WO2018149371A1 - Flow and service allocation method and apparatus for push information order, and storage medium - Google Patents

Abstract

Disclosed is a flow allocation method for a push information order, comprising: acquiring information about various orders; with regard to each of the orders, determining various supply nodes conforming to directional information about the order, and determining, according to a frequency requirement of the order, the frequency ratios of the supply nodes to the order; determining, according to the frequency ratios of the supply nodes to the order, an expected flow allocation ratio of each of the supply nodes to the order; and with regard to any one of the orders, according to the expected flow allocation ratio and the frequency ratio of each of the supply nodes conforming to the directional information about the order to the order, determining a flow allocation ratio of each of the supply nodes to the order, and sending the flow allocation ratio of the supply node to each of the orders to a push server, so that the push server performs order selection processing according to the flow allocation ratio. Also disclosed are a service allocation method and a corresponding apparatus for a push information order, and a storage medium.

Description

Push information order flow, service distribution method, device and storage medium

This application claims the priority of the Chinese patent application filed on February 15, 2017, the Chinese Patent Office, the application number is 201710080661.3, and the application name is "the flow distribution method of the push information order, the service distribution method and the device". The citations are incorporated herein by reference.

Technical field

The present application relates to the field of Internet technologies, and in particular, to a traffic of an information order, a service distribution method, an apparatus, and a storage medium.

Background technique

With the development of Internet technology, more and more data (including text, pictures, audio, video, etc.) will be pushed to various users through the Internet. For example, when a user browses a webpage using a mobile device, a PC, or the like, the user receives various data pushed by the network side, such as an advertisement in a picture or video format, public welfare information, news, and the like. In this way, the user can know the news, the content of interest, and the like in time. Such data can be referred to as push information or push media content, and the like.

Technical content

The application provides a flow of information, a service distribution method, a device and a storage medium for pushing information orders, so as to solve the technical problem of how to allocate traffic for an order when the order has frequency control requirements.

The application provides a traffic distribution method for a push information order, which is applied to a traffic distribution server, and includes:

Obtaining information of each order, including the orientation information and frequency requirements of the order;

Determining, for each order, each supply node that matches the orientation information of the order, determining a frequency ratio of the each supply node to the order according to a frequency requirement of the order, the frequency ratio being represented by each supply node The percentage of traffic in the traffic that can be used for the order;

Determining a ratio of desired traffic distribution of each supply node to the order according to a frequency ratio of each of the supply nodes to the order;

For any order, determining, according to the expected traffic distribution ratio of the order and the frequency ratio of each of the supply nodes corresponding to the orientation information of the order, determining the flow distribution ratio of each supply node to the order, The supply node sends the traffic distribution ratio to each order to the push server, so that the push server performs the menu processing according to the traffic distribution ratio.

The application also provides a service allocation method for pushing an information order, which is applied to a traffic distribution server, and includes:

Receiving an information push request sent by the terminal device, where the information push request carries the user identifier;

Determining a user feature corresponding to the identifier of the user according to the identifier of the user;

Determining a supply node that matches the user feature; determining a plurality of orders that are aligned with the supply node;

Determining, by the method according to claim 1, a distribution ratio of traffic corresponding to each of the plurality of orders by the supply node;

And providing the traffic distribution ratio to the push server, so that the push server performs a menu process according to the traffic distribution ratio.

The application also provides a traffic distribution device for pushing an information order, comprising one or more processors and one or more memories, the one or more memories comprising computer readable instructions configured by the one or one The above processor executes to implement:

Obtaining information of each order, including the orientation information and frequency requirements of the order;

Determining, for each order, each supply node that matches the orientation information of the order, determining a frequency ratio of the each supply node to the order according to a frequency requirement of the order, the frequency ratio being represented by each supply node The percentage of traffic in the traffic that can be used for the order;

Determining, according to a frequency ratio of each of the supply nodes to the order, a ratio of desired traffic distribution of each supply node to the order;

For any order, determining, according to the expected traffic distribution ratio of the order and the frequency ratio of each of the supply nodes corresponding to the orientation information of the order, determining the flow distribution ratio of each supply node to the order, The supply node sends the traffic distribution ratio to each order to the push server, so that the push server performs the menu processing according to the traffic distribution ratio.

The application also provides a service distribution device for pushing an information order, comprising one or more processors and one or more memories, the one or more memories comprising computer readable instructions configured by the one or one The above processor executes to implement:

Receiving an information push request sent by the terminal device, where the information push request carries the user identifier;

Determining a user feature corresponding to the identifier of the user according to the identifier of the user;

Determining a supply node that matches the user feature;

Determining a plurality of orders that are aligned with the supply node;

Determining, by the method according to claim 1, a distribution ratio of traffic corresponding to each of the plurality of orders by the supply node;

And providing the traffic distribution ratio to the push server, so that the push server performs a menu process according to the traffic distribution ratio.

The embodiment of the present application further provides a non-transitory computer readable storage medium storing computer readable instructions, which can cause at least one processor to perform the method as described above.

DRAWINGS

In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings to be used in the embodiments or the prior art description will be briefly described below. Obviously, the drawings in the following description are only It is a certain embodiment of the present application, and other drawings can be obtained according to the drawings without any creative labor for those skilled in the art.

1A is a system architecture diagram related to an example of the present application;

1B is a detailed system architecture diagram related to an embodiment of the present application;

Figure 2 shows the relationship between the supply node and the demand node;

3 is a flow chart of a method for distributing traffic of an information order in an example of the present application;

4 is a flow chart of obtaining an expected flow distribution ratio of an order in an example of the present application;

FIG. 5 is a flowchart of determining a traffic distribution ratio in an example of the present application;

6 is a flowchart of calculating a save requirement constraint parameter in an example of the present application;

7 is a flowchart of a service allocation method for pushing an information order in an example of the present application;

8 is a flowchart of obtaining an alternative order in an example of the present application;

9 is a schematic diagram of a traffic distribution device for pushing information orders in an example of the present application;

10 is a schematic diagram of a service distribution apparatus for pushing an information order in an example of the present application;

FIG. 11 is a schematic structural diagram of a computing device in an example of the present application.

detailed description

The technical solutions in the embodiments of the present application are clearly and completely described in the following with reference to the drawings in the embodiments of the present application. It is obvious that the described embodiments are only a part of the embodiments of the present application, and not all of the embodiments. All other embodiments obtained by a person of ordinary skill in the art based on the embodiments of the present application without departing from the inventive scope are the scope of the present application.

In the Internet-based information push technology, the online push information display service (such as the display of online advertisements) can be realized by an order that guarantees the flow (ie, the number of exposures). Among them, the media party (Publisher, also known as the supplier SSP) responsible for displaying the push information guarantees a predetermined number of exposures that satisfy the targeting predicates to the demanding party (such as the advertiser Advertiser). Here, the orientation describes the target audience characteristics of the push information corresponding to the order, and an impression of the push information to the user is referred to as an exposure. In the information push phase, it is necessary to select an order to be exposed based on the current traffic estimation result and determine how much exposure is to be assigned to each order to be exposed. For example, the core issue in the CPM (Cost Per Impressions) contract advertising system is how to allocate an available ad exposure, that is, a set of orders with contractual and audience targeting constraints. Under the estimated traffic, each order is allocated advertising traffic according to its audience targeting constraints and demand, so that all orders are not lacking or the overall deficiency is the smallest, wherein no shortage refers to the number of times within the frequency control request of the order is Delivery, the overall lack of minimum means that the total deficiencies of all orders are the smallest.

FIG. 1A is a structural diagram of a system involved in an example of the present application. As shown in FIG. 1A, the system includes at least a traffic estimation module 101, a traffic distribution module 102, a push server 103, and a data module 104.

A detailed system architecture diagram is shown in FIG. 1B. The system architecture includes a push server 103, a DSP 103, and a client 105 and a traffic distribution server 107. The traffic distribution server 107 includes a traffic estimation module 101, a traffic distribution module 102, and data. Module 104, traffic distribution server 107 is used to determine the traffic distribution ratio of the order. The traffic distribution server 107, the push server 103, the DSP 106, and the client 105 are connected via the Internet. The client may be an application client on the terminal device, for example, a social application APP, a browser, or the like. The terminal device may be a portable terminal device such as a mobile phone, a tablet, a palmtop computer, a wearable device, a PC such as a desktop computer or a notebook computer, or a smart device having an Internet access capability and a display interface, such as a smart TV. .

In the Internet, each user accesses some websites through the client 105, such as browsing a webpage or watching an online video, etc., wherein a display space for displaying push information exists on the media resource displayed on the client 105, when the client 105 When the media resource is displayed, the client 105 sends a push request to the push server 103. The push request carries the identifier of the audience user and the display bit information, and the push server 103 (for example, the advertisement server for placing the Internet advertisement) receives the push request, sends the push request to the traffic distribution server 107, and the traffic distribution server 107 pushes according to the push. The user identifier in the request acquires the feature tag of the user, and matches the push request to a supply node according to the feature tag and the display bit information of the user, thereby determining one or more orders that are aligned with the supply node. The traffic distribution server 107 returns the traffic distribution ratio of the directed node for each order. The push server 103 selects an order according to the traffic distribution ratio of each order in the one or more orders, and sends the selected order to the DSP 106, and the order received by the receiving DSP 106 is sent to the client 105 of the corresponding user for display. The push server 103 also sends the push data to the data module 104 as historical exposure data, and transmits the push data to the traffic distribution module 102, so that the traffic distribution module updates the current subscription amount of each order.

The traffic estimation module 101 may estimate the traffic of each supply node according to the historical exposure data acquired from the data module 104 to obtain the estimated traffic of each supply node, and the traffic distribution module 102 according to the received exposure request from the push server 103 ( Also referred to as a push request, a supply node that is directed to match the exposure request is obtained, thereby determining an order that is aligned with the supply node. The traffic distribution module 102 further receives the estimated traffic of each of the supply nodes from the traffic estimation module 101, calculates a traffic distribution ratio of the order, and sends the traffic distribution ratio to the push server 103, and the push server 103 can The traffic distribution ratio performs a menu processing for these orders, and can provide information push services based on the menu results (for example, the process in which the advertisement server can perform subsequent advertisement push according to the menu results).

In some examples, traffic estimation module 101 and traffic distribution module 102 estimate traffic and complete traffic distribution based on an inventory model. This inventory model describes each supply node and its corresponding estimated traffic. One supply node (also called a flow unit) can be exposed under a corresponding orientation condition. The estimated flow rate is the estimated exposure amount under the orientation condition corresponding to the supply node, and various information of the supply node is determined according to the historical exposure data (can be obtained by statistics or estimated by a method). In addition, an order can also be characterized by a demand node, which corresponds to the ordering of the order and the demand quantity (ie, the order quantity of the order).

Figure 2 shows the relationship between the supply node and the demand node. Among them, there are 6 supply nodes, they have their own orientation information and inventory quantity (also called estimated traffic), and there are N demand nodes, which have their own orientation and demand quantity (also called the order traffic volume). For example, the orientation information (also called user dimension information) of the supply node 1 is {Beijing, Sports}, which represents the user who visits the sports channel from Beijing, and the corresponding estimated traffic is 8M, which means that there is such orientation information {Beijing , the number of user visits (ie, the number of exposures counted based on historical exposure data, also known as the number of exposures) is 8M. The orientation of the demand node 1 is {sports}, which represents the user who accesses the sports channel, and the corresponding subscription amount is 15M, which means that the number of times the corresponding order is required to be exposed to the user who accesses the sports channel is 15M. As shown in FIG. 2, for any pair of supply nodes and demand nodes with connections, the orientation information of the supply node is consistent with the orientation of the demand node, and it may be considered to expose the order corresponding to the demand node to the user corresponding to the supply node. It is a key issue in information push to specifically allocate the traffic of each supply node to each demand node, and to meet the orientation constraints and reservation quantity requirements of the demand nodes.

In some instances, traffic distribution is performed through a traffic distribution scheme based on an optimization model, but it does not consider frequency control and cannot be directly applied to push of information push orders with frequency requirements. In other examples, the frequency ratio model is used for traffic distribution, which is only applied to the heuristic method HWM model, and the allocation result of the traffic allocation using the frequency ratio module is not optimal.

Based on the above technical problem, the present application provides a traffic distribution method for pushing information orders, which can be applied to the traffic distribution server 107, and is specifically applicable to the traffic distribution module 102. In an example, as shown in FIG. 3, the method includes the following steps:

Step 301: Acquire information of each order, where the information includes orientation information and frequency requirements of the order.

When the demand side and the media side (also known as the supplier) reach an order, the order describes the target audience characteristics agreed by the media party and the demand side, that is, the orientation information of the order. The demand side and the media side also agree on the predetermined number of exposures for an order. For example, the order orientation is {Sports}, which means that the user who is directed to access the sports channel has a corresponding booking amount of 15M, which means that the number of times the corresponding order is required to be exposed to the user who accesses the sports channel is 15M. The order information also includes the frequency requirement for the order, which is the number of exposures required for the order within a fixed time period required by the order. In addition, the order information includes the importance of the order and the unit price of the order for the unexposure traffic.

Step 302: For each order, determine each supply node that matches the orientation information of the order, and determine the expected traffic distribution of the order for each of the supply nodes if the frequency requirement of the order is met. proportion.

When performing the above step 302, performing, for each order, determining each supply node that matches the orientation information of the order, and determining a frequency ratio of the each supply node to the order according to the frequency requirement of the order, The frequency ratio represents a proportion of the traffic available to the order in the traffic provided by each of the supply nodes; determining, according to the frequency ratio of each of the supply nodes, the expected traffic distribution ratio of each of the supply nodes to the order, wherein The desired traffic distribution ratio is used to characterize the ratio of the predetermined amount of the order to the sum of the traffic available to the order among the flows provided by each supply node. For example, when the predetermined amount of an order is 10 times, the sum of the flows that each supply node can use to place the order is 20 times, then a value of 10/20 can be used as the desired traffic distribution ratio.

For each order, each supply node that matches the orientation information for the order is determined. For example, in Figure 2, for an order oriented {sports}, the supply nodes that match the orientation information for the order are supply nodes 1, 2, 3. Calculating the ratio of the desired traffic distribution of each of the supply nodes corresponding to the order with respect to the order, the expected traffic distribution ratio being obtained by the traffic reservation amount of the order and the available traffic of the order at each supply node.

Specifically, the frequency ratio of each supply node to the order is determined according to the frequency requirement of the order, and the frequency ratio represents a proportion of the traffic available to the order among the flows provided by each supply node. For example, the frequency requirement of a fixed order time is 3 times, the estimated flow rate of a supply node matching the order orientation is 10 times, and the flow rate of the order can be used 3 times, then the supply node is for the order. The frequency ratio is 0.3. Further, according to the frequency ratio of each supply node to the order, the expected traffic distribution ratio of each supply node to the order is determined. Allocating traffic for the order, that is, the predetermined amount of the order is evenly distributed to the available traffic. However, in the actual traffic distribution, each supply and demand node also needs to meet some constraints, and the actual traffic distribution ratio may differ from the expected traffic distribution ratio. However, the deviation between the actual flow distribution ratio and the expected flow distribution ratio can be minimized. The available traffic is used to display the traffic of the order at each supply node after considering the frequency requirement of the order.

Step 303: For any order, determining, according to the expected traffic distribution ratio of the order for each of the supply nodes corresponding to the orientation information of the order, determining the frequency requirement of the order and the flow supply and demand constraints In the case, each of the matching supply nodes allocates a proportion of the traffic used by the push server menu for the order.

When performing the above step 303, performing, for any order, determining, for each of the supply nodes, the ratio of the desired traffic distribution for the order and the frequency ratio of each of the supply nodes that match the orientation information of the order The traffic distribution ratio of the order is sent to the push server by the supply node for the traffic distribution ratio of each order, so that the push server performs the menu processing according to the traffic distribution ratio.

When determining the actual flow distribution ratio of the order, it is necessary to consider the demand constraint condition, the supply constraint condition, the frequency control constraint condition and the non-negative constraint condition. Under the condition that these constraints are met, the deviation between the solution and the expected flow distribution ratio is minimized. The proportion of traffic distribution under.

The traffic distribution method of the push information order provided by the application is based on the estimated flow rate of the supply node, and the flow distribution ratio under the frequency control condition is solved, which can be applied to the push information order with the frequency control requirement, so that the frequency control condition is Order placement is more reasonable, reducing the order replenishment rate.

In some examples, the provisioning node corresponds to a user dimension or a combination of dimensions, or corresponds to a user. When the provisioning node corresponds to a user, the determined traffic distribution ratio of the order is a user's traffic distribution for the order. Proportion; when the supply node corresponds to a user dimension or a combination of dimensions, the determined traffic distribution ratio of the order is a traffic ratio of a user dimension or a combination of dimensions for the order.

When the supply node corresponds to a user dimension or a combination of dimensions, the traffic at the supply node is the traffic of all users under the same targeting condition, so that all users under the same targeting condition use the same traffic distribution ratio regardless of the number of accesses. For example, if the order frequency is limited to 3, then for a user with 10 visits, the traffic distribution ratio should be 0.3 (ie 30% exposure can be used to display the order), and for a visit 3 Or fewer users, the frequency ratio should be 1, but if the two users are in the same orientation, they will use a common traffic distribution ratio, such as 0.5, which is obviously for both users. Not suitable. When the provisioning node corresponds to one user, the traffic distribution can obtain the traffic distribution ratio based on the user level, and the traffic distribution ratio based on the user level can make the frequency control requirement in the order delivery accurate to the user.

In some examples, the traffic distribution server 107 stores the traffic predetermined amount of each order and the estimated traffic of each supply node, wherein the estimated traffic of each supply node is determined according to historical exposure data corresponding to the supply node. For example, in FIG. 2, the supply node 1 corresponds to the user who accesses sports in Beijing, and the user who determines the access history of Beijing according to the historical exposure data is 8M, and the estimated traffic of the supply node 1 is 8M. In the above step 302, when the ratio of the frequency of each of the supply nodes to the order is determined according to the frequency ratio of the each supply node to the order, as shown in FIG. 4, the following steps may be included:

Step 401: Determine, according to the frequency ratio of each supply node to the order, the available traffic provided by the each supply node for the order.

When the above step 401 is performed, the available traffic of the order is determined according to the frequency ratio of the each supply node to the order and the estimated traffic of the each supply node.

Obtaining, according to the frequency requirement of the order, a frequency ratio of each supply node that is aligned with the order for the order, the frequency ratio characterizing traffic available to the order in the traffic provided by the corresponding supply node if the frequency requirement is met Proportion.

When an order is assigned a frequency limit, only the available traffic that satisfies its orientation is only partially available. For example, if the demand for frequency control of order j is n days k times, then for users with more visits in n days, more than k times, the number of accesses within k times is the available traffic of order j. The estimated traffic of each supply node is multiplied by the corresponding frequency ratio to obtain the available traffic at each supply node, and the available traffic at each supply node is summed to obtain the available traffic of the order.

Step 402: For each supply node, determine a desired traffic distribution ratio of the supply node to the order according to the available traffic, the frequency ratio, the traffic reservation amount of the order, and the estimated traffic of the supply node.

When performing the above step 402, performing, for each supply node, determining, according to the available traffic of the order, the frequency ratio of the supply node to the order, the traffic reservation amount of the order, and the estimated traffic of the supply node, The supply node allocates a ratio of desired traffic to the order.

As described above, the desired traffic distribution ratio is that it is desirable to distribute the predetermined amount of the order evenly to the available traffic, which can be obtained by the ratio of the traffic booking amount of the order to the available traffic of the order.

In some examples, the available traffic S _j ' corresponding to the order j determines that the available traffic is by the following formula (1):

The expected flow distribution ratio t _{ij of the} supply node i for the order j is determined by the following formula (2):

Where s _i is the estimated traffic of the supply node i, f _ij is the frequency ratio of the supply node i to the order j, Γ(j) is a set of supply nodes corresponding to the orientation information of the order j, and d _j is the order j traffic booking amount.

In some examples, in the above step 303, the supply node i determines the traffic distribution ratio x _ij used for the push server menu of the order j by the model represented by the following formula (3).

That is, the actual traffic distribution ratio in the case where the actual traffic distribution ratio is deviated from the expected traffic distribution ratio and the penalty for the undistributed traffic of the order is minimized is calculated.

Wherein, the flow supply and demand constraint condition is as shown in the following formulas (4)-(6):

Demand constraints:

Supply constraints:

Frequency control constraints:

Where s _i is the estimated traffic of the supply node i, V _j is the importance of the order j, t _ij is the expected traffic distribution ratio of the supply node i for the order j, μ _j is the undelivered traffic of the order j, p _j is The penalty unit price of the undelivered traffic, J is the order set, d _j is the traffic booking amount of the order j, and f _ij is the frequency ratio of the supply node i to the order j, the frequency ratio is characterized by satisfying the frequency requirement of the order j The traffic provided by the node i is available for the traffic proportion of the order j, Γ(j) is a set of supply nodes corresponding to the orientation information of the order j, and Γ(i) is the directional information corresponding to the supply node i. A collection of each order.

In addition, the traffic distribution model can also include non-negative constraints, such as the following formulas (7) and (8):

In some examples, the frequency ratio of a supply node to an order is determined by:

When the supply node corresponds to a user, the ratio of the frequency requirement value of the order to the estimated flow rate of the supply node is used as the frequency ratio of the supply node to the order.

When the supply node corresponds to a user dimension or a combination of dimensions, the historical exposure data corresponding to the user dimension or the combination of dimensions is obtained, and the frequency ratio of the supply node to the order is determined according to the frequency requirement of the order and the historical exposure data.

In some examples, the method for distributing traffic of a push information order provided by the present application further includes the following steps:

S11: calculating a constraint parameter corresponding to the supply constraint condition of each supply node according to a desired traffic distribution ratio of each supply node for each order and a frequency ratio of each supply node to each order, and a constraint parameter corresponding to the demand constraint condition of each order;

Determining, by the each supply node, the ratio of the desired traffic distribution for the order and the frequency ratio according to the provisioning information corresponding to the orientation information of the order, determining, for each of the supply nodes, a push server menu for the order When using the traffic distribution ratio, include the steps:

S12: Determine, according to the expected traffic distribution ratio of the supply node for the order, the frequency ratio, the constraint parameter of each supply node, and the constraint parameter of the order, determine a traffic distribution ratio of each supply node to the order.

After calculating the constraint parameters of each supply node and the constraint parameters of each order, when determining the distribution ratio of a supply node to an order, according to the constraint parameters of the supply node, the constraint parameters of the order, the order The importance of the distribution and the expected traffic distribution ratio of the supply node to the order can calculate the traffic distribution ratio of the supply node to the order. The constraint parameter of the supply node is the proportion of the constraint condition of the supply node in determining the proportion of the traffic distribution; the constraint parameter of the order is the proportion of the constraint condition of the order in determining the proportion of the traffic distribution. Wherein, for calculating the constraint parameter of the supply node and the constraint parameter of the order, the initial value of the constraint parameter of the supply node or the constraint parameter of the order may be assigned, and the expected flow distribution ratio of each supply node obtained for each order and the given The initial value is substituted into the constraints of the supply node, and the constraints of the order are iteratively solved to obtain the convergence solution of the constraint parameters of the supply node and the convergence solution of the constraint parameters of the order. The specific process of calculating the constraint parameters is described in detail below.

In some examples, the distribution ratio x _{ij of the} supply node i for use in the push server menu for order j is determined by the following formula (9):

Where α _j is the constraint parameter of the order j corresponding to the demand constraint, β _i is the constraint parameter of the supply node i corresponding to the supply constraint, f _ij is the frequency ratio of the supply node i to the order j, and s _i is the supply The estimated traffic of the node, V _j is the importance of the order j, and t _ij is the expected flow distribution ratio of the supply node i to the order j.

The flow distribution ratio formula (9) of the calculated order is obtained by the above-described traffic distribution model (3) and constraints (4), (5), and (6). In particular, to make α _j represents the demand constraint parameter, also referred to as the demand for dual, β _i denotes the supply constraints, also called supply dual, γ _ij represents a non-negative constraint parameter x _ij, also called x _ij nonnegative dual , ψ _j represents a non-negative constraint parameter μ _j, also called μ _j nonnegative dual, η _ij represents the frequency control constraint parameter, also referred to as frequency-controlled dual, then according KKT (Karush-Kuhn-Tucker) conditions, x The stability condition (stationarity) corresponding to _{ij is} as shown in the following formula (10):

The complementary relaxation conditions corresponding to γ _ij and η _ij are as shown in the following formulas (11) and (12):

The dual feasible condition for γ _ij and η _ij is the following formula (13):

γ _ij ≥0, η _ij ≥0 (13)

According to the stability condition corresponding to x _ij , that is, the formula (10) can obtain the following formula (14):

Further, according to the complementary relaxation condition corresponding to η _ij , that is, the formula (12), it is obtained that either the following formula (15) is satisfied or the following formula (16) is satisfied:

x _ij =f _ij (15)

According to the frequency control constraint, formula (6) gives the formula (17):

x _ij ≤f _ij (17)

Therefore, when the right side of the formula (16) is larger than f _ij , η _ij is increased such that x _ij =f _ij , thus obtaining the formula (18):

A similar method is used to eliminate γ _{ij to} obtain the above formula (9):

make

Can get the formula (20)

x _ij =h _ij (α _j -β _i ) (20)

After obtaining the demand constraint parameter α _j and the supply constraint parameter β _i , the flow distribution ratio of the order can be calculated according to the formula (20).

The constraint condition of the demand constraint parameter α _j , that is, the complementary relaxation condition is the formula (21):

For all j, 0 _≤ α _j ≤ p _j , and, either α _j = p _j , or ∑ _{i ∈Γ (j)} s _i x _ij = d _j (21)

The constraint condition for the supply constraint parameter β _i , ie the complementary relaxation condition is equation (22):

For all i, β _i ≥ 0, and either β _i =0, or ∑ _j∈Γ(i) x _ij =1 (22)

In some examples, when the determining the traffic distribution ratio x _ij by using the formula (20), as shown in FIG. 5, the method mainly includes the following steps:

Step 501: In the offline phase, α _{j of} each order j and β _{i of} each supply node i are calculated, and each α _{j is} saved.

In the offline phase, α _j of each order j and β _{i of} each supply node i are calculated, but only α _j is saved, because the dimension of α _{j is} the same as the dimension of the order, the dimension of β _i and the supply node The dimension is the same, the dimension of the order is generally thousands or tens of thousands of dimensions, and when a supply node represents a user, the dimension of the supply node can reach hundreds of millions or even billions of dimensions, thus only saving each α _j , saving storage.

Step 502: When the push server receives the information push request from the user, read the saved α _j and calculate β _{i of} each supply node according to each read α _j .

According to the above, in order to save storage space, only the α _j is stored in the offline phase, and when the push server receives the information push request from the user, when the traffic distribution module 102 performs traffic distribution, the stored α _{j is} read, and according to the read α _j and formula (22) is calculated for each supply node β _i, i.e. 1 β _i is calculated according to the respective _{Σ j∈Γ (i) h ij (} α j -β i) =.

Step 503: Calculate each x _ij by the formula (20) according to each of the read α _j and the calculated β _i .

Each x _{ij is} calculated using equation (20) based on each of α _j and β _i .

In some examples, wherein the constraint parameters corresponding to the supply constraints are calculated according to the expected traffic distribution ratio of each supply node for each order and the frequency ratio of each supply node to each order, and each order corresponds to The constraint parameters of the demand constraint include the following steps:

Assign an initial value to α _j for all orders;

Calculate β _{i of} each supply node according to the following formula

For all i, β _i ≥ 0, and either β _i =0, or ∑ _j∈Γ(i) h _ij (α _j -β _i )=1

Calculate α _{j of} each order according to β _{i of} each supply node and the following formula

For all j, 0 _≤ α _j ≤ p _j , and, or α _j = p _j , or ∑ _{i ∈Γ (j)} s _i h _ij (α _{j -} β _i ) = d _j

By analogy, the convergence solutions of each α _j and each β _i are iteratively solved, and the convergence solution of each α _j is used as the constraint parameter of the corresponding order j, and the convergence solution of each β _i is used as the constraint parameter of the corresponding supply node i.

Specifically, when performing the calculation of α _j and β _{i of} each order j, as shown in FIG. 6 , the method mainly includes the following steps:

Step 601: Assign an initial value to α _j of all orders.

α _j and β _i is convergent, according to the complementary slackness condition while α _j in 0≤α _j ≤p _j, gives a close to zero value α _j of all orders, but also can assign a value to 0.

Step 602: Iteratively solve the convergence solution of each α _j and each β _i according to the constraint condition of α _{j and} the constraint condition of β _i .

As said above, the constraints α _j, α _j that is complementary slackness condition of equation (21):

For all j, 0 _≤ α _j ≤ p _j , and, either α _j = p _j , or ∑ _{i ∈Γ (j)} s _i x _ij = d _j (21)

constraints of β _i, β _i that is complementary slackness condition of formula (22):

For all i, β _i ≥ 0, and either β _i =0, or ∑ _{j ∈Γ (i)} x _ij =1 (22)

When iteratively solves the convergence solution of α _j and β _i :

First, according to the initialized α _j =0, each β _{i is} calculated according to the constraint condition of β _i , that is, the formula (22), specifically, according to ∑ _{j ∈Γ (i)} x _ij =1 in the formula (22), that is, ∑ _{j ∈Γ (i)} h _ij (α _{j -} β _i ) = 1 to calculate each β _i , while β _i ≥ 0 in equation (22), and possibly β _i =0, thus when solving β _i <0 Or when there is no solution, let β _i =0.

After calculating each β _i , each α _{j is} calculated according to the constraint condition of α _j , that is, the formula (21), specifically, according to the formula _{公式 i∈Γ(j)} s _i x _ij =d in the formula (21) _j , that is, ∑ _{i ∈Γ (j)} s _i h _ij (α _{j -} β _i ) = d _j to calculate each α _j . Meanwhile, in the formula (21), 0 _≤ α _j ≤ p _j , and possibly α _j = p _j , so when α _j > p _j or no solution is solved, let α _j = p _j . Thus, after iteratively solving a certain number of times, a convergence solution of each α _j is obtained.

After calculating the convergence solution of α _j and β _i , step 603 may be further included: saving the convergence solution of each α _j .

The present application also proposes a service allocation method for pushing information orders, which can be applied to the traffic distribution module 102. In an example, as shown in FIG. 7, the method includes the following steps:

Step 701: Receive an information push request sent by the terminal device.

When the foregoing step 701 is performed, the information pushing request sent by the receiving terminal device is performed, where the information pushing request carries the user identifier; and the user feature corresponding to the identifier of the user is determined according to the identifier of the user.

After receiving the exposure request (information push request) from the push server 103, the traffic distribution module 102 acquires the feature of the user according to the identifier of the user carried in the request, and obtains a supply node that is consistent with the exposure request according to the feature of the user. In addition, the exposure request also carries the display bit information, and the matching supply node can also be determined according to the user feature and the display bit information.

Step 702: Determine a supply node that matches the user feature.

When the flow distribution module 102 receives the exposure request, the supply node corresponding to the exposure request is determined according to the characteristics of the user in the exposure request. For example, in Figure 2, users accessing the sports channel in Shanghai will be matched to the provisioning node 2.

Step 703: Determine to direct a plurality of orders that match the supply node.

In a bipartite graph such as that shown in FIG. 2, a plurality of orders that are aligned with the supply node can be determined based on the orientation information of the supply node.

Step 704: The traffic distribution ratio of each of the plurality of orders determined according to the traffic distribution method of the push information order described above.

For the plurality of orders determined in step 703, the traffic distribution ratio for each order can be obtained according to the traffic distribution method shown above.

Step 705: Provide the traffic distribution ratio to the push server, so that the push server performs a menu process according to the traffic distribution ratio.

The traffic distribution ratio of each order obtained in step 704 is supplied to the push server, and the push server performs the menu processing according to the traffic distribution ratio of each order.

The service allocation method of the push information order provided by the application is based on the estimated flow rate of the supply node, and the flow distribution ratio under the frequency control condition is solved, which can be applied to the push information order with the frequency control requirement, so that the frequency control condition is made. The order placement is more reasonable, reducing the order replenishment rate.

In some instances, for an exposure (ie, a supply node i), the sum of all calculated x _ij (j ∈Γ (i)) may not be equal to one. When the sum of x _ij is less than 1, it means that a suitable order may not be selected for display, resulting in an empty order. When the sum of x _ij is greater than 1, you need to select some of the more important orders before you deliver them. To do this, you need to assign a priority to the order definition. Specifically, as shown in FIG. 8, the method mainly includes the following steps:

Step 801: When the sum of the traffic distribution ratios of the multiple orders is greater than 1, the ratio of the traffic reservation amount of each order to the available traffic is calculated.

In this example, the ratio of the predetermined amount of the order to the available flow under the frequency control condition is used as the allocation priority of the order.

Step 802: Sort the orders according to the ratio from large to small.

The corresponding orders are sorted according to the above-mentioned priority levels calculated from each order. That is, for all j∈Γ(i), follow

Sort from big to small.

Step 803: Select the first m orders in the sorted orders, so that the flow ratios of the m orders are less than or equal to 1; wherein m is an integer greater than 1. That is, ∑ _m x _ij =1.

Step 804: Provide a traffic distribution ratio of the selected m orders to the push server, so that the push server performs a menu process in the m orders according to the traffic distribution ratio of the m orders.

The present application also proposes a traffic distribution device 900 for pushing information orders, which can be applied to the traffic distribution module 102. In an example, as shown in Figure 9, the apparatus includes:

The order information obtaining unit 901 is configured to acquire information of each order, where the information includes orientation information and frequency requirements of the order.

The desired traffic distribution ratio determining unit 902 is configured to determine, for each order, each supply node that matches the orientation information of the order, and determine that each of the supply nodes is satisfied if the frequency requirement of the order is met. The expected traffic distribution ratio for this order.

The traffic distribution ratio determining unit 903 is configured to determine, for any order, the frequency requirement that the order is satisfied according to the expected traffic distribution ratio of each supply node corresponding to the orientation information of the order for the order And in the case of a flow supply and demand constraint, each of the matching supply nodes allocates a proportion of the traffic used by the push server menu for the order.

The traffic distribution device of the push information order provided by the application, based on the estimated traffic of the supply node, solves the traffic distribution ratio under the condition of frequency control, and can be applied to the push information order with the frequency control requirement, so that the frequency control condition is Order placement is more reasonable, reducing the order replenishment rate.

In some examples, the provisioning node corresponds to a user dimension or a combination of dimensions, or corresponds to a user. When the provisioning node corresponds to a user, the determined traffic distribution ratio of the order is a user's traffic distribution for the order. Proportion; when the supply node corresponds to a user dimension or a combination of dimensions, the determined traffic distribution ratio of the order is a traffic ratio of a user dimension or a combination of dimensions for the order.

In some examples, the desired traffic distribution ratio determining unit 902 is configured to:

Determining, according to a frequency requirement of the order, a frequency ratio of each supply node to the order, the frequency ratio characterizing a proportion of traffic available to the order in the traffic provided by the corresponding supply node if the frequency requirement is met;

Determining, according to the frequency ratio of each of the supply nodes to the order, the available traffic provided by the each supply node for the order;

For each of the supply nodes, a ratio of the desired traffic distribution of the supply node to the order is determined based on the available traffic, the frequency ratio, the traffic reservation amount of the order, and the estimated traffic of the supply node.

The present application also proposes a service distribution device 1000 for pushing information orders, which can be applied to the traffic distribution module 102. In an example, as shown in FIG. 10, the apparatus includes:

The information push request receiving unit 1001 is configured to receive an information push request sent by the user,

a supply node determining unit 1002, configured to determine a supply node that matches the user;

An order determining unit 1003, configured to determine a plurality of orders that are consistent with the supply node;

The traffic distribution ratio obtaining unit 1004 is configured to receive, by the traffic distribution unit, the traffic distribution method according to claim 1, to obtain a traffic distribution ratio of each of the plurality of orders;

The traffic distribution ratio providing unit 1005 is configured to provide the traffic distribution ratio to the push server, so that the push server performs a menu process according to the traffic distribution ratio.

By using the service distribution device of the push information order provided by the application, based on the estimated flow rate of the supply node, the flow distribution ratio under the frequency control condition is solved, and the push information order with the frequency control requirement can be applied to make the frequency control condition The order placement is more reasonable, reducing the order replenishment rate.

In some examples, the traffic distribution ratio obtaining unit 1004 is configured to: when the traffic ratio sum of the multiple orders is greater than 1, further comprise: calculating a ratio of the traffic booking amount of each order to the available traffic; Sorting the orders from large to small; selecting the first m orders in the sorted orders, so that the flow ratios of the m orders are equal to 1; wherein m is an integer greater than 1. The traffic distribution ratio providing unit 1005 provides the selected traffic distribution ratio of the m orders to the push server, so that the push server performs the m order according to the traffic distribution ratio of the m orders. Menu processing.

The application also provides a non-transitory computer readable storage medium storing computer readable instructions that cause at least one processor to perform the method as described above.

Figure 11 is a diagram showing the composition of a computing device in which the traffic distribution device 900 that pushes the information order and the service distribution device 1000 that pushes the information order are located. As shown in FIG. 11, the computing device includes one or more processors (CPUs) 1102, communication modules 1104, memory 1106, user interfaces 1110, and a communication bus 1108 for interconnecting these components.

The processor 1102 can receive and transmit data through the communication module 1104 to effect network communication and/or local communication.

User interface 1110 includes one or more output devices 1112 that include one or more speakers and/or one or more visual displays. User interface 1110 also includes one or more input devices 1114 including, for example, a keyboard, a mouse, a voice command input unit or loudspeaker, a touch screen display, a touch sensitive tablet, a gesture capture camera or other input button or control, and the like.

Memory 1106 can be a high speed random access memory such as DRAM, SRAM, DDR RAM, or other random access solid state storage device; or nonvolatile memory such as one or more disk storage devices, optical disk storage devices, flash memory devices, Or other non-volatile solid-state storage devices.

The memory 1106 stores a set of instructions executable by the processor 1102, and the instruction code stored in the memory is configured to be executed by the processor to implement the traffic distribution method of the push information order in the above application, and the service allocation method of the push information order. At each step, the functions of each module in the traffic distribution device of the push information order of the present application and the service distribution device of the push information order are also realized. The memory 1106 includes:

An operating system 1116, including a program for processing various basic system services and for performing hardware related tasks;

Application 1118, including various applications for traffic distribution and order service allocation, such an application can implement the processing flow in each of the above examples, such as in a traffic distribution device 900 that can include a push information order as shown in FIG. Some or all of the units or some or all of the units in the service distribution device 1000 of the push information order shown in FIG. At least one of the units 901-903 may store machine executable instructions, and at least one of the units 1001-1005 may store the machine executable instructions. The processor 1102 can implement the functions of at least one of the units 901-903 or implement the above by executing machine executable instructions in each of the units 901-903 or at least one of the units 1001-1005 in the memory 1106. The function of at least one of the units 1001-1005.

It should be noted that not all the steps and modules in the foregoing processes and the various structural diagrams are necessary, and some steps or modules may be omitted according to actual needs. The order of execution of each step is not fixed and can be adjusted as needed. The division of each module is only for the convenience of description of the functional division. In actual implementation, one module can be implemented by multiple modules, and the functions of multiple modules can also be implemented by the same module. These modules can be located in the same device. It can also be located in different devices.

The hardware modules in the embodiments may be implemented in a hardware manner or a hardware platform plus software. The above software includes machine readable instructions stored in a non-volatile storage medium. Thus, embodiments can also be embodied as software products.

In each case, the hardware may be implemented by specialized hardware or hardware that executes machine readable instructions. For example, the hardware can be a specially designed permanent circuit or logic device (such as a dedicated processor such as an FPGA or ASIC) for performing a particular operation. The hardware may also include programmable logic devices or circuits (such as including general purpose processors or other programmable processors) that are temporarily configured by software for performing particular operations.

Additionally, each instance of the present application can be implemented by a data processing program executed by a data processing device such as a computer. Obviously, the data processing program constitutes the present application. Further, a data processing program usually stored in a storage medium is executed by directly reading a program out of a storage medium or by installing or copying the program to a storage device (such as a hard disk and or a memory) of the data processing device. Therefore, such a storage medium also constitutes the present application, and the present application also provides a non-volatile storage medium in which a data processing program is stored, which can be used to execute any of the above-mentioned method examples of the present application. An example.

The machine readable instructions corresponding to the modules of FIG. 11 may cause an operating system or the like operating on a computer to perform some or all of the operations described herein. The non-transitory computer readable storage medium may be inserted into a memory provided in an expansion board within the computer or written to a memory provided in an expansion unit connected to the computer. The CPU or the like installed on the expansion board or the expansion unit can perform part and all of the actual operations according to the instructions.

The above is only the preferred embodiment of the present application, and is not intended to limit the present application. Any modifications, equivalent substitutions, improvements, etc., which are made within the spirit and principles of the present application, should be included in the present application. Within the scope of protection.

Claims (21)

1. A traffic distribution method for pushing information orders, which is applied to a traffic distribution server, and includes:

   Obtaining information of each order, including the orientation information and frequency requirements of the order;

   Determining, for each order, each supply node that matches the orientation information of the order, determining a frequency ratio of the each supply node to the order according to a frequency requirement of the order, the frequency ratio being represented by each supply node The percentage of traffic in the traffic that can be used for the order;

   Determining a ratio of desired traffic distribution of each supply node to the order according to a frequency ratio of each of the supply nodes to the order;

   For any order, determining, according to the expected traffic distribution ratio of the order and the frequency ratio of each of the supply nodes corresponding to the orientation information of the order, determining the flow distribution ratio of each supply node to the order, The supply node sends the traffic distribution ratio to each order to the push server, so that the push server performs the menu processing according to the traffic distribution ratio.
2. The method according to claim 1, wherein the supply node corresponds to a user dimension or a combination of dimensions, or corresponds to a user;

   When the supply node corresponds to a user, the determined traffic distribution ratio of the order is a user's traffic distribution ratio for the order; when the supply node corresponds to a user dimension or a combination of dimensions, the determined order is The traffic distribution ratio is the ratio of the traffic distribution of a user dimension or dimension combination to the order.
3. The method according to claim 1, wherein the traffic distribution server stores a predetermined amount of traffic of each order and an estimated traffic of each supply node;

   The determining, according to the frequency ratio of the each supply node to the order, the ratio of the expected traffic distribution of each supply node to the order includes:

   Determining the available traffic of the order according to the frequency ratio of the each supply node to the order and the estimated traffic of the each supply node;

   Determining, for each supply node, the supply node for the order based on the available traffic of the order, the frequency ratio of the supply node to the order, the traffic reservation amount of the order, and the estimated traffic of the supply node The expected ratio of traffic distribution.
4. The method according to claim 3, wherein said available flow rate S _j ' corresponding to order j is determined by the following formula (1):

   S _j '=Σ _i∈Γ(j) min{s _i ,s _i ×f _ij } (1);

   The expected flow distribution ratio t _{ij of the} supply node i for the order j is determined by the following formula (2):

   

   Where s _i is the estimated traffic of the supply node i, f _ij is the frequency ratio of the supply node i to the order j, Γ(j) is a set of supply nodes corresponding to the orientation information of the order j, and d _j is the order j traffic booking amount.
5. The method of claim 1 wherein the method further comprises:

   Calculating a constraint parameter corresponding to the supply constraint condition of each supply node and a constraint parameter corresponding to the demand constraint condition of each order according to a ratio of a desired traffic distribution of each supply node to each order and a frequency ratio of each supply node to each order;

   Wherein, according to the expected traffic distribution ratio of the supply node and the frequency ratio of each of the supply nodes corresponding to the orientation information of the order, determining, for each of the supply nodes, the use of the push server menu for the order The traffic distribution ratio includes:

   The flow allocation ratio of each supply node to the order is determined according to the expected flow distribution ratio of the supply node for the order, the frequency ratio, the constraint parameter of each supply node, and the constraint parameter of the order.
6. The method according to claim 1, wherein said determining, according to a frequency requirement of said order, said frequency ratio of said each supply node to said order comprises:

   When the supply node corresponds to one user, the ratio of the frequency requirement value of the order to the estimated flow rate of the supply node is used as the frequency ratio of the supply node to the order;

   When the supply node corresponds to a user dimension or a combination of dimensions, the historical exposure data corresponding to the user dimension or the combination of dimensions is obtained, and the frequency ratio of the supply node to the order is determined according to the frequency requirement of the order and the historical exposure data.
7. The method according to claim 5, wherein the flow distribution ratio x _ij used by the supply node i for the push server menu for the order j is determined by the following formula (3):

   

   Where α _j is the constraint parameter of the order j corresponding to the demand constraint, β _i is the constraint parameter of the supply node i corresponding to the supply constraint, f _ij is the frequency ratio of the supply node i to the order j, and s _i is the supply The estimated traffic of the node, V _j is the importance of the order j, and t _ij is the expected traffic distribution ratio of the supply node i to the order j.
8. The method according to claim 5, wherein said calculating a constraint parameter corresponding to a supply constraint condition of each supply node according to a desired traffic distribution ratio of each supply node for each order and a frequency ratio of each supply node to each order The constraint parameters corresponding to the demand constraints for each order include:

   Assign an initial value to α _j for all orders;

   Calculate β _{i of} each supply node according to the following formula (4)

   For all i, β _i ≥ 0, and either β _i =0, or Σ _{j ∈Γ (i)} h _ij (α _{j -} β _i ) = 1 (4)

   Calculate α _{j of} each order according to β _{i of} each supply node and formula (5) below

   For all j, 0 _≤ α _j ≤ p _j , and, or α _j = p _j , or Σ _{i ∈Γ (j)} s _i h _ij (α _{j -} β _i ) = d _j (5)

   By analogy, the convergence solutions of each α _j and each β _i are iteratively solved, and the convergence solution of each α _j is used as the constraint parameter of the corresponding order j, and the convergence solution of each β _i is used as the constraint parameter of the corresponding supply node i.
9. A service allocation method for pushing information orders, which is applied to a traffic distribution server, and includes:

   Receiving an information push request sent by the terminal device, where the information push request carries the user identifier;

   Determining a user feature corresponding to the identifier of the user according to the identifier of the user;

   Determining a supply node that matches the user feature; determining a plurality of orders that are aligned with the supply node;

   Determining, by the method according to claim 1, a distribution ratio of traffic corresponding to each of the plurality of orders by the supply node;

   And providing the traffic distribution ratio to the push server, so that the push server performs a menu process according to the traffic distribution ratio.
10. The method of claim 9, wherein the method further comprises: when the traffic distribution ratios of the plurality of orders are greater than one,

    Calculate the ratio of the amount of traffic booked to the available traffic for each order;

    Sorting orders according to the ratio from large to small;

    Selecting the first m orders in the sorted orders, such that the flow ratios of the m orders are less than or equal to 1; wherein m is an integer greater than 1;

    Providing the selected traffic distribution ratio of the m orders to the push server, so that the push server performs the menu processing in the m orders according to the traffic distribution ratio of the m orders.
11. A traffic distribution device for pushing an information order, comprising one or more processors and one or more memories, the one or more memories comprising computer readable instructions configured to be executed by the one or more processors achieve:

    Obtaining information of each order, including the orientation information and frequency requirements of the order;

    Determining, for each order, each supply node that matches the orientation information of the order, determining a frequency ratio of the each supply node to the order according to a frequency requirement of the order, the frequency ratio being represented by each supply node The percentage of traffic in the traffic that can be used for the order;

    Determining a ratio of desired traffic distribution of each supply node to the order according to a frequency ratio of each of the supply nodes to the order;

    For any order, determining, according to the expected traffic distribution ratio of the order and the frequency ratio of each of the supply nodes corresponding to the orientation information of the order, determining the flow distribution ratio of each supply node to the order, The supply node sends the traffic distribution ratio to each order to the push server, so that the push server performs the menu processing according to the traffic distribution ratio.
12. The apparatus according to claim 11, wherein the provisioning node corresponds to a user dimension or a combination of dimensions, or corresponds to a user;

    When the supply node corresponds to a user, the determined traffic distribution ratio of the order is a user's traffic distribution ratio for the order; when the supply node corresponds to a user dimension or a combination of dimensions, the determined order is The traffic distribution ratio is the ratio of the traffic distribution of a user dimension or dimension combination to the order.
13. The apparatus according to claim 11, wherein said traffic distribution server stores a predetermined amount of traffic for each order and an estimated traffic of each of the supply nodes; said one or more processors executing said computer readable instructions achieve:

    Determining the available traffic of the order according to the frequency ratio of the each supply node to the order and the estimated traffic of the each supply node;

    Determining, for each supply node, the supply node for the order based on the available traffic of the order, the frequency ratio of the supply node to the order, the traffic reservation amount of the order, and the estimated traffic of the supply node The expected ratio of traffic distribution.
14. The apparatus of claim 13 wherein said one or more processors execute said computer readable instructions to:

    The available flow S _j ' corresponding to the order j is determined by the following formula (6):

    S _j '=Σ _i∈Γ(j) min{s _i ,s _i ×f _ij } (6);

    The desired flow distribution ratio t _{ij of the} supply node i for the order j is determined by the following formula (7):

    

    Where s _i is the estimated traffic of the supply node i, f _ij is the frequency ratio of the supply node i to the order j, Γ(j) is a set of supply nodes corresponding to the orientation information of the order j, and d _j is the order j traffic booking amount.
15. The apparatus of claim 11 wherein said one or more processors execute said computer readable instructions to:

    Calculating a constraint parameter corresponding to the supply constraint condition of each supply node and a constraint parameter corresponding to the demand constraint condition of each order according to a ratio of a desired traffic distribution of each supply node to each order and a frequency ratio of each supply node to each order;

    The flow allocation ratio of each supply node to the order is determined according to the expected flow distribution ratio of the supply node for the order, the frequency ratio, the constraint parameter of each supply node, and the constraint parameter of the order.
16. The apparatus of claim 11 wherein said one or more processors execute said computer readable instructions to:

    When the supply node corresponds to one user, the ratio of the frequency requirement value of the order to the estimated flow rate of the supply node is used as the frequency ratio of the supply node to the order;

    When the supply node corresponds to a user dimension or a combination of dimensions, the historical exposure data corresponding to the user dimension or the combination of dimensions is obtained, and the frequency ratio of the supply node to the order is determined according to the frequency requirement of the order and the historical exposure data.
17. The apparatus of claim 15 wherein said one or more processors execute said computer readable instructions to:

    The flow distribution ratio x _ij used by the supply node i for the push server menu for the order j is determined by the following formula (8):

    

    Where α _j is the constraint parameter of the order j corresponding to the demand constraint, β _i is the constraint parameter of the supply node i corresponding to the supply constraint, f _ij is the frequency ratio of the supply node i to the order j, and s _i is the supply The estimated traffic of the node, V _j is the importance of the order j, and t _ij is the expected traffic distribution ratio of the supply node i to the order j.
18. The apparatus of claim 15 wherein said one or more processors execute said computer readable instructions to:

    Assigning an initial value to α _j of all orders; calculating β _{i of} each supply node according to the following formula (9) for all i, β _i ≥ 0, and either β _i =0, or Σ _{j ∈Γ (i)} h _{ij (α j -β i) =} 1 (9)

    Calculate α _{j of} each order based on β _{i of} each supply node and formula (10) below

    For all j, 0 _≤ α _j ≤ p _j , and, or α _j = p _j , or Σ _{i ∈Γ (j)} s _i h _ij (α _{j -} β _i ) = d _j (10)

    By analogy, the convergence solutions of each α _j and each β _i are iteratively solved, and the convergence solution of each α _j is used as the constraint parameter of the corresponding order j, and the convergence solution of each β _i is used as the constraint parameter of the corresponding supply node i.
19. A service distribution device for pushing an information order, comprising one or more processors and one or more memories, the one or more memories comprising computer readable instructions configured to be executed by the one or more processors achieve:

    Receiving an information push request sent by the terminal device, where the information push request carries the user identifier;

    Determining a user feature corresponding to the identifier of the user according to the identifier of the user;

    Determining a supply node that matches the user feature;

    Determining a plurality of orders that are aligned with the supply node;

    Determining, by the method according to claim 1, a distribution ratio of traffic corresponding to each of the plurality of orders by the supply node;

    And providing the traffic distribution ratio to the push server, so that the push server performs a menu process according to the traffic distribution ratio.
20. The apparatus of claim 19 wherein said one or more processors execute said computer readable instructions to:

    When the sum of the flow distribution ratios of the plurality of orders is greater than 1,

    Calculate the ratio of the amount of traffic booked to the available traffic for each order;

    Sorting orders according to the ratio from large to small;

    Selecting the first m orders in the sorted orders, such that the flow ratios of the m orders are less than or equal to 1; wherein m is an integer greater than 1;

    Providing the selected traffic distribution ratio of the m orders to the push server, so that the push server performs the menu processing in the m orders according to the traffic distribution ratio of the m orders.
21. A non-transitory computer readable storage medium storing computer readable instructions for causing at least one processor to perform the method of any of claims 1-10.

Images