in large part by a warranty contract, is characterized in that an advertising system guarantees exposure to targeting conditions required by advertisers, and the exposure is time-limited to complete, and may be subject to compensation if not completed beyond the time limit. In practical application, advertisers mainly concern whether the amount required by contracts can be completed, for an advertisement system, when an exposure opportunity comes, a plurality of advertisers' singletons may meet the exposure requirement, the advertisement system needs to decide which contract should be shown in the exposure, and all other contracts are guaranteed to be completed, which relates to an advertisement distribution algorithm, and currently, most widely applied advertisement distribution algorithms are a High water level HWM (High water mark) algorithm and a SHALE algorithm, wherein the HWM algorithm is a greedy heuristic algorithm applied to advertisement distribution, and the SHALE algorithm is an algorithm for solving an advertisement distribution problem by using a KKT condition.
The two algorithms are used for distributing flow for the advertisement orders, and the following defects exist:
although the HWM algorithm can determine a satisfactory distribution scheme in a short time, the HWM algorithm often cannot determine an optimal solution, i.e., cannot distribute contracts to the maximum.
Although the algorithm SHALE can obtain a better distribution scheme than the HWM algorithm, the calculation complexity of the algorithm is far higher than that of the HWM algorithm, and the solving time is long.
Disclosure of Invention
The invention provides a method and equipment for distributing flow to advertisement orders, which are used for optimizing an initial flow distribution scheme obtained by a HWM algorithm, and continuously distributing the rest advertisement orders by adjusting the proportion of various flows distributed to each order, so that the distribution scheme can be obtained in a short time, and the compensation loss is reduced.
In a first aspect, an embodiment of the present invention provides a method for allocating traffic to an advertisement order, where the method includes:
obtaining an advertisement order, and determining an initial flow distribution scheme of the advertisement order by using a high water level HWM algorithm;
and performing at least one time of optimization based on the initial flow distribution scheme, wherein the flow distribution scheme obtained by completing the optimization each time is the flow distribution scheme before the next optimization, and when performing the optimization each time, distributing at least part of flow required by meeting the order flow demand for the rest advertisement orders according to the flow distribution scheme before optimizing and the flow required by meeting the order flow demand for the rest advertisement orders which do not meet the order flow demand.
In the above embodiment, the HWM-based algorithm is a greedy heuristic applied to advertisement distribution, and the algorithm cannot distribute advertisements to the maximum extent, and at least one time of optimization is performed on an initial traffic distribution scheme of advertisement orders determined by the HWM algorithm, and at least part of traffic is distributed to remaining advertisement orders that are not distributed in the initial traffic distribution scheme, and the remaining advertisement orders are continuously distributed on the basis of the originally distributed advertisement orders, so that the number of the remaining orders is reduced to a certain extent, and meanwhile, penalty compensation caused by the fact that the advertisement orders cannot be completed is also reduced, and loss caused to both parties signing the advertisement orders when the advertisements are distributed only through the HWM algorithm can be avoided.
As an optional implementation manner, the method further includes:
determining a first default cost of the remaining advertisement orders based on the traffic distribution scheme before optimization, and determining a second default cost of the remaining advertisement orders based on the traffic distribution scheme after optimization;
and when the first default cost is determined to be larger than the second default cost, triggering to perform next optimization, otherwise, stopping the optimization.
As an alternative embodiment, determining the first default cost/the second default cost of the remaining advertisement orders based on the pre-optimization/post-optimization traffic distribution scheme comprises:
and determining the first default cost/the second default cost of the remaining advertisement orders based on the default coefficients of the remaining advertisement orders and the remaining quantity of the remaining advertisement orders determined according to the flow distribution scheme before/after optimization.
As an optional embodiment, allocating, according to a traffic allocation scheme before optimization and traffic required by remaining advertisement orders that do not satisfy the order traffic demand when satisfying the order traffic demand, at least part of traffic required when satisfying the order traffic demand for the remaining advertisement orders, includes:
determining at least one type flow which is distributed and can be used by the remaining advertisement orders according to a flow distribution scheme before optimization and the remaining advertisement orders which do not meet the demand of ordering flow;
and distributing at least part of flow required for meeting the order flow demand for the rest advertisement orders from the at least one type flow according to the distributed proportion of the at least one type flow and the flow required by the rest advertisement orders when meeting the order flow demand.
As an optional implementation manner, according to the allocated proportion of the at least one type traffic and the traffic required by the remaining advertisement orders when the order traffic demand is met, allocating, from the at least one type traffic, at least part of traffic required when the order traffic demand is met for the remaining advertisement orders, including:
and according to the distributed proportion of the at least one type flow and the flow required by the rest advertisement orders when the order flow demand is met, selecting the distributed proportion from the at least one type flow to be smaller than a set first threshold type flow, distributing at least part of flow required when the order flow demand is met for the rest advertisement orders, wherein the distributed proportion of each type flow after distribution is not larger than a set second threshold.
As an optional embodiment, after allocating at least part of traffic required to satisfy the traffic demand of the order to the remaining advertisement orders, the method further includes:
determining other orders occupying the same type of flow as the current advertisement order;
and reallocating at least part of flow for other orders from the various types of flow which are not completely allocated currently according to the remaining flow of the various types of flow which are not completely allocated currently.
As an optional implementation manner, after redistributing at least part of the flow for the other orders, the method further includes:
determining advertisement orders in various types of flow with the proportion exceeding a set high threshold;
and redistributing at least part of the flow for the determined advertisement order from the types of flow with the current distributed proportion smaller than the set low threshold.
In a second aspect, an embodiment of the present invention provides an apparatus for allocating traffic to an advertisement order, where the apparatus includes: a processor and a memory, wherein the memory stores program code that, when executed by the processor, causes the processor to perform the steps of:
obtaining an advertisement order, and determining an initial flow distribution scheme of the advertisement order by using a high water level HWM algorithm;
and performing at least one time of optimization based on the initial flow distribution scheme, wherein the flow distribution scheme obtained by completing the optimization each time is the flow distribution scheme before the next optimization, and when performing the optimization each time, distributing at least part of flow required by meeting the order flow demand for the rest advertisement orders according to the flow distribution scheme before optimizing and the flow required by meeting the order flow demand for the rest advertisement orders which do not meet the order flow demand.
As an optional implementation, the processor is further configured to:
determining a first default cost of the remaining advertisement orders based on the traffic distribution scheme before optimization, and determining a second default cost of the remaining advertisement orders based on the traffic distribution scheme after optimization;
and when the first default cost is determined to be larger than the second default cost, triggering to carry out next optimization, otherwise, stopping the optimization.
As an optional implementation manner, the processor is specifically configured to:
and determining the first default cost/the second default cost of the remaining advertisement orders based on the default coefficients of the remaining advertisement orders and the remaining quantity of the remaining advertisement orders determined according to the flow distribution scheme before/after optimization.
As an optional implementation, the processor is specifically configured to:
determining at least one type flow which is distributed and can be used by the rest advertisement orders according to the flow distribution scheme before optimization and the rest advertisement orders which do not meet the order flow demand;
and distributing at least part of flow required for meeting the order flow demand for the rest advertisement orders from the at least one type flow according to the distributed proportion of the at least one type flow and the flow required by the rest advertisement orders when meeting the order flow demand.
As an optional implementation manner, the processor is specifically configured to:
and according to the distributed proportion of the at least one type flow and the flow required by the rest advertisement orders when the order flow demand is met, selecting the distributed proportion from the at least one type flow to be smaller than a set first threshold type flow, distributing at least part of flow required when the order flow demand is met for the rest advertisement orders, wherein the distributed proportion of each type flow after distribution is not larger than a set second threshold.
As an optional implementation manner, the processor is specifically configured to:
determining other orders occupying the same type of flow as the current advertisement order;
and reallocating at least part of flow for other orders from the various types of flow which are not completely allocated currently according to the remaining flow of the various types of flow which are not completely allocated currently.
As an optional implementation manner, the processor is specifically further configured to:
determining advertisement orders in various types of flow with the current distribution proportion exceeding a set high threshold;
and redistributing at least part of the flow for the determined advertisement order from the types of flow with the current distributed proportion smaller than the set low threshold.
In a third aspect, an embodiment of the present invention further provides an apparatus for allocating traffic to an advertisement order, where the apparatus includes an initial traffic allocation scheme module and an initial traffic allocation optimization scheme module, where:
the initial flow distribution scheme module is used for acquiring the advertisement orders and determining the initial flow distribution scheme of the advertisement orders by utilizing a high water level HWM algorithm;
and the optimized initial flow distribution scheme module is used for carrying out at least one time of optimization based on the initial flow distribution scheme, and the flow distribution scheme obtained by completing the optimization each time is the flow distribution scheme before the next optimization, wherein each time of optimization, at least part of flow required by satisfying the order flow demand is distributed for the rest advertisement orders according to the flow distribution scheme before optimization and the flow required by satisfying the order flow demand by the rest advertisement orders which do not satisfy the order flow demand.
As an optional implementation, the apparatus is further configured to:
determining a first default cost of the remaining advertisement orders based on the traffic distribution scheme before optimization, and determining a second default cost of the remaining advertisement orders based on the traffic distribution scheme after optimization;
and when the first default cost is determined to be larger than the second default cost, triggering to perform next optimization, otherwise, stopping the optimization.
As an optional implementation, the apparatus is specifically configured to:
and determining the first default cost/the second default cost of the remaining advertisement orders based on the default coefficients of the remaining advertisement orders and the remaining quantity of the remaining advertisement orders determined according to the flow distribution scheme before/after optimization.
As an optional implementation manner, the module for optimizing the initial traffic distribution scheme is specifically configured to:
determining at least one type flow which is distributed and can be used by the rest advertisement orders according to the flow distribution scheme before optimization and the rest advertisement orders which do not meet the order flow demand;
and distributing at least part of flow required by meeting the order flow demand for the rest advertisement orders from the at least one type flow according to the distributed proportion of the at least one type flow and the flow required by the rest advertisement orders when meeting the order flow demand.
As an optional implementation manner, the module for optimizing the initial traffic allocation scheme is specifically configured to:
and according to the distributed proportion of the at least one type flow and the flow required by the rest advertisement orders when the order flow demand is met, selecting the distributed proportion from the at least one type flow to be smaller than a set first threshold type flow, distributing at least part of flow required when the order flow demand is met for the rest advertisement orders, wherein the distributed proportion of each type flow after distribution is not larger than a set second threshold.
As an optional implementation manner, the module for optimizing the initial traffic distribution scheme is specifically configured to:
determining other orders occupying the same type of flow as the current advertisement order;
and reallocating at least part of the flow for each other order from the types of the flow which is not completely allocated currently according to the residual flow of the types of the flow which is not completely allocated currently.
As an optional implementation manner, the module for optimizing the initial traffic distribution scheme is further specifically configured to:
determining advertisement orders in various types of flow with the proportion exceeding a set high threshold;
and redistributing at least part of the flow for the determined advertisement order from the types of flow which are distributed with the proportion smaller than the set low threshold currently.
In a fourth aspect, an embodiment of the present invention further provides a computer storage medium, on which a computer program is stored, where the computer program, when executed by a processing unit, implements the steps of the method in the first aspect.
In addition, for technical effects brought by any one implementation manner of the second aspect to the fourth aspect, reference may be made to technical effects brought by different implementation manners of the first aspect, and details are not described herein again.
Drawings
FIG. 1A is a diagram of method steps for allocating traffic to an advertisement order.
FIG. 1B is a flow chart of method steps for maximizing distribution of traffic to advertisement orders.
FIG. 2A is a flow chart of optimization of the remaining orders that are not allocated.
Fig. 2B is a flow chart of the optimization of the remaining traffic.
FIG. 3 is a block diagram of allocating traffic to an advertisement order.
Detailed Description
In order to make the objects, technical solutions and advantages of the present invention clearer, the present invention will be described in further detail with reference to the accompanying drawings, and it is apparent that the described embodiments are only a part of the embodiments of the present invention, not all of the embodiments. All other embodiments, which can be obtained by a person skilled in the art without making any creative effort based on the embodiments in the present invention, belong to the protection scope of the present invention.
Some of the words that appear in the text are explained below:
1. the term "and/or" in the embodiments of the present invention describes an association relationship of associated objects, and indicates that three relationships may exist, for example, a and/or B may indicate: a exists alone, A and B exist simultaneously, and B exists alone. The character "/" generally indicates that the former and latter associated objects are in an "or" relationship.
2. In the embodiment of the present invention, the term "HWM algorithm" is a High Water Mark (HWM) algorithm, which is a greedy heuristic algorithm applied to advertisement distribution.
3. The term "shal algorithm" in the embodiment of the present invention is a SHALE algorithm, and is an algorithm for solving an advertisement distribution problem by using a KKT condition, where the KKT (Karush-Kuhn-Tucher) condition is a necessary condition for a nonlinear programming (nonlinear programming) optimal solution in an optimization theory.
4. In the embodiment of the present invention, "advertisement order", "order" and "advertisement" have the same meaning, and for simplification of description, the "order" is collectively described in the specification.
The application scenario described in the embodiment of the present invention is for more clearly illustrating the technical solution of the embodiment of the present invention, and does not form a limitation on the technical solution provided in the embodiment of the present invention, and it can be known by a person skilled in the art that, with the occurrence of a new application scenario, the technical solution provided in the embodiment of the present invention is also applicable to similar technical problems. In the description of the present invention, the term "plurality" means two or more unless otherwise specified.
Example one
HWM Algorithm in a mathematical model of the advertisement distribution problem, as shown in the following equation, the HWM algorithm is a minimum of the following equation for the traffic distribution scheme x ij And optimizing the uniformity of the advertising order putting and whether the order quantity required by the order contract can be finished.
S j ＝∑ i∈Γ(j) s i ；θ ij ＝d j /S j ；
In the above equation, the flow distribution scheme x ij Is the proportion of flow i assigned to order j, x ij Greater than 0 and less than 1;
s i is the size of the flow i, d j Required for order jAdvertisement exposure amount, S j Is an ideal value for the order to be evenly distributed to each available flow, and Γ (j) is a flow set satisfying order j;
V j is the priority of order j, i.e. the degree of priority to which the order needs to be evenly distributed, u j Is the amount of order, p, that order j fails to allocate j And the penalty coefficient is a penalty coefficient which cannot be distributed to the order j, wherein the penalty coefficient represents a default coefficient of the advertisement order, the default cost of different orders is different, and the penalty coefficient is used for embodying in an HWM algorithm order distribution mathematical model.
In this embodiment, x obtained based on the HWM algorithm order allocation mathematical model ij As an initial flow allocation scheme, in order to correspond to the description of the flow i in the HWM algorithm, the flow is differentiated according to types, the flow i indicates that the type of the flow is i, and different orders are indicated by orders j, where i and j are positive integers, as shown in fig. 1A, the specific steps are as follows:
step 101: obtaining an advertisement order, and determining an initial flow distribution scheme of the advertisement order by using a high-water-level HWM algorithm;
in practice, the initial flow allocation scheme for an order, as determined by the HWM algorithm, actually determines x ij I.e. determining the proportion of flow i allocated to order j, wherein i is greater than zero and less than the total number of flow types, j is greater than zero and less than the total number of order types, x ij Is greater than zero and less than 1.
Step 102: and performing at least one time of optimization based on the initial flow distribution scheme, wherein the flow distribution scheme obtained by completing the optimization each time is the flow distribution scheme before the next optimization, and when performing the optimization each time, distributing at least part of flow required by meeting the order flow demand for the rest advertisement orders according to the flow distribution scheme before optimizing and the flow required by meeting the order flow demand for the rest advertisement orders which do not meet the order flow demand.
After an initial flow distribution scheme is obtained by using a HWM algorithm, the unallocated remaining orders and the unallocated remaining flow in the initial distribution scheme can be obtained at the same time, optimization is performed at least once based on the initial flow distribution scheme, if the optimization is performed only once, the number of the remaining advertisement orders which do not meet the order flow demand can be determined according to the proportion of the flow i in the initial flow distribution scheme to the order j, and the flow required by at least part of orders can be completed, the flow required by the allocated orders is not reduced, and at least part of flow required when the order flow demand is met is allocated to the remaining orders, so that the number of the remaining orders can be reduced, and the loss of compensation can be correspondingly reduced;
if the optimization is continued after one optimization, the flow distribution scheme obtained by completing the optimization each time is the flow distribution scheme before the next optimization, the number of the remaining orders can be reduced after each optimization, all orders can be completed as far as possible, and the loss of compensation is reduced to the greatest extent.
In the embodiment of the invention, the total flow required by the distributed orders is not changed, and in one case, the flow sizes of different flows used by the same distributed order are changed by adjusting the flow sizes of different flows used by the distributed order; alternatively, the remaining available flows may be allocated to the remaining orders only for the remaining orders without any adjustment to the different flows used by the allocated orders. By using any one of the above cases or a combination of the two cases, the flow can be distributed to the remaining orders, and the purpose of reducing the compensation loss can be achieved.
In order to allocate all orders to the greatest extent, the embodiment of the present invention further provides an implementable manner, which includes the following specific steps:
step 103: determining a first default cost of the remaining advertisement orders based on the traffic distribution scheme before optimization, and determining a second default cost of the remaining advertisement orders based on the traffic distribution scheme after optimization;
specifically, the default cost can obtain the quantity of the remaining advertisement orders according to the flow distribution scheme before/after optimization, and meanwhile, different default coefficients can be determined for different advertisement orders according to different contracted amounts, and based on the default coefficients of the remaining advertisement orders and the remaining quantity of the remaining advertisement orders determined according to the flow distribution scheme before/after optimization, the first default cost/the second default cost of the remaining advertisement orders can be determined, and the default cost can be represented by the following formula:
∑ j p j u j wherein u is j Is the amount of order, p, that order j fails to allocate j Is the penalty factor that order j fails to assign.
Step 104: and when the first default cost is determined to be larger than the second default cost, triggering to carry out next optimization, otherwise, stopping the optimization.
Specifically, whether the next optimization is triggered is judged according to the default cost, when the first default cost is larger than the second default cost, the default cost obtained by the current optimization is small and the default cost obtained before the optimization is shown, and the possibility of further reducing the default cost is further optimized through continuous optimization, so that the optimization is continuously performed by using the step 102, and at least part of flow required when the order flow demand is met is distributed to the rest of advertisement orders according to the flow distribution scheme before the optimization and the flow required when the order flow demand is met by the rest of advertisement orders which do not meet the order flow demand.
As shown in fig. 1B, the embodiment of step 103 and step 104 can allocate all advertisement orders to the greatest extent in cooperation with step 101 and step 102, so as to achieve the best effect and reduce the loss due to default to the greatest extent. Through step 101, step 102, step 103 and step 104, when the distribution scheme is judged not to be the optimal distribution scheme according to the default cost, different flow rates used by the distributed orders are adjusted, and the rest advertisement orders are distributed continuously until the minimum default cost is reached, so that the default cost is reduced to the maximum extent.
In addition, the initial solution obtained by the HWM algorithm is optimized according to the default cost of different contracts, the contracts can be distributed to the maximum extent while the default cost is considered, the quality of the advertisement distribution scheme obtained by the HWM is further improved, the same order distribution effect as the SHALE algorithm can be obtained, and the calculation time used when the SHALE algorithm is used for distributing the flow of the advertisement orders is shorter than that used when the SHALE algorithm is used for distributing the flow of the advertisement orders.
When optimization is performed each time, according to a flow distribution scheme before optimization and the flow required by the remaining advertisement orders which do not meet the order flow demand when the order flow demand is met, distributing at least part of flow required by the remaining advertisement orders when the order flow demand is met, and specifically implementing the following steps:
in one aspect, optimization is performed for the remaining orders that are not allocated.
Determining at least one type flow which is distributed and can be used by the remaining advertisement orders according to a flow distribution scheme before optimization and the remaining advertisement orders which do not meet the demand of ordering flow;
and distributing at least part of flow required by meeting the order flow demand for the rest advertisement orders from the at least one type flow according to the distributed proportion of the at least one type flow and the flow required by the rest advertisement orders when meeting the order flow demand.
Specifically, according to whether the traffic for allocating traffic for the remaining advertisement orders is unallocated remaining traffic or allocated traffic, there are the following two traffic allocation methods:
the method I is characterized in that the residual flow is used for distributing the flow for the residual advertisement orders;
specifically, according to the flow distribution scheme before optimization, the number of the remaining orders and the remaining flow type i which can be used by the current remaining orders are determined;
when the residual flow i is distributed by 30%, the residual flow of 70% can be distributed to the residual orders, meanwhile, the flow distributed to the residual orders by the residual flow i is related to the flow required by the residual orders when the residual orders meet the order flow demand, namely if the residual orders j need 5M flow to meet the order flow demand, but the flow i can be distributed to the residual orders j by the residual 10M flow, the final flow i is distributed to the residual orders j5M flow; similarly, if the remaining order j requires 10M traffic to satisfy the order traffic demand, but only 5M traffic remains for the traffic i to allocate the remaining order j, the final traffic i is allocated to the remaining order j5M traffic.
Distributing the flow for the rest advertisement orders by using the distributed flow k;
specifically, according to the flow distribution scheme before optimization, determining the number of remaining orders and the distributed flow type k which can be used by the current remaining order j, redistributing the distributed flow k, and redistributing the remaining flow to other orders g occupying the flow k, so that the partial flow of the distributed flow k can be redistributed to the remaining orders, wherein the size of the partial flow of the flow k which can be redistributed to the remaining orders is determined according to the minimum value of the flow kg which is distributed to the order g by the flow k and the flow ig which can be redistributed to the order g by the remaining flow k;
and the third mode is combined with the first mode and the second mode, and the flow is distributed to the rest advertisement orders by using the distributed flow and the unallocated flow.
As an optional implementation manner, according to the allocated proportion of the at least one type traffic and the traffic required by the remaining advertisement orders when the order traffic demand is met, allocating, from the at least one type traffic, at least part of traffic required when the order traffic demand is met for the remaining advertisement orders, including:
and according to the distributed proportion of the at least one type flow and the flow required by the rest advertisement orders when the order flow demand is met, selecting the distributed proportion from the at least one type flow to be smaller than a set first threshold type flow, distributing at least part of flow required when the order flow demand is met for the rest advertisement orders, wherein the distributed proportion of each type flow after distribution is not larger than a set second threshold.
Specifically, a flow type with a smaller allocated proportion of the type flow is selected from a plurality of types of flow which can meet the remaining order, namely the flow type with a larger allocated flow is the allocated flow of the remaining order; meanwhile, the proportion of the distributed flow of the type left after the flow type distributes the flow for the left order is not larger than a second threshold value, and when the second threshold value is 100%, the flow type is determined to be the flow distributed for the left order, and then no residual flow exists.
The third embodiment is shown in fig. 2A, and the flow is distributed to the unfinished ordersWherein M is the total number of flow types, N is the total number of orders, i and p represent flow types, i represents all flow types, p represents the remaining flow types, j represents any one of the remaining orders, k represents any one of the whole orders, and Γ i,j =1 or 0, wherein Γ i,j =1 means that flow i can be allocated to order j, Γ i,j =0 indicates that a flow i cannot be allocated to an order j, wherein the steps in the flow chart 2A are explained:
step 200: the first remaining order j =1 is selected.
Step 201: the first flow i =1 is selected.
Step 202: and judging whether the flow i can be distributed to the order j, if so, executing the step 206, and otherwise, executing the step 203.
Step 203: the next flow is selected, optionally the second flow i =2.
Step 204: and judging whether the flow i is less than or equal to M, if so, executing step 202, otherwise, executing step 205.
Step 205: and ending the distribution of the current remaining order and continuing to distribute the flow of the next remaining order.
Step 206: the first order k =1 is selected.
Step 207: judging whether the flow i is distributed to an order k, wherein k is not equal to j, if the flow i is distributed to the order k, and k is not equal to j, executing a step 208, otherwise, executing a step 210;
step 208: judging whether residual flow p can be distributed to the order k, if so, executing the step 209, otherwise, executing the step 210;
step 209: using the residual flow p to redistribute the flow i originally distributed to the order k, redistributing the residual flow to the partial or all flow occupied by the order k on the flow i, and simultaneously distributing the occupied partial/all flow in the flow i to the residual order j;
specifically, the residual flow p is redistributed to the partial/whole flow a occupied by the order k on the flow i, and the flow i redistributes the flow a unoccupied by the order k to the residual order j, wherein the flow a refers to the size of the flow a.
Step 210: the next order is selected, and optionally the second order k =2.
Step 211: and judging whether the order k is less than or equal to N, if so, executing step 207, and otherwise, executing step 203.
On the other hand, after the optimization is performed on the unallocated remaining orders, in order to balance the allocated flow proportions in all types of remaining flows, the embodiment further provides the following manner in cooperation with the manner of performing the optimization on the unallocated remaining orders, and performs balanced optimization on the types of flows which have been allocated and have no remaining, so that the remaining flows are generated after the flows are reallocated by the types of flows having no remaining, and thus the optimal optimization effect is achieved.
Specifically, after allocating at least part of traffic required to satisfy the flow demand of the order to the remaining advertisement orders, the method further includes:
determining other orders occupying the same type of flow with the current advertisement order;
and reallocating at least part of the flow for each other order from the types of the flow which is not completely allocated currently according to the residual flow of the types of the flow which is not completely allocated currently.
The current advertisement order is any one of all advertisement orders, and the allocated flow proportion in the residual flows of all types can be balanced to a certain extent after at least part of flow is reallocated for each other order, so that the residual flows can be reallocated to the residual orders to a certain extent.
As an optional implementation manner, after redistributing at least part of the flow for the other orders, the method further includes:
determining advertisement orders in various types of flow with the current distribution proportion exceeding a set high threshold;
and redistributing at least part of the flow for the determined advertisement order from the types of flow which are distributed with the proportion smaller than the set low threshold currently.
Specifically, when the remaining traffic is redistributed, the traffic can be adjusted for each type of distributed traffic, and the traffic which is not left after being originally distributed to the order j is redistributed according to the distributable traffic in the current remaining traffic type; distributing part or all of the current residual flow to the order j so that no residual flow is generated after the order j is originally distributed, and adjusting the flow of each type of distributed flow.
The specific implementation steps are as shown in fig. 2B, redistributing the allocated traffic, where M is the total number of traffic types, N is the total number of orders, i represents a traffic type, i represents the remaining traffic type, k represents any traffic type in all traffic types, j represents any order in all orders, Γ represents k,j =1 or 0, wherein Γ kj =1 means that flow k can be allocated to order j, Γ k,j =0 indicates that a flow k cannot be allocated to the order j, wherein the steps in the flow chart 2B are explained:
step 212: the first remaining flow i =1 is selected.
Step 213: the first order j =1 is selected.
Step 214: a determination is made as to whether flow i can be assigned to order j, if so, step 218 is performed, otherwise step 215 is performed.
Step 215: the next order is selected, optionally the second order j =2.
Step 216: and judging whether the order j is less than or equal to N, if so, executing the step 214, otherwise, executing the step 217.
Step 217: and finishing the distribution of the current residual flow and continuing to distribute the next residual flow.
Step 218: the first flow k =1 is selected.
Step 219: and judging whether the flow k is distributed to the order j and k is not equal to i, if the flow k is distributed to the order j and k is not equal to j, executing the step 220, and otherwise, executing the step 222.
Step 220: and after the flow k is distributed to the order j, judging whether the flow k has no residual flow, if so, executing a step 221, otherwise, executing a step 222.
Step 221: and redistributing the flow k originally allocated to the order j by using the residual flow i, and redistributing part or all of the flow occupied by the order j on the flow k by using the residual flow i, so that the flow k originally allocated to the order j generates the residual flow.
Step 222: the next flow is selected, and optionally the second flow k =2.
Step 223: and judging whether the flow k is less than or equal to M, if so, executing the step 219, otherwise, executing the step 215.
Example two
Based on the same inventive concept, the embodiment of the present invention further provides a device for allocating traffic to an advertisement order, and since the device is the device in the method according to the embodiment of the present invention, and the principle of the device for solving the problem is similar to that of the method, the implementation of the device may refer to the implementation in the method embodiment, and repeated details are not repeated.
The apparatus comprises: a processor and a memory, wherein the memory stores program code that, when executed by the processor, causes the processor to perform the steps of:
obtaining an advertisement order, and determining an initial flow distribution scheme of the advertisement order by using a high-water-level HWM algorithm;
and performing at least one time of optimization based on the initial flow distribution scheme, wherein the flow distribution scheme obtained by completing the optimization each time is the flow distribution scheme before the next optimization, and when performing the optimization each time, distributing at least part of flow required by meeting the order flow demand for the rest advertisement orders according to the flow distribution scheme before optimizing and the flow required by meeting the order flow demand for the rest advertisement orders which do not meet the order flow demand.
As an optional implementation, the processor is further configured to:
determining a first default cost of the remaining advertisement orders based on the traffic distribution scheme before optimization, and determining a second default cost of the remaining advertisement orders based on the traffic distribution scheme after optimization;
and when the first default cost is determined to be larger than the second default cost, triggering to perform next optimization, otherwise, stopping the optimization.
As an optional implementation, the processor is specifically configured to:
and determining the first default cost/the second default cost of the remaining advertisement orders based on the default coefficients of the remaining advertisement orders and the remaining quantity of the remaining advertisement orders determined according to the flow distribution scheme before/after optimization.
As an optional implementation manner, the processor is specifically configured to:
determining at least one type flow which is distributed and can be used by the remaining advertisement orders according to a flow distribution scheme before optimization and the remaining advertisement orders which do not meet the demand of ordering flow;
and distributing at least part of flow required by meeting the order flow demand for the rest advertisement orders from the at least one type flow according to the distributed proportion of the at least one type flow and the flow required by the rest advertisement orders when meeting the order flow demand.
As an optional implementation manner, the processor is specifically configured to:
and according to the distributed proportion of the at least one type flow and the flow required by the rest advertisement orders when the order flow demand is met, selecting the distributed proportion from the at least one type flow to be smaller than a set first threshold type flow, distributing at least part of flow required when the order flow demand is met for the rest advertisement orders, wherein the distributed proportion of each type flow after distribution is not larger than a set second threshold.
As an optional implementation, the processor is specifically configured to:
determining other orders occupying the same type of flow with the current advertisement order;
and reallocating at least part of the flow for each other order from the types of the flow which is not completely allocated currently according to the residual flow of the types of the flow which is not completely allocated currently.
As an optional implementation manner, the processor is specifically further configured to:
determining advertisement orders in various types of flow with the current distribution proportion exceeding a set high threshold;
and redistributing at least part of the flow for the determined advertisement order from the types of flow with the current distributed proportion smaller than the set low threshold.
EXAMPLE III
Based on the same inventive concept, the present invention further provides an apparatus for allocating traffic to an advertisement order, as shown in fig. 3, the apparatus includes an initial traffic allocation scheme module 301 and an optimized initial traffic allocation scheme module 302, wherein:
an initial traffic distribution scheme module 301, configured to obtain an advertisement order, and determine an initial traffic distribution scheme of the advertisement order by using a high-water level HWM algorithm;
and an optimized initial traffic distribution scheme module 302, configured to perform at least one optimization based on the initial traffic distribution scheme, where a traffic distribution scheme obtained after completing the optimization each time is a traffic distribution scheme before next optimization, and allocate at least part of traffic required for satisfying the order traffic demand to remaining advertisement orders according to the traffic distribution scheme before optimizing and traffic required for satisfying the order traffic demand for remaining advertisement orders that do not satisfy the order traffic demand each time.
As an optional implementation, the apparatus is further configured to:
determining a first default cost of the remaining advertisement orders based on the traffic distribution scheme before optimization, and determining a second default cost of the remaining advertisement orders based on the traffic distribution scheme after optimization;
and when the first default cost is determined to be larger than the second default cost, triggering to perform next optimization, otherwise, stopping the optimization.
As an optional implementation, the apparatus is specifically configured to:
and determining the first default cost/the second default cost of the remaining advertisement orders based on the default coefficients of the remaining advertisement orders and the remaining quantity of the remaining advertisement orders determined according to the flow distribution scheme before/after optimization.
As an optional implementation manner, the module for optimizing the initial traffic allocation scheme is specifically configured to:
determining at least one type flow which is distributed and can be used by the remaining advertisement orders according to a flow distribution scheme before optimization and the remaining advertisement orders which do not meet the demand of ordering flow;
and distributing at least part of flow required by meeting the order flow demand for the rest advertisement orders from the at least one type flow according to the distributed proportion of the at least one type flow and the flow required by the rest advertisement orders when meeting the order flow demand.
As an optional implementation manner, the module 302 for optimizing the initial traffic allocation scheme is specifically configured to:
and according to the distributed proportion of the at least one type flow and the flow required by the rest advertisement orders when the order flow demand is met, selecting the distributed proportion from the at least one type flow to be smaller than a set first threshold type flow, distributing at least part of flow required when the order flow demand is met for the rest advertisement orders, wherein the distributed proportion of each type flow after distribution is not larger than a set second threshold.
As an optional implementation manner, the module 302 for optimizing the initial traffic distribution scheme is specifically configured to:
determining other orders occupying the same type of flow as the current advertisement order;
and reallocating at least part of the flow for each other order from the types of the flow which is not completely allocated currently according to the residual flow of the types of the flow which is not completely allocated currently.
As an optional implementation manner, the module 302 for optimizing the initial traffic allocation scheme is further specifically configured to:
determining advertisement orders in various types of flow with the current distribution proportion exceeding a set high threshold;
and redistributing at least part of the flow for the determined advertisement order from the types of flow with the current distributed proportion smaller than the set low threshold.
Example four
Based on the same inventive concept, an embodiment of the present invention further provides a computer storage medium, on which a computer program is stored, which when executed by a processing unit implements the following steps:
obtaining an advertisement order, and determining an initial flow distribution scheme of the advertisement order by using a high-water-level HWM algorithm;
and performing at least one time of optimization based on the initial flow distribution scheme, wherein the flow distribution scheme obtained by completing the optimization each time is the flow distribution scheme before the next optimization, and when performing the optimization each time, distributing at least part of flow required by meeting the order flow demand for the rest advertisement orders according to the flow distribution scheme before optimizing and the flow required by meeting the order flow demand for the rest advertisement orders which do not meet the order flow demand.
As will be appreciated by one skilled in the art, embodiments of the present invention may be provided as a method, system, or computer program product. Accordingly, the present invention may take the form of an entirely hardware embodiment, an entirely software embodiment or an embodiment combining software and hardware aspects. Furthermore, the present invention may take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, optical storage, and the like) having computer-usable program code embodied therein.
The present invention is described with reference to flowchart illustrations and/or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It will be understood that each flow and/or block of the flowchart illustrations and/or block diagrams, and combinations of flows and/or blocks in the flowchart illustrations and/or block diagrams, can be implemented by computer program instructions. These computer program instructions may be provided to a processor of a general purpose computer, special purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, create means for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
These computer program instructions may also be stored in a computer-readable memory that can direct a computer or other programmable data processing apparatus to function in a particular manner, such that the instructions stored in the computer-readable memory produce an article of manufacture including instruction means which implement the function specified in the flowchart flow or flows and/or block diagram block or blocks.
These computer program instructions may also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
It will be apparent to those skilled in the art that various changes and modifications may be made in the present invention without departing from the spirit and scope of the invention. Thus, if such modifications and variations of the present invention fall within the scope of the claims of the present invention and their equivalents, the present invention is also intended to include such modifications and variations.
Claims (5)
1. A method of allocating traffic to an ad order, the method comprising:
obtaining an advertisement order, and determining an initial flow distribution scheme of the advertisement order by using a high-water-level HWM algorithm; wherein the HWM algorithm order is assigned with x obtained by a mathematical model ij As an initial flow distribution scheme, i represents that the type of flow is i, j represents different orders, and both i and j are positive integers;
performing at least one optimization based on the initial flow distribution scheme, wherein the flow distribution scheme obtained by completing the optimization each time is a flow distribution scheme before the next optimization, and when the optimization is performed each time, determining at least one type of flow which is distributed and can be used by the rest advertisement orders according to the flow distribution scheme before the optimization and the rest advertisement orders which do not meet the order flow demand;
according to the distributed proportion of the at least one type flow and the flow required by the rest advertisement orders when the order flow requirements are met, selecting the distributed proportion from the at least one type flow to be smaller than a set first threshold type flow, distributing at least part of flow required when the order flow requirements are met for the rest advertisement orders, and the distributed proportion of each type flow after distribution is not larger than a set second threshold;
determining a first default cost of the remaining advertisement orders based on the traffic distribution scheme before optimization, and determining a second default cost of the remaining advertisement orders based on the traffic distribution scheme after optimization;
and when the first default cost is determined to be larger than the second default cost, adjusting each type of distributed flow and triggering next optimization, otherwise, stopping the optimization.
2. The method of claim 1, wherein determining the first/second default cost for the remaining ad orders based on the pre/post-optimization traffic distribution plan comprises:
and determining the first default cost/the second default cost of the remaining advertisement orders based on the default coefficients of the remaining advertisement orders and the remaining quantity of the remaining advertisement orders determined according to the flow distribution scheme before/after optimization.
3. The method of claim 1, further comprising, after allocating at least a portion of the traffic needed to satisfy the order traffic demand for the remaining ad orders,:
determining other orders occupying the same type of flow with the current advertisement order;
and reallocating at least part of the flow for each other order from the types of the flow which is not completely allocated currently according to the residual flow of the types of the flow which is not completely allocated currently.
4. The method of claim 3, further comprising, after redistributing at least a portion of the flow for the other respective orders:
determining advertisement orders in various types of flow with the current distribution proportion exceeding a set high threshold;
and redistributing at least part of the flow for the determined advertisement order from the types of flow with the current distributed proportion smaller than the set low threshold.
5. An apparatus for allocating traffic to an advertising order, the apparatus comprising: a processor and a memory, wherein the memory stores program code that, when executed by the processor, causes the processor to perform the steps of:
obtaining an advertisement order, and determining an initial flow distribution scheme of the advertisement order by using a high-water-level HWM algorithm; wherein, the HWM algorithm order distribution mathematical model obtains x ij As an initial flow distribution scheme, i represents that the type of flow is i, j represents different orders, and i and j are positive integers;
performing at least one optimization based on the initial flow distribution scheme, wherein the flow distribution scheme obtained by completing the optimization each time is the flow distribution scheme before the next optimization, and when performing the optimization each time,
determining at least one type flow which is distributed and can be used by the remaining advertisement orders according to a flow distribution scheme before optimization and the remaining advertisement orders which do not meet the demand of ordering flow;
according to the distributed proportion of the at least one type flow and the flow required by the rest advertisement orders when the order flow requirements are met, selecting the distributed proportion from the at least one type flow to be smaller than a set first threshold type flow, distributing at least part of flow required when the order flow requirements are met for the rest advertisement orders, wherein the distributed proportion of each type flow after distribution is not larger than a set second threshold;
determining a first default cost of the remaining advertisement orders based on the traffic distribution scheme before optimization, and determining a second default cost of the remaining advertisement orders based on the traffic distribution scheme after optimization;
and when the first default cost is determined to be larger than the second default cost, adjusting each type of distributed flow and triggering next optimization, otherwise, stopping the optimization.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910015119.9A CN109829753B (en) | 2019-01-08 | 2019-01-08 | Method and equipment for distributing flow to advertisement orders |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910015119.9A CN109829753B (en) | 2019-01-08 | 2019-01-08 | Method and equipment for distributing flow to advertisement orders |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN109829753A CN109829753A (en) | 2019-05-31 |
| CN109829753B true CN109829753B (en) | 2023-04-14 |
Family
ID=66861599
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910015119.9A Active CN109829753B (en) | 2019-01-08 | 2019-01-08 | Method and equipment for distributing flow to advertisement orders |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109829753B (en) |
Families Citing this family (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111612541A (en) * | 2020-05-22 | 2020-09-01 | 贝壳技术有限公司 | Event distribution method and device, computer readable storage medium and electronic equipment |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105791157A (en) * | 2016-04-20 | 2016-07-20 | 腾讯科技（深圳）有限公司 | Flow distribution method, distribution system and server |
| CN108053265A (en) * | 2017-12-29 | 2018-05-18 | 北京奇艺世纪科技有限公司 | A kind of distribution method and system of publicity orders amount |
| CN108282418A (en) * | 2017-01-06 | 2018-07-13 | 腾讯科技（深圳）有限公司 | Media flow allocation method and device |
| CN108428144A (en) * | 2017-02-15 | 2018-08-21 | 腾讯科技（北京）有限公司 | Flow allocation method, service distribution method and the device of pushed information order |
| CN109067817A (en) * | 2018-05-31 | 2018-12-21 | 北京五八信息技术有限公司 | Media content flow distribution method, device, electronic equipment and server |
-
2019
- 2019-01-08 CN CN201910015119.9A patent/CN109829753B/en active Active
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105791157A (en) * | 2016-04-20 | 2016-07-20 | 腾讯科技（深圳）有限公司 | Flow distribution method, distribution system and server |
| CN108282418A (en) * | 2017-01-06 | 2018-07-13 | 腾讯科技（深圳）有限公司 | Media flow allocation method and device |
| CN108428144A (en) * | 2017-02-15 | 2018-08-21 | 腾讯科技（北京）有限公司 | Flow allocation method, service distribution method and the device of pushed information order |
| CN108053265A (en) * | 2017-12-29 | 2018-05-18 | 北京奇艺世纪科技有限公司 | A kind of distribution method and system of publicity orders amount |
| CN109067817A (en) * | 2018-05-31 | 2018-12-21 | 北京五八信息技术有限公司 | Media content flow distribution method, device, electronic equipment and server |
Also Published As
| Publication number | Publication date |
|---|---|
| CN109829753A (en) | 2019-05-31 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US8392928B1 (en) | Automated workload placement recommendations for a data center | |
| CN105809323A (en) | Task allocation method and system | |
| CN110351375B (en) | Data processing method and device, computer device and readable storage medium | |
| US20130290063A1 (en) | Optimizing Allocations In A Workforce Allocation Plan | |
| CN105718316A (en) | Job scheduling method and apparatus | |
| CN110866687A (en) | Task allocation method and device | |
| CN110636394B (en) | Virtual optical network mapping method, device, equipment and medium | |
| CN109976901A (en) | A kind of resource regulating method, device, server and readable storage medium storing program for executing | |
| CN110554964A (en) | Web service crowdsourcing test task allocation method based on deep reinforcement learning | |
| CN109829753B (en) | Method and equipment for distributing flow to advertisement orders | |
| CN106504012A (en) | A kind of method of adjustment of characteristic parameter, device and electronic equipment | |
| CN109840807B (en) | Traffic distribution method and equipment for advertisement orders in pre-buried scene | |
| CN113095885A (en) | Information delivery data processing method and device | |
| CN110609735A (en) | Task management method, device and computer program product | |
| CN108846635A (en) | A kind of method and system of resource bid | |
| CN106936733B (en) | Bandwidth resource allocation method and device | |
| CN105592160B (en) | Resource allocation method towards service consumer under a kind of cloud computing environment | |
| KR102312642B1 (en) | Apparatus and method for scheduling process based on machine learning | |
| KR101639765B1 (en) | Method for Adjusting and Distributing Liquidity | |
| CN117273940A (en) | Financial product combination transaction method and system based on real-time calculation | |
| CN109784687B (en) | Smart cloud manufacturing task scheduling method, readable storage medium and terminal | |
| CN109636097A (en) | A kind of distribution method and device of product design task | |
| JP7447919B2 (en) | Negotiation devices, negotiation methods and programs | |
| CN112085328B (en) | Risk assessment method, system, electronic equipment and storage medium | |
| WO2020183709A1 (en) | Automatic negotiation device, automatic negotiation method, and computer-readable recording medium |
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 | ||
| TR01 | Transfer of patent right |
Effective date of registration: 20240723 Address after: 266000 No.399, Songling Road, Laoshan District, Qingdao, Shandong Province Patentee after: QINGDAO JUKANYUN TECHNOLOGY CO.,LTD. Country or region after: China Address before: 266061 Songling Road, Laoshan District, Qingdao, Shandong Province, No. 399 Patentee before: JUHAOKAN TECHNOLOGY Co.,Ltd. Country or region before: China |
|
| TR01 | Transfer of patent right |