Advertisement resource putting method and computer equipment

Abstract

The invention relates to an advertisement resource putting method and computer equipment. According to the application, when the execution result of the advertisement resource executing the corresponding putting strategy has the failure data, the suggested correction data for correcting the difference data of the failure of executing the target putting strategy by the target advertisement resource is fed back to the media end, and the confirmation information is received; and judging whether the difference data of the target advertisement resource is agreed to be modified into the suggested correction data, executing the corrected advertisement resource according to the target putting strategy when the difference data of the target advertisement resource is agreed, and executing the target advertisement resource according to the corrected putting strategy when the difference data of the target advertisement resource is not agreed. The advertisement resource modification mode and the delivery strategy modification mode are optimized according to the comparison result of the execution effect index value and the historical data, the advertisement resources and / or the delivery strategy content can be adjusted in time to improve the resource delivery efficiency, the modification mode can be optimized and adjusted, and the time consumed for confirming the correction information is shortened.

Description

Technical Field

[0001] The present application relates to the technical field of resource delivery, and in particular to an advertisement resource delivery method and computer equipment. Background Art

[0002] For resource delivery, it is necessary to classify and deliver resources according to resource demand parties in order to improve resource delivery efficiency. For example, for the delivery of advertising resources on the media side, the advertising promoter will formulate corresponding delivery strategies based on the content of the advertising resource plan, such as bids, gender and age of the delivered targets, etc. At present, advertising resources will be adjusted at any time based on the delivery purpose, and the delivery strategy will also be adjusted according to the delivery feedback effect. Therefore, the advertising resources provided by the media side to the advertising promoter may not match the delivery strategy of the advertising promoter, resulting in the failure of the advertising resource delivery demand party. For a large number of advertising resources, the specific content of the delivery strategy needs to be adjusted. At present, the business personnel of the advertising promoter need to determine the adjustment content of the delivery strategy and feedback it to different agents on the media side for different adjustments. This is not only time-consuming, but also has a high probability of error. Moreover, it is impossible to guarantee that each agent on the media side can make adjustments in time, resulting in low resource delivery efficiency. Summary of the invention

[0003] Based on this, an advertising resource delivery method and a computer device are provided to solve the technical problems that the current adjustment of advertising resources and / or delivery strategy content causes mismatches, resulting in failure of advertising resource delivery demand, and determining the adjustment content of the delivery strategy is time-consuming and error-prone, resulting in low resource delivery efficiency.

[0004] In one aspect, a method for delivering advertising resources is provided, the method comprising:

[0005] Receiving advertising resources to be delivered and delivery strategies selected corresponding to the advertising resources, adding the advertising resources to an advertising resource delivery list in order, and recording the delivery strategy corresponding to each advertising resource in the advertising resource delivery list;

[0006] Executing the corresponding delivery strategy for each advertising resource in turn according to the advertising resource delivery list, obtaining the execution result of each advertising resource, and recording the execution result in the advertising resource delivery list;

[0007] Collecting statistics on the failure data of the target advertising resources in the execution results to execute the target delivery strategy, analyzing the difference data of the failure of the target advertising resources to execute the target delivery strategy, obtaining and feeding back the suggested correction data for correcting the difference data;

[0008] receiving the confirmation information, and determining whether to agree to modify the difference data of the target advertising resource to the suggested revised data;

[0009] In response to agreeing to modify the difference data of the target advertising resource to the recommended revised data, the returned revised advertising resource is obtained, the revised advertising resource is executed according to the target delivery strategy and a first execution effect indicator value is obtained, the first execution effect indicator value is compared with a first historical execution effect indicator value of the historical advertising resource executed according to the target delivery strategy, and the advertising resource modification method is optimized according to the comparison result;

[0010] In response to disagreement with modifying the difference data of the target advertising resources into the recommended revised data, the target delivery strategy is revised according to the target advertising resources to form a revised delivery strategy, the target advertising resources are executed according to the revised delivery strategy and a second execution effect indicator value is obtained, the second execution effect indicator value is compared with the second historical execution effect indicator value of the historical target advertising resources executed according to their corresponding delivery strategies, and the delivery strategy modification method is optimized according to the comparison result.

[0011] In one embodiment, the step of executing the corresponding delivery strategy for each advertising resource in turn according to the advertising resource delivery list, obtaining the execution result of each advertising resource, and recording the execution result in the advertising resource delivery list includes:

[0012] Acquire data of each advertising resource and its corresponding delivery strategy according to the advertising resource delivery list, and determine whether the data of the advertising resource and / or the delivery strategy has changed;

[0013] In response to changes in the data of the advertising resource and / or the delivery strategy, the advertising resource executes the delivery strategy and records the execution result in the advertising resource delivery list;

[0014] In response to the fact that the data of the advertising resources and the delivery strategy have not changed, it is determined whether the advertising resources have result data after executing the delivery strategy. If the result data after executing the delivery strategy exists, the advertising resources do not execute the delivery strategy; if not, the advertising resources execute the delivery strategy, and the execution result is recorded in the advertising resource delivery list.

[0015] In one embodiment, the steps of receiving the advertising resources to be delivered and the delivery strategies selected corresponding to the advertising resources, adding the advertising resources to an advertising resource delivery list in order, and recording the delivery strategies corresponding to each advertising resource in the advertising resource delivery list include:

[0016] Receiving advertisement resources to be delivered, parsing the advertisement resources to obtain advertisement structure data, wherein the advertisement structure data includes advertisement content and advertisement delivery plan configuration data;

[0017] Converting the advertisement structure data into a data transmission unit message, and performing standardization processing on the data transmission unit message to form a standardized transmission unit message;

[0018] Adding the primary key ID of the object table in the standardized transmission unit message to the advertising resource delivery list in the order of receipt;

[0019] Acquire a delivery strategy corresponding to the advertising resource according to the advertising delivery plan configuration data in the standardized transmission unit message, and record the delivery strategy corresponding to the advertising resource in the advertising resource delivery list, wherein the delivery strategy includes advertising delivery period, bid, delivery crowd attributes and regional targeting data;

[0020] After executing the corresponding delivery strategy in response to the advertising resource, the execution times and the execution results are recorded in the advertising resource delivery list.

[0021] In one embodiment, the step of acquiring the delivery strategy corresponding to the advertising resource according to the advertising delivery plan configuration data in the standardized transmission unit message, and recording the delivery strategy corresponding to the advertising resource in the advertising resource delivery list comprises:

[0022] Obtain historical user exposure data, historical advertising resource time-sharing consumption report data, and historical user quality data;

[0023] Constructing an advertising plan effect data model of the relationship between user quality and consumption in different time periods based on the historical user exposure data, historical advertising resource time-sharing consumption report data, and historical user quality data;

[0024] Using the advertisement plan effect data model of the relationship between user quality and consumption in different time periods, obtaining a delivery strategy corresponding to the advertisement resource according to the advertisement delivery plan configuration data in the standardized transmission unit message;

[0025] A strategy name and a version number are set for the delivery strategy, and the strategy name and the version number of the delivery strategy are recorded in the advertisement resource delivery list.

[0026] In one embodiment, the step of constructing a time-division user quality and consumption relationship advertising plan effect data model according to the historical user exposure data, the historical advertising resource time-division consumption report data and the historical user quality data includes:

[0027] Using a Q-learning algorithm to construct an advertising plan effect data model of the relationship between user quality and consumption in different time periods, and training the advertising plan effect data model of the relationship between user quality and consumption in different time periods until convergence;

[0028] Acquire the advertising delivery plan configuration data in the standardized transmission unit message, use the historical user exposure data, the historical advertising resource time-sharing consumption report data and the historical user quality score data as effect data, and encode the advertising delivery plan configuration data and the effect data into a numerical vector;

[0029] The action of adjusting the delivery strategy is defined in the advertising plan effect data model of the relationship between user quality and consumption in different time periods. The action of adjusting the delivery strategy includes changing the delivery time period, increasing the bid, decreasing the bid, changing the attributes of the delivery population and the regional targeting data.

[0030] In one embodiment, the step of training the advertising plan effect data model of the relationship between user quality and consumption in different time periods until convergence includes:

[0031] Set the time-divided user quality and consumption relationship advertising plan effect data model and use the Bellman equation to update the Q value: The Q value Q(s,a) represents the expected value of the cumulative reward that may be obtained in the future after taking action a in state s, α is the learning rate, r is the immediate reward, γ is the discount factor, s is the current state, a is the current action, s′ is the next state, a′ is the next action, It is to choose the action that can maximize the Q value in the next state s′;

[0032] When training the advertising plan effect data model of the relationship between user quality and consumption in different time periods, initialize the Q value Q(s,a) to zero, observe the current state s, select an action a according to the current delivery strategy, execute action a, observe the immediate reward r and the next state s′, use the Bellman equation to update the Q value, transfer the state to s′, and repeat the above steps until the Q value converges.

[0033] In one embodiment, the step of using the advertisement plan effect data model of the relationship between user quality and consumption in different time periods to obtain the delivery strategy corresponding to the advertisement resource according to the advertisement delivery plan configuration data in the standardized transmission unit message includes:

[0034] Using the advertising plan effect data model of the relationship between user quality and consumption in different time periods, adjusting the bid of the advertising resources in real time according to the delivery strategy;

[0035] Obtain real-time user exposure data and real-time advertising resource consumption report data;

[0036] According to the real-time user exposure data and the real-time advertising resource time-sharing consumption report data, the advertising plan effect data model of the relationship between user quality and consumption in different time periods is corrected in real time;

[0037] The revised advertising plan effect data model of the relationship between user quality and consumption in different time periods is used to reform the delivery strategy corresponding to the advertising resources.

[0038] In one embodiment, the step of modifying the advertising plan effect data model of the relationship between user quality and consumption in different time periods in real time according to the real-time user exposure data and the real-time advertising resource time-sharing consumption report data comprises:

[0039] Obtaining hyperparameters of the advertising plan effect data model of the relationship between user quality and consumption in different time periods, wherein the hyperparameters include a learning rate, a discount factor, and a network structure;

[0040] The performance of the advertising plan effect data model of the relationship between user quality and consumption in different time periods is evaluated by using a test set or an online A / B test method, and the advertising plan effect data model of the relationship between user quality and consumption in different time periods is optimized by adjusting the hyperparameters.

[0041] In one embodiment, the step of optimizing the delivery strategy modification method according to the comparison result includes:

[0042] Analyze the best click-through rate, conversion rate, cost per conversion and return on investment in the comparison result, and if the second execution effect indicator value is greater than the second historical execution effect indicator value of the historical target advertising resource executed according to its corresponding delivery strategy, it is determined as positive feedback, otherwise it is determined as negative feedback;

[0043] When the probability of the positive feedback is greater than the preset probability threshold, it is set not to feedback the suggested correction data to receive confirmation information, and the advertising delivery time period, bid, delivery population attributes and regional targeting data of the target delivery strategy are automatically corrected according to the target advertising resources to form a corrected delivery strategy, and the target advertising resources are executed according to the corrected delivery strategy.

[0044] In another aspect, a computer device is provided, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the computer program, the following steps are implemented:

[0045] Receiving advertising resources to be delivered and delivery strategies selected corresponding to the advertising resources, adding the advertising resources to an advertising resource delivery list in order, and recording the delivery strategy corresponding to each advertising resource in the advertising resource delivery list;

[0046] Executing the corresponding delivery strategy for each advertising resource in turn according to the advertising resource delivery list, obtaining the execution result of each advertising resource, and recording the execution result in the advertising resource delivery list;

[0047] Collecting statistics on the failure data of the target advertising resources in the execution results to execute the target delivery strategy, analyzing the difference data of the failure of the target advertising resources to execute the target delivery strategy, obtaining and feeding back the suggested correction data for correcting the difference data;

[0048] receiving the confirmation information, and determining whether to agree to modify the difference data of the target advertising resource to the suggested revised data;

[0049] In response to agreeing to modify the difference data of the target advertising resource to the recommended revised data, the returned revised advertising resource is obtained, the revised advertising resource is executed according to the target delivery strategy and a first execution effect indicator value is obtained, the first execution effect indicator value is compared with a first historical execution effect indicator value of the historical advertising resource executed according to the target delivery strategy, and the advertising resource modification method is optimized according to the comparison result;

[0050] In response to disagreement with modifying the difference data of the target advertising resources into the recommended revised data, the target delivery strategy is revised according to the target advertising resources to form a revised delivery strategy, the target advertising resources are executed according to the revised delivery strategy and a second execution effect indicator value is obtained, the second execution effect indicator value is compared with the second historical execution effect indicator value of the historical target advertising resources executed according to their corresponding delivery strategies, and the delivery strategy modification method is optimized according to the comparison result.

[0051] The above-mentioned advertising resource delivery method and computer device, when there is failure data in the execution result of the advertising resource executing its corresponding delivery strategy, feed back suggested correction data for correcting the difference data of the failure of the target advertising resource to execute the target delivery strategy to the media end, receive confirmation information, judge whether to agree to modify the difference data of the target advertising resource to the suggested correction data, and if agreed, execute the corrected advertising resource according to the target delivery strategy, and if not agreed, execute the target advertising resource according to the corrected delivery strategy, optimize the advertising resource modification method and the delivery strategy modification method according to the comparison result of the execution effect indicator value and the historical data, not only can the advertising resources and / or delivery strategy content be adjusted in time to improve the resource delivery efficiency, but also the adjustment and modification method can be optimized to reduce the time spent on confirming the correction information. BRIEF DESCRIPTION OF THE DRAWINGS

[0052] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0053] Figure 1This is an application environment diagram of the advertising resource delivery method in one embodiment of the present application;

[0054] Figure 2 This is a logic diagram of an advertising resource delivery method in one embodiment of the present application;

[0055] Figure 3 This is a flowchart of a method for delivering advertising resources in one embodiment of the present application;

[0056] Figure 4 A logic diagram for optimizing the delivery strategy in one embodiment of the present application;

[0057] Figure 5 This is a structural block diagram of an advertising resource delivery device in one embodiment of the present application;

[0058] Figure 6 This is a diagram of the internal structure of a computer device in one embodiment of the present application. DETAILED DESCRIPTION

[0059] In order to make the purpose, technical solution and advantages of the present application more clearly understood, the present application is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and are not used to limit the present application.

[0060] The advertising resource delivery method provided in this application can be applied to Figure 1 In the application environment shown. Among them, the media terminal 101 communicates with the resource delivery terminal 102 through the network. The media terminal 101 has multiple agents, including advertising pushers such as Douyin, Toutiao, and Baidu. The media terminal 101 pushes the advertising resources to be delivered to the resource delivery terminal 102, and synchronizes the delivery strategy selected for the advertising resources. The resource delivery terminal 102 executes its corresponding delivery strategy on the advertising resources. Among them, the media terminal 101 and the resource delivery terminal 102 can be implemented with independent servers or a server cluster composed of multiple servers.

[0061] In one embodiment, Figure 2 , Figure 3 As shown, a method for delivering advertising resources is provided, and the method is applied to Figure 1 The resource delivery end 102 in the example is used as an example to illustrate, including the following steps:

[0062] Step S1, receiving advertising resources to be delivered and delivery strategies corresponding to the advertising resources, adding the advertising resources to an advertising resource delivery list in order, and recording the delivery strategy corresponding to each advertising resource in the advertising resource delivery list;

[0063] Step S2, executing the corresponding delivery strategy for each advertising resource in turn according to the advertising resource delivery list, obtaining the execution result of each advertising resource, and recording the execution result in the advertising resource delivery list;

[0064] Step S3, counting the failure data of the target advertising resources in the execution results in executing the target delivery strategy, analyzing the difference data of the target advertising resources failing to execute the target delivery strategy, obtaining the suggested correction data for correcting the difference data and feeding back the result;

[0065] Step S4, receiving confirmation information, and determining whether to agree to modify the difference data of the target advertising resource to the suggested revised data;

[0066] Step S5, in response to agreeing to modify the difference data of the target advertising resource to the suggested modified data, obtaining the returned modified advertising resource, executing the modified advertising resource according to the target delivery strategy and obtaining a first execution effect indicator value, comparing the first execution effect indicator value with a first historical execution effect indicator value of the historical advertising resource executed according to the target delivery strategy, and optimizing the advertising resource modification method according to the comparison result;

[0067] Step S6, in response to disagreement with modifying the difference data of the target advertising resources into the recommended revised data, the target delivery strategy is modified according to the target advertising resources to form a revised delivery strategy, the target advertising resources are executed according to the revised delivery strategy and a second execution effect indicator value is obtained, the second execution effect indicator value is compared with the second historical execution effect indicator value of the historical target advertising resources executed according to their corresponding delivery strategies, and the delivery strategy modification method is optimized according to the comparison result.

[0068] Specifically, when there is failure data in the execution result of the advertising resources executing their corresponding delivery strategies, suggested correction data for correcting the difference data of the failure of the target advertising resources to execute the target delivery strategy is fed back to the media end, confirmation information is received, and it is determined whether to agree to modify the difference data of the target advertising resources to the suggested correction data. If agreed, the modified advertising resources will be executed according to the target delivery strategy; if not agreed, the target advertising resources will be executed according to the corrected delivery strategy. According to the comparison result of the execution effect indicator value and the historical data, the advertising resource modification method and the delivery strategy modification method are optimized. Not only can the advertising resources and / or delivery strategy content be adjusted in time to improve the resource delivery efficiency, but also the adjustment and modification method can be optimized to reduce the time spent on confirming the correction information.

[0069] In this embodiment, the steps of executing the corresponding delivery strategy for each advertising resource in turn according to the advertising resource delivery list, obtaining the execution result of each advertising resource, and recording the execution result in the advertising resource delivery list include:

[0070] Acquire data of each advertising resource and its corresponding delivery strategy according to the advertising resource delivery list, and determine whether the data of the advertising resource and / or the delivery strategy has changed;

[0071] In response to changes in the data of the advertising resource and / or the delivery strategy, the advertising resource executes the delivery strategy and records the execution result in the advertising resource delivery list;

[0072] In response to the fact that the data of the advertising resources and the delivery strategy have not changed, it is determined whether the advertising resources have result data after executing the delivery strategy. If the result data after executing the delivery strategy exists, the advertising resources do not execute the delivery strategy; if not, the advertising resources execute the delivery strategy, and the execution result is recorded in the advertising resource delivery list.

[0073] In this embodiment, the steps of receiving the advertising resources to be delivered and the delivery strategies selected corresponding to the advertising resources, adding the advertising resources to the advertising resource delivery list in order, and recording the delivery strategies corresponding to each advertising resource in the advertising resource delivery list include:

[0074] Receiving advertisement resources to be delivered, parsing the advertisement resources to obtain advertisement structure data, wherein the advertisement structure data includes advertisement content and advertisement delivery plan configuration data;

[0075] Converting the advertisement structure data into a data transmission unit message, and performing standardization processing on the data transmission unit message to form a standardized transmission unit message;

[0076] Adding the primary key ID of the object table in the standardized transmission unit message to the advertising resource delivery list in the order of receipt;

[0077] Acquire a delivery strategy corresponding to the advertising resource according to the advertising delivery plan configuration data in the standardized transmission unit message, and record the delivery strategy corresponding to the advertising resource in the advertising resource delivery list, wherein the delivery strategy includes advertising delivery period, bid, delivery crowd attributes and regional targeting data;

[0078] After executing the corresponding delivery strategy in response to the advertising resource, the execution times and the execution results are recorded in the advertising resource delivery list.

[0079] Among them, the data transmission unit message is preferably a Kafka message, and the message in Kafka consists of two parts: a message body and a message header.

[0080] The message body is the actual data part transmitted, which can be any form of data, such as text, pictures, videos, etc. In Kafka, the message body is usually stored in the form of a byte array.

[0081] The message header provides some additional information that may be very important to consumers. For example, it can contain metadata of the message, such as timestamp, sequence number, version information, etc. This information can help consumers better understand or process the message.

[0082] In Kafka, a complete message usually contains: key, value, timestamp, partition, and offset.

[0083] Key: Optional, used to distinguish different message streams. In the same topic, messages with the same key will be sent to the same partition, which helps ensure the order of messages.

[0084] Value: The actual content of the message, which can be any form of data.

[0085] Timestamp: The time when the message was created.

[0086] Partition: The partition to which the message is assigned.

[0087] Offset: A unique identifier within a partition that identifies the location of a message within the partition.

[0088] The purpose of forming a standardized transmission unit message after standardizing the data transmission unit message is to eliminate the name differences of the configuration data of the advertising delivery plan in the advertising resources to be delivered formed by various agents, so as to achieve the uniformity of the names in the standardized transmission unit message. For example, the names for the user age attribute include the following:

[0089] Children: 0-12 years old, including infants (0-1 years old), toddlers (1-3 years old), children (6-12 years old), etc.

[0090] Adolescents: 13-19 years old, commonly referred to as teenagers or young people.

[0091] Youth: 20-39 years old. People in this stage are in the development stage both physically and mentally.

[0092] Middle age: 40-59 years old, people in this stage are usually called middle-aged people or prime-aged people.

[0093] Elderly: over 60 years old, including the elderly (60-74 years old), the octogenarians (over 75 years old), etc.

[0094] Wherein, by adding the primary key ID of the object table in the standardized transmission unit message to the advertising resource delivery list in the order of reception, it is possible to query the corresponding advertising resource in the advertising resource delivery list by the primary key ID.

[0095] In this embodiment, the step of acquiring the delivery strategy corresponding to the advertising resource according to the advertising delivery plan configuration data in the standardized transmission unit message, and recording the delivery strategy corresponding to the advertising resource in the advertising resource delivery list includes:

[0096] Obtain historical user exposure data, historical advertising resource time-sharing consumption report data, and historical user quality data;

[0097] Constructing an advertising plan effect data model of the relationship between user quality and consumption in different time periods based on the historical user exposure data, historical advertising resource time-sharing consumption report data, and historical user quality data;

[0098] Using the advertisement plan effect data model of the relationship between user quality and consumption in different time periods, obtaining a delivery strategy corresponding to the advertisement resource according to the advertisement delivery plan configuration data in the standardized transmission unit message;

[0099] A strategy name and a version number are set for the delivery strategy, and the strategy name and the version number of the delivery strategy are recorded in the advertisement resource delivery list.

[0100] In this embodiment, the step of constructing a time-division user quality and consumption relationship advertising plan effect data model according to the historical user exposure data, the historical advertising resource time-division consumption report data and the historical user quality data includes:

[0101] Using a Q-learning algorithm to construct an advertising plan effect data model of the relationship between user quality and consumption in different time periods, and training the advertising plan effect data model of the relationship between user quality and consumption in different time periods until convergence;

[0102] Acquire the advertising delivery plan configuration data in the standardized transmission unit message, use the historical user exposure data, the historical advertising resource time-sharing consumption report data and the historical user quality score data as effect data, and encode the advertising delivery plan configuration data and the effect data into a numerical vector;

[0103] The action of adjusting the delivery strategy is defined in the advertising plan effect data model of the relationship between user quality and consumption in different time periods. The action of adjusting the delivery strategy includes changing the delivery time period, increasing the bid, decreasing the bid, changing the attributes of the delivery population and the regional targeting data.

[0104] Q-learning is a model-free reinforcement learning algorithm used to solve Markov decision process (MDP) problems. Its goal is to enable the agent to learn an optimal strategy through interaction with the environment to maximize the cumulative reward. The core idea of ​​Q-learning is to evaluate the long-term value of taking an action in a certain state by continuously updating a function called Q-value. The goal of the Q-value function is to find the optimal strategy, that is, to select the action that maximizes the Q-value in each state. The core of Q-learning is to update the Q-value using the Bellman equation.

[0105] In a Markov decision process (MDP), State: a set of states of the environment, representing different situations that the agent is in. Action: a set of actions that the agent can take in each state. Reward: the immediate feedback returned by the environment after the agent performs an action. Transition Probability: the probability of transitioning to the next state after taking an action in a certain state.

[0106] In this embodiment, the step of training the advertising plan effect data model of the relationship between user quality and consumption in different time periods until convergence includes:

[0107] Set the time-divided user quality and consumption relationship advertising plan effect data model and use the Bellman equation to update the Q value: The Q value Q(s,a) represents the expected value of the cumulative reward that may be obtained in the future after taking action a in state s, α is the learning rate, r is the immediate reward, γ is the discount factor, s is the current state, a is the current action, s′ is the next state, a′ is the next action, It is to choose the action that can maximize the Q value in the next state s′;

[0108] When training the advertising plan effect data model of the relationship between user quality and consumption in different time periods, initialize the Q value Q(s,a) to zero, observe the current state s, select an action a according to the current delivery strategy, execute action a, observe the immediate reward r and the next state s′, use the Bellman equation to update the Q value, transfer the state to s′, and repeat the above steps until the Q value converges; wherein selecting an action a according to the current delivery strategy includes: setting the current delivery strategy to an ∈-greedy strategy, using the ∈-greedy strategy to select an action, that is, randomly selecting an action with a probability of ∈, and selecting the current optimal action with a probability of 1-∈.

[0109] like Figure 4 As shown, Figure 4 A logical diagram for optimizing delivery strategies.

[0110] In this embodiment, the step of using the advertisement plan effect data model of the relationship between user quality and consumption in different time periods to obtain the delivery strategy corresponding to the advertisement resource according to the advertisement delivery plan configuration data in the standardized transmission unit message includes:

[0111] Using the advertising plan effect data model of the relationship between user quality and consumption in different time periods, adjusting the bid of the advertising resources in real time according to the delivery strategy;

[0112] Obtain real-time user exposure data and real-time advertising resource consumption report data;

[0113] According to the real-time user exposure data and the real-time advertising resource time-sharing consumption report data, the advertising plan effect data model of the relationship between user quality and consumption in different time periods is corrected in real time;

[0114] The revised advertising plan effect data model of the relationship between user quality and consumption in different time periods is used to reform the delivery strategy corresponding to the advertising resources.

[0115] Verification rules are created through the advertising plan effect data model based on the relationship between user quality and consumption in different time periods. The advertising resource delivery plan of the agency is verified in real time, and the agency is assisted in making real-time adjustments, saving the agency time in back-end operations on the media side and making adjustments anytime and anywhere.

[0116] In this embodiment, the step of correcting the advertising plan effect data model of the relationship between user quality and consumption in different time periods in real time according to the real-time user exposure data and the real-time advertising resource time-sharing consumption report data includes:

[0117] Obtaining hyperparameters of the advertising plan effect data model of the relationship between user quality and consumption in different time periods, wherein the hyperparameters include a learning rate, a discount factor, and a network structure;

[0118] The performance of the advertising plan effect data model of the relationship between user quality and consumption in different time periods is evaluated by using a test set or an online A / B test method, and the advertising plan effect data model of the relationship between user quality and consumption in different time periods is optimized by adjusting the hyperparameters.

[0119] First, initialize the Q-value table (Q-learning).

[0120] Secondly, loop (each time step): Observe the state: Get the current ad placement configuration and effect data. Select action: Select action according to the current strategy (such as ∈-greedy strategy). Execute action: Adjust ad placement configuration. Observe reward and new state: Get new ad effect data and state. Store experience: Save experience (state, action, reward, new state). Update Q-value: Update Q-value using Bellman equation.

[0121] Finally, the termination condition is: reaching the predetermined number of training steps or the advertising effect reaches the expected level. Through the reinforcement learning model, the advertising delivery strategy can be dynamically adjusted to maximize the effect of advertising resource delivery.

[0122] In this embodiment, the step of optimizing the delivery strategy modification method according to the comparison result includes:

[0123] Analyze the best click-through rate, conversion rate, cost per conversion and return on investment in the comparison result, and if the second execution effect indicator value is greater than the second historical execution effect indicator value of the historical target advertising resource executed according to its corresponding delivery strategy, it is determined as positive feedback, otherwise it is determined as negative feedback;

[0124] When the probability of the positive feedback is greater than the preset probability threshold, it is set not to feedback the suggested correction data to receive confirmation information, and the advertising delivery time period, bid, delivery population attributes and regional targeting data of the target delivery strategy are automatically corrected according to the target advertising resources to form a corrected delivery strategy, and the target advertising resources are executed according to the corrected delivery strategy.

[0125] The preset probability threshold is preferably 90%. Based on positive and negative feedback, if the effect is worse after accepting the suggested revised data, the model will continue to be optimized and modified. If the effect is better after accepting the suggested revised data, it means that the suggested revised data is effective. When the probability of positive feedback reaches more than 90% execution, there is no need to confirm the modification content. Instead, the system will directly modify the parameters corresponding to the target delivery strategy, so that no one is required to automatically process it.

[0126] The method of analyzing the optimal click-through rate, conversion rate, cost per conversion and return on investment in the comparison results is as follows.

[0127] Click-through rate (CTR): CTR = Clicks / Impressions; Clicks refers to the number of clicks, and Impressions refers to the number of impressions;

[0128] Conversion rate (CVR): CVR = Conversions / ClicksCVR; Conversions is the number of conversions, ClicksCVR is the number of clicks;

[0129] Cost per conversion (CPA): CPA = Cost / Conversions; Cost is the investment cost, Conversions is the conversion volume;

[0130] Return on investment (ROI): ROI = ROI = (Revenue-Cost) / Cost, where Revenue is the return on investment and Cost is the investment cost.

[0131] In the above-mentioned advertising resource delivery method, when there is failure data in the execution result of the advertising resource executing its corresponding delivery strategy, suggested correction data for correcting the difference data of the failure of the target advertising resource to execute the target delivery strategy is fed back to the media end, confirmation information is received, and it is determined whether to agree to modify the difference data of the target advertising resource to the suggested correction data. If agreed, the modified advertising resource is executed according to the target delivery strategy; if not agreed, the target advertising resource is executed according to the modified delivery strategy, and the advertising resource modification method and the delivery strategy modification method are optimized according to the comparison result of the execution effect indicator value and the historical data. Not only can the advertising resources and / or delivery strategy content be adjusted in time to improve the resource delivery efficiency, but also the adjustment and modification method can be optimized to reduce the time spent in confirming the correction information.

[0132] It should be understood that although Figure 2-Figure 4 The steps in the flowchart are shown in sequence as indicated by the arrows, but these steps are not necessarily executed in the order indicated by the arrows. Unless otherwise specified in this document, there is no strict order restriction for the execution of these steps, and these steps can be executed in other orders. Moreover, Figure 2-Figure 4 At least part of the steps may include multiple sub-steps or multiple stages. These sub-steps or stages are not necessarily executed at the same time, but can be executed at different times. The execution order of these sub-steps or stages is not necessarily sequential, but can be executed in turn or alternately with other steps or at least part of the sub-steps or stages of other steps.

[0133] In one embodiment, Figure 5As shown, an advertising resource delivery device 10 is provided, comprising: a data receiving module 1, a delivery strategy execution statistics module 2, an invalid data correction module 3, a data correction interaction module 4, an advertisement modification optimization module 5 and a strategy modification optimization module 6.

[0134] The data receiving module 1 is used to receive the advertising resources to be delivered and the delivery strategies selected corresponding to the advertising resources, add the advertising resources to the advertising resource delivery list in order, and record the delivery strategy corresponding to each advertising resource in the advertising resource delivery list.

[0135] The delivery strategy execution statistics module 2 is used to execute the corresponding delivery strategy for each advertising resource in turn according to the advertising resource delivery list, obtain the execution result of each advertising resource, and record the execution result in the advertising resource delivery list.

[0136] The failure data correction module 3 is used to count the failure data of the target advertising resources in the execution results to execute the target delivery strategy, analyze the difference data of the target advertising resources failing to execute the target delivery strategy, obtain the suggested correction data for correcting the difference data and feedback it.

[0137] The data correction interaction module 4 is used to receive confirmation information and determine whether to agree to modify the difference data of the target advertising resource to the recommended correction data.

[0138] The advertising modification optimization module 5 is used to obtain the returned revised advertising resources in response to agreeing to modify the difference data of the target advertising resources to the recommended revised data, execute the revised advertising resources according to the target delivery strategy and obtain a first execution effect indicator value, compare the first execution effect indicator value with the first historical execution effect indicator value of the historical advertising resources executed according to the target delivery strategy, and optimize the advertising resource modification method according to the comparison result.

[0139] The strategy modification optimization module 6 is used to, in response to disagreement with modifying the difference data of the target advertising resources to the recommended revised data, modify the target delivery strategy according to the target advertising resources to form a revised delivery strategy, execute the target advertising resources according to the revised delivery strategy and obtain a second execution effect indicator value, compare the second execution effect indicator value with the second historical execution effect indicator value of the historical target advertising resources executed according to their corresponding delivery strategies, and optimize the delivery strategy modification method according to the comparison result.

[0140] In this embodiment, the steps of executing the corresponding delivery strategy for each advertising resource in turn according to the advertising resource delivery list, obtaining the execution result of each advertising resource, and recording the execution result in the advertising resource delivery list include:

[0141] Acquire data of each advertising resource and its corresponding delivery strategy according to the advertising resource delivery list, and determine whether the data of the advertising resource and / or the delivery strategy has changed;

[0142] In response to changes in the data of the advertising resource and / or the delivery strategy, the advertising resource executes the delivery strategy and records the execution result in the advertising resource delivery list;

[0143] In response to the fact that the data of the advertising resources and the delivery strategy have not changed, it is determined whether the advertising resources have result data after executing the delivery strategy. If the result data after executing the delivery strategy exists, the advertising resources do not execute the delivery strategy; if not, the advertising resources execute the delivery strategy, and the execution result is recorded in the advertising resource delivery list.

[0144] In this embodiment, the steps of receiving the advertising resources to be delivered and the delivery strategies selected corresponding to the advertising resources, adding the advertising resources to the advertising resource delivery list in order, and recording the delivery strategies corresponding to each advertising resource in the advertising resource delivery list include:

[0145] Receiving advertisement resources to be delivered, parsing the advertisement resources to obtain advertisement structure data, wherein the advertisement structure data includes advertisement content and advertisement delivery plan configuration data;

[0146] Converting the advertisement structure data into a data transmission unit message, and performing standardization processing on the data transmission unit message to form a standardized transmission unit message;

[0147] Adding the primary key ID of the object table in the standardized transmission unit message to the advertising resource delivery list in the order of receipt;

[0148] Acquire a delivery strategy corresponding to the advertising resource according to the advertising delivery plan configuration data in the standardized transmission unit message, and record the delivery strategy corresponding to the advertising resource in the advertising resource delivery list, wherein the delivery strategy includes advertising delivery period, bid, delivery crowd attributes and regional targeting data;

[0149] After executing the corresponding delivery strategy in response to the advertising resource, the execution times and the execution results are recorded in the advertising resource delivery list.

[0150] In this embodiment, the step of acquiring the delivery strategy corresponding to the advertising resource according to the advertising delivery plan configuration data in the standardized transmission unit message, and recording the delivery strategy corresponding to the advertising resource in the advertising resource delivery list includes:

[0151] Obtain historical user exposure data, historical advertising resource time-sharing consumption report data, and historical user quality data;

[0152] Constructing an advertising plan effect data model of the relationship between user quality and consumption in different time periods based on the historical user exposure data, historical advertising resource time-sharing consumption report data, and historical user quality data;

[0153] Using the advertisement plan effect data model of the relationship between user quality and consumption in different time periods, obtaining a delivery strategy corresponding to the advertisement resource according to the advertisement delivery plan configuration data in the standardized transmission unit message;

[0154] A strategy name and a version number are set for the delivery strategy, and the strategy name and the version number of the delivery strategy are recorded in the advertisement resource delivery list.

[0155] In this embodiment, the step of constructing a time-division user quality and consumption relationship advertising plan effect data model according to the historical user exposure data, the historical advertising resource time-division consumption report data and the historical user quality data includes:

[0156] Using a Q-learning algorithm to construct an advertising plan effect data model of the relationship between user quality and consumption in different time periods, and training the advertising plan effect data model of the relationship between user quality and consumption in different time periods until convergence;

[0157] Acquire the advertising delivery plan configuration data in the standardized transmission unit message, use the historical user exposure data, the historical advertising resource time-sharing consumption report data and the historical user quality score data as effect data, and encode the advertising delivery plan configuration data and the effect data into a numerical vector;

[0158] The action of adjusting the delivery strategy is defined in the advertising plan effect data model of the relationship between user quality and consumption in different time periods. The action of adjusting the delivery strategy includes changing the delivery time period, increasing the bid, decreasing the bid, changing the attributes of the delivery population and the regional targeting data.

[0159] In this embodiment, the step of training the advertising plan effect data model of the relationship between user quality and consumption in different time periods until convergence includes:

[0160] Set the time-divided user quality and consumption relationship advertising plan effect data model and use the Bellman equation to update the Q value: The Q value Q(s,a) represents the expected value of the cumulative reward that may be obtained in the future after taking action a in state s, α is the learning rate, r is the immediate reward, γ is the discount factor, s is the current state, a is the current action, s′ is the next state, a′ is the next action, It is to choose the action that can maximize the Q value in the next state s′;

[0161] When training the advertising plan effect data model of the relationship between user quality and consumption in different time periods, initialize the Q value Q(s,a) to zero, observe the current state s, select an action a according to the current delivery strategy, execute action a, observe the immediate reward r and the next state s′, use the Bellman equation to update the Q value, transfer the state to s′, and repeat the above steps until the Q value converges; wherein selecting an action a according to the current delivery strategy includes: setting the current delivery strategy to an ∈-greedy strategy, using the ∈-greedy strategy to select an action, that is, randomly selecting an action with a probability of ∈, and selecting the current optimal action with a probability of 1-∈.

[0162] In this embodiment, the step of using the advertisement plan effect data model of the relationship between user quality and consumption in different time periods to obtain the delivery strategy corresponding to the advertisement resource according to the advertisement delivery plan configuration data in the standardized transmission unit message includes:

[0163] Using the advertising plan effect data model of the relationship between user quality and consumption in different time periods, adjusting the bid of the advertising resources in real time according to the delivery strategy;

[0164] Obtain real-time user exposure data and real-time advertising resource consumption report data;

[0165] According to the real-time user exposure data and the real-time advertising resource time-sharing consumption report data, the advertising plan effect data model of the relationship between user quality and consumption in different time periods is corrected in real time;

[0166] The revised advertising plan effect data model of the relationship between user quality and consumption in different time periods is used to reform the delivery strategy corresponding to the advertising resources.

[0167] In this embodiment, the step of correcting the advertising plan effect data model of the relationship between user quality and consumption in different time periods in real time according to the real-time user exposure data and the real-time advertising resource time-sharing consumption report data includes:

[0168] Obtaining hyperparameters of the advertising plan effect data model of the relationship between user quality and consumption in different time periods, wherein the hyperparameters include a learning rate, a discount factor, and a network structure;

[0169] The performance of the advertising plan effect data model of the relationship between user quality and consumption in different time periods is evaluated by using a test set or an online A / B test method, and the advertising plan effect data model of the relationship between user quality and consumption in different time periods is optimized by adjusting the hyperparameters.

[0170] In this embodiment, the step of optimizing the delivery strategy modification method according to the comparison result includes:

[0171] Analyze the best click-through rate, conversion rate, cost per conversion and return on investment in the comparison result, and if the second execution effect indicator value is greater than the second historical execution effect indicator value of the historical target advertising resource executed according to its corresponding delivery strategy, it is determined as positive feedback, otherwise it is determined as negative feedback;

[0172] When the probability of the positive feedback is greater than the preset probability threshold, it is set not to feedback the suggested correction data to receive confirmation information, and the advertising delivery time period, bid, delivery population attributes and regional targeting data of the target delivery strategy are automatically corrected according to the target advertising resources to form a corrected delivery strategy, and the target advertising resources are executed according to the corrected delivery strategy.

[0173] In the above-mentioned advertising resource delivery device, when there is failure data in the execution result of the advertising resource executing its corresponding delivery strategy, suggested correction data for correcting the difference data of the failure of the target advertising resource to execute the target delivery strategy is fed back to the media end, confirmation information is received, and it is determined whether to agree to modify the difference data of the target advertising resource to the suggested correction data. If agreed, the modified advertising resource will be executed according to the target delivery strategy, and if not agreed, the target advertising resource will be executed according to the modified delivery strategy. The advertising resource modification method and the delivery strategy modification method are optimized according to the comparison result of the execution effect indicator value and the historical data, which can not only timely adjust the advertising resources and / or delivery strategy content to improve resource delivery efficiency, but also optimize the adjustment and modification method and reduce the time spent on confirming the correction information.

[0174] For the specific definition of the advertising resource delivery device, please refer to the definition of the advertising resource delivery method above, which will not be repeated here. Each module in the above advertising resource delivery device can be implemented in whole or in part by software, hardware and a combination thereof. The above modules can be embedded in or independent of the processor in the computer device in the form of hardware, or can be stored in the memory of the computer device in the form of software, so that the processor can call and execute the operations corresponding to the above modules.

[0175] In one embodiment, a computer program product is provided, comprising a computer program, which, when executed by a processor, implements the following steps:

[0176] Receiving advertising resources to be delivered and delivery strategies selected corresponding to the advertising resources, adding the advertising resources to an advertising resource delivery list in order, and recording the delivery strategy corresponding to each advertising resource in the advertising resource delivery list;

[0177] Executing the corresponding delivery strategy for each advertising resource in turn according to the advertising resource delivery list, obtaining the execution result of each advertising resource, and recording the execution result in the advertising resource delivery list;

[0178] Collecting statistics on the failure data of the target advertising resources in the execution results to execute the target delivery strategy, analyzing the difference data of the failure of the target advertising resources to execute the target delivery strategy, obtaining and feeding back the suggested correction data for correcting the difference data;

[0179] receiving the confirmation information, and determining whether to agree to modify the difference data of the target advertising resource to the suggested revised data;

[0180] In response to agreeing to modify the difference data of the target advertising resource to the recommended revised data, the returned revised advertising resource is obtained, the revised advertising resource is executed according to the target delivery strategy and a first execution effect indicator value is obtained, the first execution effect indicator value is compared with a first historical execution effect indicator value of the historical advertising resource executed according to the target delivery strategy, and the advertising resource modification method is optimized according to the comparison result;

[0181] In response to disagreement with modifying the difference data of the target advertising resources into the recommended revised data, the target delivery strategy is revised according to the target advertising resources to form a revised delivery strategy, the target advertising resources are executed according to the revised delivery strategy and a second execution effect indicator value is obtained, the second execution effect indicator value is compared with the second historical execution effect indicator value of the historical target advertising resources executed according to their corresponding delivery strategies, and the delivery strategy modification method is optimized according to the comparison result.

[0182] In one embodiment, the computer program further implements the following steps when executed by a processor:

[0183] The step of executing the corresponding delivery strategy for each advertising resource in turn according to the advertising resource delivery list, obtaining the execution result of each advertising resource, and recording the execution result in the advertising resource delivery list comprises:

[0184] Acquire data of each advertising resource and its corresponding delivery strategy according to the advertising resource delivery list, and determine whether the data of the advertising resource and / or the delivery strategy has changed;

[0185] In response to changes in the data of the advertising resource and / or the delivery strategy, the advertising resource executes the delivery strategy and records the execution result in the advertising resource delivery list;

[0186] In response to the fact that the data of the advertising resources and the delivery strategy have not changed, it is determined whether the advertising resources have result data after executing the delivery strategy. If the result data after executing the delivery strategy exists, the advertising resources do not execute the delivery strategy; if not, the advertising resources execute the delivery strategy, and the execution result is recorded in the advertising resource delivery list.

[0187] In one embodiment, when the computer program is executed by a processor, the following steps are also implemented:

[0188] The steps of receiving the advertisement resources to be delivered and the delivery strategies selected corresponding to the advertisement resources, adding the advertisement resources to the advertisement resource delivery list in order, and recording the delivery strategies corresponding to each advertisement resource in the advertisement resource delivery list include:

[0189] Receiving advertisement resources to be delivered, parsing the advertisement resources to obtain advertisement structure data, wherein the advertisement structure data includes advertisement content and advertisement delivery plan configuration data;

[0190] Converting the advertisement structure data into a data transmission unit message, and performing standardization processing on the data transmission unit message to form a standardized transmission unit message;

[0191] Adding the primary key ID of the object table in the standardized transmission unit message to the advertising resource delivery list in the order of receipt;

[0192] Acquire a delivery strategy corresponding to the advertising resource according to the advertising delivery plan configuration data in the standardized transmission unit message, and record the delivery strategy corresponding to the advertising resource in the advertising resource delivery list, wherein the delivery strategy includes advertising delivery period, bid, delivery crowd attributes and regional targeting data;

[0193] After executing the corresponding delivery strategy in response to the advertising resource, the execution times and the execution results are recorded in the advertising resource delivery list.

[0194] In one embodiment, when the computer program is executed by a processor, the following steps are also implemented:

[0195] The step of acquiring the delivery strategy corresponding to the advertising resource according to the advertising delivery plan configuration data in the standardized transmission unit message, and recording the delivery strategy corresponding to the advertising resource in the advertising resource delivery list comprises:

[0196] Obtain historical user exposure data, historical advertising resource time-sharing consumption report data, and historical user quality data;

[0197] Constructing an advertising plan effect data model of the relationship between user quality and consumption in different time periods based on the historical user exposure data, historical advertising resource time-sharing consumption report data, and historical user quality data;

[0198] Using the advertisement plan effect data model of the relationship between user quality and consumption in different time periods, obtaining a delivery strategy corresponding to the advertisement resource according to the advertisement delivery plan configuration data in the standardized transmission unit message;

[0199] A strategy name and a version number are set for the delivery strategy, and the strategy name and the version number of the delivery strategy are recorded in the advertisement resource delivery list.

[0200] In one embodiment, when the computer program is executed by a processor, the following steps are also implemented:

[0201] The step of constructing a time-division user quality and consumption relationship advertising plan effect data model according to the historical user exposure data, the historical advertising resource time-division consumption report data and the historical user quality data includes:

[0202] Using a Q-learning algorithm to construct an advertising plan effect data model of the relationship between user quality and consumption in different time periods, and training the advertising plan effect data model of the relationship between user quality and consumption in different time periods until convergence;

[0203] Acquire the advertising delivery plan configuration data in the standardized transmission unit message, use the historical user exposure data, the historical advertising resource time-sharing consumption report data and the historical user quality score data as effect data, and encode the advertising delivery plan configuration data and the effect data into a numerical vector;

[0204] The action of adjusting the delivery strategy is defined in the advertising plan effect data model of the relationship between user quality and consumption in different time periods. The action of adjusting the delivery strategy includes changing the delivery time period, increasing the bid, decreasing the bid, changing the attributes of the delivery population and the regional targeting data.

[0205] In one embodiment, when the computer program is executed by a processor, the following steps are also implemented:

[0206] The step of training the advertising plan effect data model of the relationship between user quality and consumption in different time periods until convergence comprises:

[0207] Set the time-divided user quality and consumption relationship advertising plan effect data model and use the Bellman equation to update the Q value: The Q value Q(s,a) represents the expected value of the cumulative reward that may be obtained in the future after taking action a in state s, α is the learning rate, r is the immediate reward, γ is the discount factor, s is the current state, a is the current action, s′ is the next state, a′ is the next action, It is to choose the action that can maximize the Q value in the next state s′;

[0208] When training the advertising plan effect data model of the relationship between user quality and consumption in different time periods, initialize the Q value Q(s,a) to zero, observe the current state s, select an action a according to the current delivery strategy, execute action a, observe the immediate reward r and the next state s′, use the Bellman equation to update the Q value, transfer the state to s′, and repeat the above steps until the Q value converges; wherein selecting an action a according to the current delivery strategy includes: setting the current delivery strategy to an ∈-greedy strategy, using the ∈-greedy strategy to select an action, that is, randomly selecting an action with a probability of ∈, and selecting the current optimal action with a probability of 1-∈.

[0209] In one embodiment, when the computer program is executed by a processor, the following steps are also implemented:

[0210] The step of using the advertisement plan effect data model of the relationship between user quality and consumption in different time periods to obtain the delivery strategy corresponding to the advertisement resource according to the advertisement delivery plan configuration data in the standardized transmission unit message includes:

[0211] Using the advertising plan effect data model of the relationship between user quality and consumption in different time periods, adjusting the bid of the advertising resources in real time according to the delivery strategy;

[0212] Obtain real-time user exposure data and real-time advertising resource consumption report data;

[0213] According to the real-time user exposure data and the real-time advertising resource time-sharing consumption report data, the advertising plan effect data model of the relationship between user quality and consumption in different time periods is corrected in real time;

[0214] The revised advertising plan effect data model of the relationship between user quality and consumption in different time periods is used to reform the delivery strategy corresponding to the advertising resources.

[0215] In one embodiment, when the computer program is executed by a processor, the following steps are also implemented:

[0216] The step of correcting the advertising plan effect data model of the relationship between user quality and consumption in different time periods in real time according to the real-time user exposure data and the real-time advertising resource time-sharing consumption report data comprises:

[0217] Obtaining hyperparameters of the advertising plan effect data model of the relationship between user quality and consumption in different time periods, wherein the hyperparameters include a learning rate, a discount factor, and a network structure;

[0218] The performance of the advertising plan effect data model of the relationship between user quality and consumption in different time periods is evaluated by using a test set or an online A / B test method, and the advertising plan effect data model of the relationship between user quality and consumption in different time periods is optimized by adjusting the hyperparameters.

[0219] In one embodiment, when the computer program is executed by a processor, the following steps are also implemented:

[0220] The step of optimizing the delivery strategy modification method according to the comparison result includes:

[0221] Analyze the best click-through rate, conversion rate, cost per conversion and return on investment in the comparison result, and if the second execution effect indicator value is greater than the second historical execution effect indicator value of the historical target advertising resource executed according to its corresponding delivery strategy, it is determined as positive feedback, otherwise it is determined as negative feedback;

[0222] When the probability of the positive feedback is greater than the preset probability threshold, it is set not to feedback the suggested correction data to receive confirmation information, and the advertising delivery time period, bid, delivery population attributes and regional targeting data of the target delivery strategy are automatically corrected according to the target advertising resources to form a corrected delivery strategy, and the target advertising resources are executed according to the corrected delivery strategy.

[0223] For the specific limitations on the steps implemented when the computer program is executed by the processor, please refer to the limitations on the method of advertising resource delivery mentioned above, which will not be repeated here.

[0224] In one embodiment, a computer device is provided. The computer device may be a resource delivery terminal, and its internal structure diagram may be as follows: Figure 6 As shown. The computer device includes a processor, a memory, a network interface and a database connected through a system bus. Among them, the processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program and a database. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The database of the computer device is used to store advertising resource delivery data. The network interface of the computer device is used to communicate with an external terminal through a network connection. When the computer program is executed by the processor, a method for delivering advertising resources is implemented.

[0225] Those skilled in the art will understand that Figure 6 The structure shown in the figure is only a block diagram of a part of the structure related to the solution of the present application, and does not constitute a limitation on the computer device to which the solution of the present application is applied. The specific computer device may include more or fewer components than those shown in the figure, or combine certain components, or have a different arrangement of components.

[0226] In one embodiment, a computer device is provided, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the following steps when executing the computer program:

[0227] Receiving advertising resources to be delivered and delivery strategies selected corresponding to the advertising resources, adding the advertising resources to an advertising resource delivery list in order, and recording the delivery strategy corresponding to each advertising resource in the advertising resource delivery list;

[0228] Executing the corresponding delivery strategy for each advertising resource in turn according to the advertising resource delivery list, obtaining the execution result of each advertising resource, and recording the execution result in the advertising resource delivery list;

[0229] Collecting statistics on the failure data of the target advertising resources in the execution results to execute the target delivery strategy, analyzing the difference data of the failure of the target advertising resources to execute the target delivery strategy, obtaining and feeding back the suggested correction data for correcting the difference data;

[0230] receiving the confirmation information, and determining whether to agree to modify the difference data of the target advertising resource to the suggested revised data;

[0231] In response to agreeing to modify the difference data of the target advertising resource to the recommended revised data, the returned revised advertising resource is obtained, the revised advertising resource is executed according to the target delivery strategy and a first execution effect indicator value is obtained, the first execution effect indicator value is compared with a first historical execution effect indicator value of the historical advertising resource executed according to the target delivery strategy, and the advertising resource modification method is optimized according to the comparison result;

[0232] In response to disagreement with modifying the difference data of the target advertising resources into the recommended revised data, the target delivery strategy is revised according to the target advertising resources to form a revised delivery strategy, the target advertising resources are executed according to the revised delivery strategy and a second execution effect indicator value is obtained, the second execution effect indicator value is compared with the second historical execution effect indicator value of the historical target advertising resources executed according to their corresponding delivery strategies, and the delivery strategy modification method is optimized according to the comparison result.

[0233] In one embodiment, when the processor executes the computer program, the processor further implements the following steps:

[0234] The step of executing the corresponding delivery strategy for each advertising resource in turn according to the advertising resource delivery list, obtaining the execution result of each advertising resource, and recording the execution result in the advertising resource delivery list comprises:

[0235] Acquire data of each advertising resource and its corresponding delivery strategy according to the advertising resource delivery list, and determine whether the data of the advertising resource and / or the delivery strategy has changed;

[0236] In response to changes in the data of the advertising resource and / or the delivery strategy, the advertising resource executes the delivery strategy and records the execution result in the advertising resource delivery list;

[0237] In response to the fact that the data of the advertising resources and the delivery strategy have not changed, it is determined whether the advertising resources have result data after executing the delivery strategy. If the result data after executing the delivery strategy exists, the advertising resources do not execute the delivery strategy; if not, the advertising resources execute the delivery strategy, and the execution result is recorded in the advertising resource delivery list.

[0238] In one embodiment, when the processor executes the computer program, the processor further implements the following steps:

[0239] The steps of receiving the advertisement resources to be delivered and the delivery strategies selected corresponding to the advertisement resources, adding the advertisement resources to the advertisement resource delivery list in order, and recording the delivery strategies corresponding to each advertisement resource in the advertisement resource delivery list include:

[0240] Receiving advertisement resources to be delivered, parsing the advertisement resources to obtain advertisement structure data, wherein the advertisement structure data includes advertisement content and advertisement delivery plan configuration data;

[0241] Converting the advertisement structure data into a data transmission unit message, and performing standardization processing on the data transmission unit message to form a standardized transmission unit message;

[0242] Adding the primary key ID of the object table in the standardized transmission unit message to the advertising resource delivery list in the order of receipt;

[0243] Acquire a delivery strategy corresponding to the advertising resource according to the advertising delivery plan configuration data in the standardized transmission unit message, and record the delivery strategy corresponding to the advertising resource in the advertising resource delivery list, wherein the delivery strategy includes advertising delivery period, bid, delivery crowd attributes and regional targeting data;

[0244] After executing the corresponding delivery strategy in response to the advertising resource, the execution times and the execution results are recorded in the advertising resource delivery list.

[0245] In one embodiment, when the processor executes the computer program, the processor further implements the following steps:

[0246] The step of acquiring the delivery strategy corresponding to the advertising resource according to the advertising delivery plan configuration data in the standardized transmission unit message, and recording the delivery strategy corresponding to the advertising resource in the advertising resource delivery list comprises:

[0247] Obtain historical user exposure data, historical advertising resource time-sharing consumption report data, and historical user quality data;

[0248] Constructing an advertising plan effect data model of the relationship between user quality and consumption in different time periods based on the historical user exposure data, historical advertising resource time-sharing consumption report data, and historical user quality data;

[0249] Using the advertisement plan effect data model of the relationship between user quality and consumption in different time periods, obtaining a delivery strategy corresponding to the advertisement resource according to the advertisement delivery plan configuration data in the standardized transmission unit message;

[0250] A strategy name and a version number are set for the delivery strategy, and the strategy name and the version number of the delivery strategy are recorded in the advertisement resource delivery list.

[0251] In one embodiment, when the processor executes the computer program, the processor further implements the following steps:

[0252] The step of constructing a time-division user quality and consumption relationship advertising plan effect data model according to the historical user exposure data, the historical advertising resource time-division consumption report data and the historical user quality data includes:

[0253] Using a Q-learning algorithm to construct an advertising plan effect data model of the relationship between user quality and consumption in different time periods, and training the advertising plan effect data model of the relationship between user quality and consumption in different time periods until convergence;

[0254] Acquire the advertising delivery plan configuration data in the standardized transmission unit message, use the historical user exposure data, the historical advertising resource time-sharing consumption report data and the historical user quality score data as effect data, and encode the advertising delivery plan configuration data and the effect data into a numerical vector;

[0255] The action of adjusting the delivery strategy is defined in the advertising plan effect data model of the relationship between user quality and consumption in different time periods. The action of adjusting the delivery strategy includes changing the delivery time period, increasing the bid, decreasing the bid, changing the attributes of the delivery population and the regional targeting data.

[0256] In one embodiment, when the processor executes the computer program, the processor further implements the following steps:

[0257] The step of training the advertising plan effect data model of the relationship between user quality and consumption in different time periods until convergence comprises:

[0258] Set the time-divided user quality and consumption relationship advertising plan effect data model and use the Bellman equation to update the Q value: The Q value Q(s,a) represents the expected value of the cumulative reward that may be obtained in the future after taking action a in state s, α is the learning rate, r is the immediate reward, γ is the discount factor, s is the current state, a is the current action, s′ is the next state, a′ is the next action, It is to choose the action that can maximize the Q value in the next state s′;

[0259] When training the advertising plan effect data model of the relationship between user quality and consumption in different time periods, initialize the Q value Q(s,a) to zero, observe the current state s, select an action a according to the current delivery strategy, execute action a, observe the immediate reward r and the next state s′, use the Bellman equation to update the Q value, transfer the state to s′, and repeat the above steps until the Q value converges; wherein selecting an action a according to the current delivery strategy includes: setting the current delivery strategy to an ∈-greedy strategy, using the ∈-greedy strategy to select an action, that is, randomly selecting an action with a probability of ∈, and selecting the current optimal action with a probability of 1-∈.

[0260] In one embodiment, when the processor executes the computer program, the processor further implements the following steps:

[0261] The step of using the advertisement plan effect data model of the relationship between user quality and consumption in different time periods to obtain the delivery strategy corresponding to the advertisement resource according to the advertisement delivery plan configuration data in the standardized transmission unit message includes:

[0262] Using the advertising plan effect data model of the relationship between user quality and consumption in different time periods, adjusting the bid of the advertising resources in real time according to the delivery strategy;

[0263] Obtain real-time user exposure data and real-time advertising resource consumption report data;

[0264] According to the real-time user exposure data and the real-time advertising resource time-sharing consumption report data, the advertising plan effect data model of the relationship between user quality and consumption in different time periods is corrected in real time;

[0265] The revised advertising plan effect data model of the relationship between user quality and consumption in different time periods is used to reform the delivery strategy corresponding to the advertising resources.

[0266] In one embodiment, when the processor executes the computer program, the processor further implements the following steps:

[0267] The step of correcting the advertising plan effect data model of the relationship between user quality and consumption in different time periods in real time according to the real-time user exposure data and the real-time advertising resource time-sharing consumption report data comprises:

[0268] Obtaining hyperparameters of the advertising plan effect data model of the relationship between user quality and consumption in different time periods, wherein the hyperparameters include a learning rate, a discount factor, and a network structure;

[0269] The performance of the advertising plan effect data model of the relationship between user quality and consumption in different time periods is evaluated by using a test set or an online A / B test method, and the advertising plan effect data model of the relationship between user quality and consumption in different time periods is optimized by adjusting the hyperparameters.

[0270] In one embodiment, when the processor executes the computer program, the processor further implements the following steps:

[0271] The step of optimizing the delivery strategy modification method according to the comparison result includes:

[0272] Analyze the best click-through rate, conversion rate, cost per conversion and return on investment in the comparison result, and if the second execution effect indicator value is greater than the second historical execution effect indicator value of the historical target advertising resource executed according to its corresponding delivery strategy, it is determined as positive feedback, otherwise it is determined as negative feedback;

[0273] When the probability of the positive feedback is greater than the preset probability threshold, it is set not to feedback the suggested correction data to receive confirmation information, and the advertising delivery time period, bid, delivery population attributes and regional targeting data of the target delivery strategy are automatically corrected according to the target advertising resources to form a corrected delivery strategy, and the target advertising resources are executed according to the corrected delivery strategy.

[0274] For the specific limitations on the steps implemented when the processor executes the computer program, please refer to the limitations on the method of advertising resource delivery mentioned above, which will not be repeated here.

[0275] In one embodiment, a computer readable storage medium is provided, on which a computer program is stored, and when the computer program is executed by a processor, the following steps are implemented:

[0276] Receiving advertising resources to be delivered and delivery strategies selected corresponding to the advertising resources, adding the advertising resources to an advertising resource delivery list in order, and recording the delivery strategy corresponding to each advertising resource in the advertising resource delivery list;

[0277] Executing the corresponding delivery strategy for each advertising resource in turn according to the advertising resource delivery list, obtaining the execution result of each advertising resource, and recording the execution result in the advertising resource delivery list;

[0278] Collecting statistics on the failure data of the target advertising resources in the execution results to execute the target delivery strategy, analyzing the difference data of the failure of the target advertising resources to execute the target delivery strategy, obtaining and feeding back the suggested correction data for correcting the difference data;

[0279] receiving the confirmation information, and determining whether to agree to modify the difference data of the target advertising resource to the suggested revised data;

[0280] In response to agreeing to modify the difference data of the target advertising resource to the recommended revised data, the returned revised advertising resource is obtained, the revised advertising resource is executed according to the target delivery strategy and a first execution effect indicator value is obtained, the first execution effect indicator value is compared with a first historical execution effect indicator value of the historical advertising resource executed according to the target delivery strategy, and the advertising resource modification method is optimized according to the comparison result;

[0281] In response to disagreement with modifying the difference data of the target advertising resources into the recommended revised data, the target delivery strategy is revised according to the target advertising resources to form a revised delivery strategy, the target advertising resources are executed according to the revised delivery strategy and a second execution effect indicator value is obtained, the second execution effect indicator value is compared with the second historical execution effect indicator value of the historical target advertising resources executed according to their corresponding delivery strategies, and the delivery strategy modification method is optimized according to the comparison result.

[0282] In one embodiment, when the computer program is executed by a processor, the following steps are also implemented:

[0283] The step of executing the corresponding delivery strategy for each advertising resource in turn according to the advertising resource delivery list, obtaining the execution result of each advertising resource, and recording the execution result in the advertising resource delivery list comprises:

[0284] Acquire data of each advertising resource and its corresponding delivery strategy according to the advertising resource delivery list, and determine whether the data of the advertising resource and / or the delivery strategy has changed;

[0285] In response to changes in the data of the advertising resource and / or the delivery strategy, the advertising resource executes the delivery strategy and records the execution result in the advertising resource delivery list;

[0286] In response to the fact that the data of the advertising resources and the delivery strategy have not changed, it is determined whether the advertising resources have result data after executing the delivery strategy. If the result data after executing the delivery strategy exists, the advertising resources do not execute the delivery strategy; if not, the advertising resources execute the delivery strategy, and the execution result is recorded in the advertising resource delivery list.

[0287] In one embodiment, when the computer program is executed by a processor, the following steps are also implemented:

[0288] The steps of receiving the advertisement resources to be delivered and the delivery strategies selected corresponding to the advertisement resources, adding the advertisement resources to the advertisement resource delivery list in order, and recording the delivery strategies corresponding to each advertisement resource in the advertisement resource delivery list include:

[0289] Receiving advertisement resources to be delivered, parsing the advertisement resources to obtain advertisement structure data, wherein the advertisement structure data includes advertisement content and advertisement delivery plan configuration data;

[0290] Converting the advertisement structure data into a data transmission unit message, and performing standardization processing on the data transmission unit message to form a standardized transmission unit message;

[0291] Adding the primary key ID of the object table in the standardized transmission unit message to the advertising resource delivery list in the order of receipt;

[0292] Acquire a delivery strategy corresponding to the advertising resource according to the advertising delivery plan configuration data in the standardized transmission unit message, and record the delivery strategy corresponding to the advertising resource in the advertising resource delivery list, wherein the delivery strategy includes advertising delivery period, bid, delivery crowd attributes and regional targeting data;

[0293] After executing the corresponding delivery strategy in response to the advertising resource, the execution times and the execution results are recorded in the advertising resource delivery list.

[0294] In one embodiment, when the computer program is executed by a processor, the following steps are also implemented:

[0295] The step of acquiring the delivery strategy corresponding to the advertising resource according to the advertising delivery plan configuration data in the standardized transmission unit message, and recording the delivery strategy corresponding to the advertising resource in the advertising resource delivery list comprises:

[0296] Obtain historical user exposure data, historical advertising resource time-sharing consumption report data, and historical user quality data;

[0297] Constructing an advertising plan effect data model of the relationship between user quality and consumption in different time periods based on the historical user exposure data, historical advertising resource time-sharing consumption report data, and historical user quality data;

[0298] Using the advertisement plan effect data model of the relationship between user quality and consumption in different time periods, obtaining a delivery strategy corresponding to the advertisement resource according to the advertisement delivery plan configuration data in the standardized transmission unit message;

[0299] A strategy name and a version number are set for the delivery strategy, and the strategy name and the version number of the delivery strategy are recorded in the advertisement resource delivery list.

[0300] In one embodiment, when the computer program is executed by a processor, the following steps are also implemented:

[0301] The step of constructing a time-division user quality and consumption relationship advertising plan effect data model according to the historical user exposure data, the historical advertising resource time-division consumption report data and the historical user quality data includes:

[0302] Using a Q-learning algorithm to construct an advertising plan effect data model of the relationship between user quality and consumption in different time periods, and training the advertising plan effect data model of the relationship between user quality and consumption in different time periods until convergence;

[0303] Acquire the advertising delivery plan configuration data in the standardized transmission unit message, use the historical user exposure data, the historical advertising resource time-sharing consumption report data and the historical user quality score data as effect data, and encode the advertising delivery plan configuration data and the effect data into a numerical vector;

[0304] The action of adjusting the delivery strategy is defined in the advertising plan effect data model of the relationship between user quality and consumption in different time periods. The action of adjusting the delivery strategy includes changing the delivery time period, increasing the bid, decreasing the bid, changing the attributes of the delivery population and the regional targeting data.

[0305] In one embodiment, when the computer program is executed by a processor, the following steps are also implemented:

[0306] The step of training the advertising plan effect data model of the relationship between user quality and consumption in different time periods until convergence comprises:

[0307] Set the time-divided user quality and consumption relationship advertising plan effect data model and use the Bellman equation to update the Q value: The Q value Q(s,a) represents the expected value of the cumulative reward that may be obtained in the future after taking action a in state s, α is the learning rate, r is the immediate reward, γ is the discount factor, s is the current state, a is the current action, s′ is the next state, a′ is the next action, It is to choose the action that can maximize the Q value in the next state s′;

[0308] When training the advertising plan effect data model of the relationship between user quality and consumption in different time periods, initialize the Q value Q(s,a) to zero, observe the current state s, select an action a according to the current delivery strategy, execute action a, observe the immediate reward r and the next state s′, use the Bellman equation to update the Q value, transfer the state to s′, and repeat the above steps until the Q value converges; wherein selecting an action a according to the current delivery strategy includes: setting the current delivery strategy to an ∈-greedy strategy, using the ∈-greedy strategy to select an action, that is, randomly selecting an action with a probability of ∈, and selecting the current optimal action with a probability of 1-∈.

[0309] In one embodiment, when the computer program is executed by a processor, the following steps are also implemented:

[0310] The step of using the advertisement plan effect data model of the relationship between user quality and consumption in different time periods to obtain the delivery strategy corresponding to the advertisement resource according to the advertisement delivery plan configuration data in the standardized transmission unit message includes:

[0311] Using the advertising plan effect data model of the relationship between user quality and consumption in different time periods, adjusting the bid of the advertising resources in real time according to the delivery strategy;

[0312] Obtain real-time user exposure data and real-time advertising resource consumption report data;

[0313] According to the real-time user exposure data and the real-time advertising resource time-sharing consumption report data, the advertising plan effect data model of the relationship between user quality and consumption in different time periods is corrected in real time;

[0314] The revised advertising plan effect data model of the relationship between user quality and consumption in different time periods is used to reform the delivery strategy corresponding to the advertising resources.

[0315] In one embodiment, when the computer program is executed by a processor, the following steps are also implemented:

[0316] The step of correcting the advertising plan effect data model of the relationship between user quality and consumption in different time periods in real time according to the real-time user exposure data and the real-time advertising resource time-sharing consumption report data comprises:

[0317] Obtaining hyperparameters of the advertising plan effect data model of the relationship between user quality and consumption in different time periods, wherein the hyperparameters include a learning rate, a discount factor, and a network structure;

[0318] The performance of the advertising plan effect data model of the relationship between user quality and consumption in different time periods is evaluated by using a test set or an online A / B test method, and the advertising plan effect data model of the relationship between user quality and consumption in different time periods is optimized by adjusting the hyperparameters.

[0319] In one embodiment, when the computer program is executed by a processor, the following steps are also implemented:

[0320] The step of optimizing the delivery strategy modification method according to the comparison result includes:

[0321] Analyze the best click-through rate, conversion rate, cost per conversion and return on investment in the comparison result, and if the second execution effect indicator value is greater than the second historical execution effect indicator value of the historical target advertising resource executed according to its corresponding delivery strategy, it is determined as positive feedback, otherwise it is determined as negative feedback;

[0322] When the probability of the positive feedback is greater than the preset probability threshold, it is set not to feedback the suggested correction data to receive confirmation information, and the advertising delivery time period, bid, delivery population attributes and regional targeting data of the target delivery strategy are automatically corrected according to the target advertising resources to form a corrected delivery strategy, and the target advertising resources are executed according to the corrected delivery strategy.

[0323] For the specific limitations on the steps implemented when the computer program is executed by the processor, please refer to the limitations on the method of advertising resource delivery mentioned above, which will not be repeated here.

[0324] Those skilled in the art can understand that all or part of the processes in the above-mentioned embodiment methods can be completed by instructing the relevant hardware through a computer program, and the computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, any reference to memory, storage, database or other media used in the embodiments provided in this application can include non-volatile and / or volatile memory. Non-volatile memory can include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM) or flash memory. Volatile memory can include random access memory (RAM) or external cache memory. As an illustration and not limitation, RAM is available in many forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link (Synchlink) DRAM (SLDRAM), memory bus (Rambus) direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM).

[0325] The technical features of the above embodiments may be combined arbitrarily. To make the description concise, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0326] The above-mentioned embodiments only express several implementation methods of the present application, and the descriptions thereof are relatively specific and detailed, but they cannot be understood as limiting the scope of the invention patent. It should be pointed out that, for a person of ordinary skill in the art, several variations and improvements can be made without departing from the concept of the present application, and these all belong to the protection scope of the present application. Therefore, the protection scope of the patent of the present application shall be subject to the attached claims.

Claims

1. A method for delivering advertising resources, characterized in that: include: Receiving advertising resources to be delivered and delivery strategies selected corresponding to the advertising resources, adding the advertising resources to an advertising resource delivery list in order, and recording the delivery strategy corresponding to each advertising resource in the advertising resource delivery list; Executing the corresponding delivery strategy for each advertising resource in turn according to the advertising resource delivery list, obtaining the execution result of each advertising resource, and recording the execution result in the advertising resource delivery list; Collecting statistics on the failure data of the target advertising resources in the execution results to execute the target delivery strategy, analyzing the difference data of the failure of the target advertising resources to execute the target delivery strategy, obtaining and feeding back the suggested correction data for correcting the difference data; receiving the confirmation information, and determining whether to agree to modify the difference data of the target advertising resource to the suggested revised data; In response to agreeing to modify the difference data of the target advertising resource to the recommended revised data, the returned revised advertising resource is obtained, the revised advertising resource is executed according to the target delivery strategy and a first execution effect indicator value is obtained, the first execution effect indicator value is compared with a first historical execution effect indicator value of the historical advertising resource executed according to the target delivery strategy, and the advertising resource modification method is optimized according to the comparison result; In response to disagreement with modifying the difference data of the target advertising resources into the recommended revised data, the target delivery strategy is revised according to the target advertising resources to form a revised delivery strategy, the target advertising resources are executed according to the revised delivery strategy and a second execution effect indicator value is obtained, the second execution effect indicator value is compared with the second historical execution effect indicator value of the historical target advertising resources executed according to their corresponding delivery strategies, and the delivery strategy modification method is optimized according to the comparison result.

2. The method for delivering advertising resources according to claim 1, characterized in that: The step of executing the corresponding delivery strategy for each advertising resource in turn according to the advertising resource delivery list, obtaining the execution result of each advertising resource, and recording the execution result in the advertising resource delivery list comprises: Acquire data of each advertising resource and its corresponding delivery strategy according to the advertising resource delivery list, and determine whether the data of the advertising resource and / or the delivery strategy has changed; In response to changes in the data of the advertising resource and / or the delivery strategy, the advertising resource executes the delivery strategy and records the execution result in the advertising resource delivery list; In response to the fact that the data of the advertising resources and the delivery strategy have not changed, it is determined whether the advertising resources have result data after executing the delivery strategy. If so, the advertising resources do not execute the delivery strategy. If not, the advertising resources execute the delivery strategy, and the execution result is recorded in the advertising resource delivery list.

3. The method for delivering advertising resources according to claim 1, characterized in that: The steps of receiving the advertisement resources to be delivered and the delivery strategies selected corresponding to the advertisement resources, adding the advertisement resources to the advertisement resource delivery list in order, and recording the delivery strategies corresponding to each advertisement resource in the advertisement resource delivery list include: Receiving advertisement resources to be delivered, parsing the advertisement resources to obtain advertisement structure data, wherein the advertisement structure data includes advertisement content and advertisement delivery plan configuration data; Converting the advertisement structure data into a data transmission unit message, and performing standardization processing on the data transmission unit message to form a standardized transmission unit message; Adding the primary key ID of the object table in the standardized transmission unit message to the advertising resource delivery list in the order of receipt; Acquire a delivery strategy corresponding to the advertising resource according to the advertising delivery plan configuration data in the standardized transmission unit message, and record the delivery strategy corresponding to the advertising resource in the advertising resource delivery list, wherein the delivery strategy includes an advertising delivery period, a bid, delivery crowd attributes, and regional targeting data; After executing the corresponding delivery strategy in response to the advertising resource, the execution times and the execution results are recorded in the advertising resource delivery list.

4. The method for delivering advertising resources according to claim 3, characterized in that: The step of acquiring the delivery strategy corresponding to the advertising resource according to the advertising delivery plan configuration data in the standardized transmission unit message, and recording the delivery strategy corresponding to the advertising resource in the advertising resource delivery list comprises: Obtain historical user exposure data, historical advertising resource time-sharing consumption report data, and historical user quality data; Constructing an advertising plan effect data model of the relationship between user quality and consumption in different time periods based on the historical user exposure data, historical advertising resource time-sharing consumption report data, and historical user quality data; Utilizing the advertisement plan effect data model of the relationship between user quality and consumption in different time periods, obtaining a delivery strategy corresponding to the advertisement resource according to the advertisement delivery plan configuration data in the standardized transmission unit message; A strategy name and a version number are set for the delivery strategy, and the strategy name and the version number of the delivery strategy are recorded in the advertisement resource delivery list.

5. The method for delivering advertising resources according to claim 3, characterized in that: The step of constructing a time-division user quality and consumption relationship advertising plan effect data model according to the historical user exposure data, the historical advertising resource time-division consumption report data and the historical user quality data includes: Using a Q learning algorithm to construct an advertising plan effect data model of the relationship between user quality and consumption in different time periods, and training the advertising plan effect data model of the relationship between user quality and consumption in different time periods until convergence; Acquire the advertising plan configuration data in the standardized transmission unit message, use the historical user exposure data, the historical advertising resource time-sharing consumption report data and the historical user quality score data as effect data, and encode the advertising plan configuration data and the effect data into a numerical vector; The action of adjusting the delivery strategy is defined in the advertising plan effect data model of the relationship between user quality and consumption in different time periods. The action of adjusting the delivery strategy includes changing the delivery time period, increasing the bid, decreasing the bid, changing the attributes of the delivery population and the regional targeting data.

6. The method for delivering advertising resources according to claim 5, characterized in that: The step of training the advertising plan effect data model of the relationship between user quality and consumption in different time periods until convergence comprises: Set the time-divided user quality and consumption relationship advertising plan effect data model and use the Bellman equation to update the Q value: The Q value Q(s,a) represents the expected value of the cumulative reward that may be obtained in the future after taking action a in state s, α is the learning rate, r is the immediate reward, γ is the discount factor, s is the current state, a is the current action, s′ is the next state, a′ is the next action, It is to choose the action that can maximize the Q value in the next state s′; When training the advertising plan effect data model of the relationship between user quality and consumption in different time periods, initialize the Q value Q(s,a) to zero, observe the current state s, select an action a according to the current delivery strategy, execute action a, observe the immediate reward r and the next state s′, use the Bellman equation to update the Q value, transfer the state to s′, and repeat the above steps until the Q value converges.

7. The method for delivering advertising resources according to claim 4, characterized in that: The step of using the advertisement plan effect data model of the relationship between user quality and consumption in different time periods to obtain the delivery strategy corresponding to the advertisement resource according to the advertisement delivery plan configuration data in the standardized transmission unit message includes: Using the advertising plan effect data model of the relationship between user quality and consumption in different time periods, adjusting the bid of the advertising resources in real time according to the delivery strategy; Obtain real-time user exposure data and real-time advertising resource consumption report data; According to the real-time user exposure data and the real-time advertising resource time-sharing consumption report data, the advertising plan effect data model of the relationship between user quality and consumption in different time periods is corrected in real time; The revised advertising plan effect data model of the relationship between user quality and consumption in different time periods is used to reform the delivery strategy corresponding to the advertising resources.

8. The method for delivering advertising resources according to claim 7, characterized in that: The step of correcting the advertising plan effect data model of the relationship between user quality and consumption in different time periods in real time according to the real-time user exposure data and the real-time advertising resource time-sharing consumption report data comprises: Obtaining hyperparameters of the advertising plan effect data model of the relationship between user quality and consumption in different time periods, wherein the hyperparameters include a learning rate, a discount factor, and a network structure; The performance of the advertising plan effect data model of the relationship between user quality and consumption in different time periods is evaluated by using a test set or an online A / B test method, and the advertising plan effect data model of the relationship between user quality and consumption in different time periods is optimized by adjusting the hyperparameters.

9. The method for delivering advertising resources according to claim 1, characterized in that: The step of optimizing the delivery strategy modification method according to the comparison result includes: Analyze the best click-through rate, conversion rate, cost per conversion and return on investment in the comparison result, and if the second execution effect indicator value is greater than the second historical execution effect indicator value of the historical target advertising resource executed according to its corresponding delivery strategy, it is determined as positive feedback, otherwise it is determined as negative feedback; When the probability of the positive feedback is greater than the preset probability threshold, it is set not to feedback the suggested correction data to receive confirmation information, and the advertising delivery time period, bid, delivery population attributes and regional targeting data of the target delivery strategy are automatically corrected according to the target advertising resources to form a corrected delivery strategy, and the target advertising resources are executed according to the corrected delivery strategy.

10. A computer device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that: When the processor executes the computer program, the steps of the method according to any one of claims 1 to 9 are implemented.

Images