Method, device, equipment and storage medium for controlling execution of RTA policy rules

Abstract

The application relates to a method, device, equipment and storage medium for controlling execution of an RTA strategy rule. The method comprises the following steps: acquiring all channels currently executing the RTA strategy rule and all RTA strategy rules corresponding to each channel; acquiring, for a target channel, the channel advertisement consumption amount and the channel available quota within a first time length before the current time and the historical data and the executed RTA strategy rule, and inputting into a convolutional neural network model; comparing the channel advertisement consumption amount and the channel available quota within the first time length before the current time with the channel advertisement consumption amount and the channel available quota in the historical data to determine whether the RTA strategy rule needs to be adjusted; if yes, determining the value of the fusion model part in the real-time RTA strategy rule according to the historical data; and if no, keeping the RTA strategy rule currently executed by the target channel and outputting. The method reduces the workload of business personnel and ensures the accuracy of result determination.

Description

Technical Field

[0001] This application relates to the field of advertising delivery technology, and in particular to a method, apparatus, computer device, and storage medium for controlling the execution of RTA strategy rules. Background Technology

[0002] In the advertising business, RTA refers to advertisers using API interfaces to filter and intervene in advertising traffic in real time. RTA, short for Realtime API, sends customer identification requests to advertisers during the targeting process to filter customers and meet their real-time personalized delivery needs. The RTA service is a system advertisers use to control whether users participate in ad bidding. The advertiser's system provides an RTA request interface to the media, allowing the media to initiate an RTA request to the advertiser regarding ad placement information, inquiring whether a specific user should participate in the ad placement bidding.

[0003] Currently, advertisers determine whether a user participates in bidding primarily by using a model score to execute pre-configured business rules. This involves pre-setting several fixed strategy rules, outputting a model score based on ad spend and initial credit limit, and then having business personnel select the appropriate strategy rule from these pre-defined, fixed rules.

[0004] However, because these fixed strategy rules cannot be modified or adjusted in real time, it is very cumbersome for business personnel to adjust the rules according to the day's advertising performance. Moreover, the inability to modify and adjust the rules in real time results in the inability to adjust the strategy rules or to postpone the adjustment. The inability to adjust the strategy rules in real time based on the real-time advertising performance data ultimately leads to deviations in the RTA service's results regarding whether users participate. Summary of the Invention

[0005] Therefore, it is necessary to provide a method, apparatus, computer device, and storage medium for controlling the execution of RTA policy rules to address the above-mentioned technical problems. This can solve the technical problem that after configuring RTA policy rules in traditional RTA services, the policy rules cannot be adjusted in real time based on real-time advertising performance data, which ultimately leads to deviations in the RTA service's assessment of whether users participate.

[0006] On the one hand, a method for controlling the execution of RTA policy rules is provided, the method comprising:

[0007] Get all channels currently executing RTA policy rules and all RTA policy rules corresponding to each channel;

[0008] The system obtains the channel advertising spending amount and available credit for the first time period before the current time, as well as the corresponding RTA strategy rules, and inputs them into the convolutional neural network model.

[0009] The system obtains historical data on advertising spending and available credit for the target channel, along with the corresponding RTA strategy rules, and inputs them into a convolutional neural network model. The available credit for the channel includes the initial loan amount and / or the initial credit limit granted by the channel.

[0010] The convolutional neural network model compares the channel advertising consumption amount and available channel credit in the first time period before the current time with the channel advertising consumption amount and available channel credit in historical data to determine whether the RTA strategy rules currently being executed by the target channel need to be adjusted.

[0011] If it is necessary to adjust the RTA strategy rules currently being executed by the target channel, then the value of the fusion model score in the real-time RTA strategy rules is determined based on historical data; if it is not necessary to adjust the RTA strategy rules currently being executed by the target channel, then the RTA strategy rules currently being executed by the target channel are kept as real-time RTA strategy rules.

[0012] When the media corresponding to the target channel initiates an RTA service inquiry, the real-time RTA policy rules are obtained and executed.

[0013] In one embodiment, the convolutional neural network model compares the channel advertising spending and available channel credit within a first time period prior to the current time with the channel advertising spending and available channel credit in historical data to determine whether the RTA strategy rule currently being executed by the target channel needs to be adjusted. This includes the following steps:

[0014] The ratio of channel advertising spending to available channel credit within the first hour preceding the current time is calculated as the real-time consumption rate parameter.

[0015] The ratio of channel advertising spending to available channel credit within a second time period prior to the current time in historical data is used as the first baseline spending rate parameter; the second time period is longer than the first time period.

[0016] When the real-time consumption rate parameter is greater than the consumption rate parameter, it is determined that the RTA strategy rule currently being executed by the target channel needs to be adjusted.

[0017] In one embodiment, when the real-time consumption rate parameter is less than or equal to the first baseline consumption rate parameter, the method further includes:

[0018] The ratio of channel advertising spending to available channel credit within a third time period prior to the current time in historical data is used as the second baseline consumption rate parameter; the third time period is longer than the second time period.

[0019] When the real-time consumption rate parameter is greater than the consumption rate parameter, it is determined that the RTA strategy rule currently being executed by the target channel needs to be adjusted.

[0020] When the real-time consumption rate parameter is less than or equal to the second baseline consumption rate parameter, it is determined that there is no need to adjust the RTA strategy rule currently being executed by the target channel.

[0021] In one embodiment, if it is necessary to adjust the RTA policy rule currently being executed by the target channel, the step of determining the value of the fusion model score in the real-time RTA policy rule based on historical data includes:

[0022] Obtain the value and level of the available quota for the target channel corresponding to the RTA policy rule currently being executed;

[0023] The available credit limit for the target channel is set to increase to the value within the next level;

[0024] Set the value of the fusion model score in the real-time RTA strategy rule based on the latest value of the available quota of the target channel.

[0025] In one embodiment, after obtaining the real-time RTA policy rules and executing the steps when the media corresponding to the target channel initiates an RTA service query, the method further includes:

[0026] Determine whether to participate in the bidding and respond to RTA service inquiries initiated by the media corresponding to the target channel;

[0027] Collect data on channel advertising spending and available credit after the target channel executes the real-time RTA strategy rules, and feed this data back to the convolutional neural network model;

[0028] The convolutional neural network model uses collected channel advertising spending and available channel credit data as historical data to iteratively update the real-time RTA strategy rules.

[0029] In one embodiment, the convolutional neural network model uses collected channel advertising spending and available channel credit data as historical data to iteratively update real-time RTA policy rules, including:

[0030] Obtain the available budget of the target channel before the value is increased, and the channel advertising consumption amount corresponding to the increase to the next level, and compare them;

[0031] The first quota value is the value of the available quota corresponding to the maximum advertising expenditure of the channel.

[0032] Select multiple second credit limits within the range of [first credit limit value - first threshold, first credit limit value + first threshold].

[0033] The available quota of the target channel is updated based on multiple second quota values. The values ​​are set as multiple experimental RTA strategy rules, and the value of the fusion model score is set in each experimental RTA strategy rule.

[0034] After executing all experimental RTA strategy rules on a weekly basis, the channel advertising consumption amount and available channel credit data after executing each experimental RTA strategy rule are collected and fed back to the convolutional neural network model.

[0035] The available credit for the channel corresponding to the maximum channel advertising expenditure in all experimental RTA strategy rules is compared and used as the credit value for the second iteration update;

[0036] This process is repeated until the available credit for the channel corresponding to the maximum channel advertising spending is obtained, which is then used as the final credit limit.

[0037] In one embodiment, the step of obtaining all channels currently executing RTA policy rules and all RTA policy rules corresponding to each channel includes:

[0038] Get the names of all channels currently executing the RTA policy rules, including Tencent, Toutiao, and Baidu;

[0039] Obtain all RTA strategy rules corresponding to each channel and the value of the fusion model score in each RTA strategy rule;

[0040] Establish a database by mapping the channel name to the values ​​of the fusion model points in the corresponding executed RTA strategy rules.

[0041] On the other hand, an apparatus for controlling the execution of RTA policy rules is provided, the apparatus comprising:

[0042] The current RTA policy rule acquisition module is used to acquire all channels currently executing RTA policy rules and all RTA policy rules corresponding to each channel;

[0043] The current execution strategy data acquisition module is used to obtain the channel advertising consumption amount and available quota of the target channel within the first time period before the current time, as well as the corresponding RTA strategy rules, and input them into the convolutional neural network model;

[0044] The historical execution strategy data acquisition module is used to acquire the channel advertising consumption amount and available credit of the target channel from historical data, as well as the corresponding RTA strategy rules, and input them into the convolutional neural network model; the available credit of the channel includes the channel's first loan amount and / or the channel's initial credit limit.

[0045] The strategy rule adjustment judgment module is used by the convolutional neural network model to compare the channel advertising consumption amount and channel available quota in the first time period before the current time with the channel advertising consumption amount and channel available quota in historical data, and to determine whether it is necessary to adjust the RTA strategy rule currently being executed by the target channel.

[0046] The real-time RTA strategy rule acquisition module is used to determine the value of the fusion model part in the real-time RTA strategy rule based on historical data if it is necessary to adjust the RTA strategy rule currently being executed by the target channel; otherwise, it keeps the RTA strategy rule currently being executed by the target channel as the real-time RTA strategy rule.

[0047] The Real-Time RTA Policy Rule Output and Execution Module is used to obtain and execute real-time RTA policy rules when the media corresponding to the target channel initiates an RTA service query.

[0048] In another aspect, 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 executes the computer program to perform the following steps:

[0049] Get all channels currently executing RTA policy rules and all RTA policy rules corresponding to each channel;

[0050] The system obtains the channel advertising spending amount and available credit for the first time period before the current time, as well as the corresponding RTA strategy rules, and inputs them into the convolutional neural network model.

[0051] The system obtains historical data on advertising spending and available credit for the target channel, along with the corresponding RTA strategy rules, and inputs them into a convolutional neural network model. The available credit for the channel includes the initial loan amount and / or the initial credit limit granted by the channel.

[0052] The convolutional neural network model compares the channel advertising consumption amount and available channel credit in the first time period before the current time with the channel advertising consumption amount and available channel credit in historical data to determine whether the RTA strategy rules currently being executed by the target channel need to be adjusted.

[0053] If it is necessary to adjust the RTA strategy rules currently being executed by the target channel, then the value of the fusion model score in the real-time RTA strategy rules is determined based on historical data; if it is not necessary to adjust the RTA strategy rules currently being executed by the target channel, then the RTA strategy rules currently being executed by the target channel are kept as real-time RTA strategy rules.

[0054] When the media corresponding to the target channel initiates an RTA service inquiry, the real-time RTA policy rules are obtained and executed.

[0055] In another aspect, a computer-readable storage medium is provided having a computer program stored thereon, which, when executed by a processor, performs the following steps:

[0056] Get all channels currently executing RTA policy rules and all RTA policy rules corresponding to each channel;

[0057] The system obtains the channel advertising spending amount and available credit for the first time period before the current time, as well as the corresponding RTA strategy rules, and inputs them into the convolutional neural network model.

[0058] The system obtains historical data on advertising spending and available credit for the target channel, along with the corresponding RTA strategy rules, and inputs them into a convolutional neural network model. The available credit for the channel includes the initial loan amount and / or the initial credit limit granted by the channel.

[0059] The convolutional neural network model compares the channel advertising consumption amount and available channel credit in the first time period before the current time with the channel advertising consumption amount and available channel credit in historical data to determine whether the RTA strategy rules currently being executed by the target channel need to be adjusted.

[0060] If it is necessary to adjust the RTA strategy rules currently being executed by the target channel, then the value of the fusion model score in the real-time RTA strategy rules is determined based on historical data; if it is not necessary to adjust the RTA strategy rules currently being executed by the target channel, then the RTA strategy rules currently being executed by the target channel are kept as real-time RTA strategy rules.

[0061] When the media corresponding to the target channel initiates an RTA service inquiry, the real-time RTA policy rules are obtained and executed.

[0062] The aforementioned method, apparatus, computer equipment, and storage medium for controlling the execution of RTA strategy rules, combined with a convolutional neural network model, enable real-time adjustment of RTA strategy rules based on channel advertising spending and available channel credit. This reduces the workload of business personnel, minimizes the possibility of errors in manual strategy configuration, and ensures the accuracy of RTA service's judgment on user participation by adjusting RTA strategy rules in a timely manner. Attached Figure Description

[0063] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0064] Figure 1 This is an application environment diagram of a method for controlling the execution of RTA policy rules in one embodiment;

[0065] Figure 2 This is a flowchart illustrating a method for controlling the execution of RTA policy rules in one embodiment;

[0066] Figure 3 This is a flowchart illustrating a method for controlling the execution of RTA policy rules when the available credit limit for a channel includes the initial loan amount for that channel, as described in one embodiment.

[0067] Figure 4 This is a flowchart illustrating a method for controlling the execution of RTA policy rules when the available credit limit for a channel includes the initial credit limit for the channel, as described in one embodiment.

[0068] Figure 5 This is a flowchart illustrating the steps for obtaining all channels currently executing RTA policy rules and all RTA policy rules corresponding to each channel in one embodiment.

[0069] Figure 6 This is a flowchart illustrating the steps of the convolutional neural network model in one embodiment to compare the channel advertising consumption amount and available channel credit within the first time period before the current time with the channel advertising consumption amount and available channel credit in historical data to determine whether the RTA strategy rule currently being executed by the target channel needs to be adjusted.

[0070] Figure 7 This is a flowchart illustrating the steps for determining the value of the fusion model in the real-time RTA policy rule based on historical data if it is necessary to adjust the RTA policy rule currently being executed by the target channel in one embodiment.

[0071] Figure 8 This is a flowchart illustrating the steps of the convolutional neural network model in one embodiment to iteratively update the real-time RTA strategy rules by using collected channel advertising consumption amount and channel available credit data as historical data.

[0072] Figure 9 This is a structural block diagram of a device for controlling the execution of RTA policy rules in one embodiment;

[0073] Figure 10This is an internal structural diagram of a computer device in one embodiment. Detailed Implementation

[0074] To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description is provided in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the scope of this application.

[0075] As described in the background section, traditional RTA services, after configuring RTA strategy rules, cannot adjust these rules based on real-time advertising performance data. Even if the RTA strategy rules are adjusted, the adjustment will be delayed, and the inability to adjust the rules in real time based on advertising performance data will ultimately lead to deviations in the RTA service's assessment of user participation.

[0076] The above situation arises because traditional RTA services do not combine ad spend with the final performance data generated by the advertiser, and all RTA strategy rules are not adjusted in a timely manner within a certain period of time.

[0077] Therefore, this invention creatively proposes a method for controlling the execution of RTA strategy rules. This method combines real-time ad consumption data and the resulting initial loan amount and / or credit limit during actual ad delivery to continuously adjust and optimize the RTA strategy rules, thereby enabling better differentiated adjustments for target customer groups and better control over ad delivery costs. If ad consumption is found to be high within a certain time period, but the actual initial loan amount is relatively low, the model score for participating users can be increased, thereby reducing user participation and increasing the number of times high-quality customers participate.

[0078] The method for controlling the execution of RTA policy rules provided in this application can be applied to, for example... Figure 1 In the application environment shown, terminal 102 communicates with media advertising platform 103 via a network, and media advertising platform 103 communicates with advertiser server 104 via a network. Terminal 102 can be, but is not limited to, various personal computers, laptops, smartphones, tablets, and portable wearable devices. Customers can browse advertising information from platforms such as Toutiao and Tencent through terminal 102. Media advertising platform 103 includes servers from platforms such as Toutiao, Tencent, and Baidu. Media advertising platform 103 and advertiser server 104 can be implemented using independent servers or a server cluster consisting of multiple servers.

[0079] In this embodiment, the advertiser server 104 provides RTA services to the media advertising platform 103 through an API interface; the media advertising platform 103 sends an advertising resource request to the idle advertising space generated by the customer during the browsing of the webpage, and initiates an RTA consultation through the API interface when processing the advertising resource request.

[0080] In one embodiment, such as Figure 2 As shown, a method for controlling the execution of RTA policy rules is provided, which can be applied to... Figure 1 Taking advertiser server 104 as an example, the following steps are included:

[0081] Step S1: Obtain all channels currently executing RTA policy rules and all RTA policy rules corresponding to each channel;

[0082] Step S2: Obtain the channel advertising consumption amount and available quota for the first time period before the current time for the target channel, as well as the corresponding RTA strategy rules, and input them into the convolutional neural network model (AI model);

[0083] Step S3: Obtain the channel advertising consumption amount and available credit limit from the historical data of the target channel, as well as the corresponding RTA strategy rules, and input them into the convolutional neural network model; the available credit limit includes the channel's initial loan amount and / or the channel's initial credit limit.

[0084] Step S4: The convolutional neural network model compares the channel advertising consumption amount and available channel credit in the first time period before the current time with the channel advertising consumption amount and available channel credit in the historical data to determine whether it is necessary to adjust the RTA strategy rules currently being executed by the target channel.

[0085] Step S5: If it is necessary to adjust the RTA strategy rule currently being executed by the target channel, then determine the value of the fusion model score in the real-time RTA strategy rule based on historical data; if it is not necessary to adjust the RTA strategy rule currently being executed by the target channel, then keep the RTA strategy rule currently being executed by the target channel as the real-time RTA strategy rule.

[0086] Step S6: When the media corresponding to the target channel initiates an RTA service query, obtain the real-time RTA policy rules and execute them.

[0087] like Figure 3 As shown, in one embodiment, the available credit limit for the channel includes the initial loan amount for the channel.

[0088] At this point, the convolutional neural network model compares the channel advertising spending and available channel credit within the first time period prior to the current time with the channel advertising spending and available channel credit in historical data to determine whether the RTA strategy rule currently being executed by the target channel needs to be adjusted. This includes the following steps:

[0089] The ratio of channel advertising spending to channel initial loan amount (CPS1) within the first hour prior to the current time is used as a real-time parameter for each sales transaction.

[0090] The ratio of channel advertising spending to channel initial loan amount in the second time period before the current time in historical data (CPS2) is used as the benchmark parameter for each sales transaction; the first time period is less than the second time period.

[0091] When the real-time sales parameter per transaction is greater than the baseline sales parameter per transaction, it is determined that the RTA strategy rule currently being executed by the target channel needs to be adjusted.

[0092] When the real-time sales parameter per transaction is less than or equal to the baseline sales parameter per transaction, it is determined that there is no need to adjust the RTA strategy rules currently being executed by the target channel.

[0093] CPS is short for Cost Per Sales.

[0094] like Figure 4 As shown, in another embodiment, the available credit line for the channel includes the initial credit line for the channel.

[0095] At this point, the convolutional neural network model compares the channel advertising spending and available channel credit within the first time period prior to the current time with the channel advertising spending and available channel credit in historical data to determine whether the RTA strategy rule currently being executed by the target channel needs to be adjusted. This includes the following steps:

[0096] The ratio of channel advertising expenditure to initial channel credit limit (CPCL1) within the first hour preceding the current time is calculated as the real-time cost per credit limit parameter.

[0097] The ratio of channel advertising expenditure to initial channel credit limit (CPCL2) within the second time period prior to the current time in historical data is used as the benchmark cost per credit limit parameter; the first time period is shorter than the second time period.

[0098] When the real-time cost per credit limit parameter is greater than the baseline cost per credit limit parameter, it is determined that the RTA strategy rule currently being executed by the target channel needs to be adjusted.

[0099] When the real-time cost per credit limit parameter is less than or equal to the baseline cost per credit limit parameter, it is determined that there is no need to adjust the RTA strategy rules currently being executed by the target channel.

[0100] CPCL is short for costpercreditlimit.

[0101] In another embodiment, the available credit limit for the channel includes the initial loan amount and the initial credit line granted by the channel. In this case, the convolutional neural network model compares the channel advertising consumption amount and available credit limit within the first time period prior to the current time with the channel advertising consumption amount and available credit limit in historical data to determine whether it is necessary to adjust the RTA strategy rule steps currently being executed by the target channel. This includes a combination of the two judgment methods described above, i.e. Figure 3 , Figure 4 The combination methods shown will not be elaborated upon here.

[0102] like Figure 5 As shown, in this embodiment, the step of obtaining all channels currently executing RTA policy rules and all RTA policy rules corresponding to each channel includes:

[0103] Step S11: Obtain the names of all channels currently executing the RTA policy rules, including Tencent, Toutiao, and Baidu.

[0104] Step S12: Obtain all RTA strategy rules corresponding to each channel and the value of the fusion model score in each RTA strategy rule; where each RTA strategy rule has its own focus.

[0105] Step S13: Establish a mapping set between the channel name and the value of the fusion model in the corresponding executed RTA strategy rule to form a database.

[0106] like Figure 6 As shown, in this embodiment, the convolutional neural network model compares the channel advertising spending and available channel credit within the first time period before the current time with the channel advertising spending and available channel credit in historical data to determine whether the RTA strategy rule currently being executed by the target channel needs to be adjusted. This includes the following steps:

[0107] Step S41: Calculate the ratio of channel advertising spending to available channel credit within the first time period prior to the current time as the real-time spending rate parameter;

[0108] Step S42: Calculate the ratio of channel advertising consumption amount to channel available credit within the second duration before the current time in historical data as the first benchmark consumption rate parameter; the second duration is longer than the first duration;

[0109] Step S43: When the real-time consumption rate parameter is greater than the consumption rate parameter, it is determined that the RTA strategy rule currently being executed by the target channel needs to be adjusted.

[0110] like Figure 6 As shown, in this embodiment, when the real-time consumption rate parameter is less than or equal to the first baseline consumption rate parameter, the method further includes:

[0111] Step S44: Calculate the ratio of channel advertising consumption amount to channel available credit within the third time period before the current time in historical data as the second benchmark consumption rate parameter; the third time period is longer than the second time period;

[0112] Step S45: When the real-time consumption rate parameter is greater than the consumption rate parameter, it is determined that the RTA strategy rule currently being executed by the target channel needs to be adjusted.

[0113] Step S46: When the real-time consumption rate parameter is less than or equal to the second baseline consumption rate parameter, it is determined that there is no need to adjust the RTA strategy rule currently being executed by the target channel.

[0114] For example, the first timeframe is the 30 minutes of the current day, updated every 30 minutes; the second timeframe is yesterday (one day); the third timeframe is the past 3 or 7 days. These timeframes can be set according to actual needs. Preferably, the ratio of channel advertising spending to available channel credit over the past 7 days is used as the baseline spending rate parameter, ensuring minimal variation and more stable analysis results. In other words, when no third timeframe is set, the second timeframe is preferably the past 7 days.

[0115] like Figure 7 As shown in this embodiment, if it is necessary to adjust the RTA policy rule currently being executed by the target channel, the step of determining the value of the fusion model score in the real-time RTA policy rule based on historical data includes:

[0116] Step S51: Obtain the value and level of the available quota for the target channel corresponding to the RTA policy rule currently being executed.

[0117] Step S52: Increase the available credit limit of the target channel to the value within the next level;

[0118] Step S53: Set the value of the fusion model score in the real-time RTA strategy rule according to the latest value of the available quota of the target channel.

[0119] like Figure 2 As shown, in this embodiment, after obtaining the real-time RTA policy rules and executing the steps when the media corresponding to the target channel initiates an RTA service query, the method further includes:

[0120] Step S7: Determine whether to participate in the bidding and respond to the RTA service inquiry initiated by the media corresponding to the target channel;

[0121] Step S8: Collect data on channel advertising spending and available credit after the target channel executes the real-time RTA strategy rules, and feed it back into the convolutional neural network model;

[0122] Step S9: The convolutional neural network model uses the collected channel advertising consumption amount and channel available quota data as historical data to iteratively update the real-time RTA strategy rules.

[0123] like Figure 8 As shown, in this embodiment, the convolutional neural network model uses the collected channel advertising spending and available channel credit data as historical data to iteratively update the real-time RTA strategy rules, including:

[0124] Step S91: Obtain the available channel credit of the target channel before the value is increased and the channel advertising consumption amount corresponding to the value of the next level, and compare them.

[0125] Step S92: Compare the value of the available credit for the channel corresponding to the maximum value of the channel advertising consumption amount with the value of the first credit limit.

[0126] Step S93: Select multiple second credit limit values ​​within the range of [first credit limit value - first threshold, first credit limit value + first threshold].

[0127] Step S94: Update the channel available quota of the target channel according to multiple second quota values ​​to multiple experimental RTA strategy rules, and set the value of the fusion model score in each experimental RTA strategy rule.

[0128] Step S95: After executing all experimental RTA strategy rules on a weekly basis, collect data on channel advertising spending and available credit after executing each experimental RTA strategy rule for the target channel and feed it back into the convolutional neural network model.

[0129] Step S96: Compare the available channel credit corresponding to the maximum channel advertising consumption amount in all experimental RTA strategy rules with the value of the credit limit updated in the second iteration.

[0130] Step S97, repeat this process until the value of the available channel credit corresponding to the maximum channel advertising consumption amount is obtained as the final credit value.

[0131] The above-mentioned method for controlling the execution of RTA strategy rules combines a convolutional neural network model to achieve real-time adjustment of RTA strategy rules based on channel advertising spending and available channel credit. This reduces the workload of business personnel, reduces the possibility of errors in manual strategy configuration, and ensures the accuracy of RTA service's judgment on whether users participate by adjusting RTA strategy rules in a timely manner.

[0132] It should be understood that, although Figures 2-8 The steps in the flowchart are shown sequentially as indicated by the arrows, but these steps are not necessarily executed in the order indicated by the arrows. Unless otherwise specified herein, there is no strict order in which these steps are executed, and they can be performed in other orders. Figures 2-8 At least some of the steps in the process 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 some of the sub-steps or stages of other steps.

[0133] In one embodiment, such as Figure 9 As shown, a device 10 for controlling the execution of RTA policy rules is provided, including: a current RTA policy rule acquisition module 1, a current execution policy data acquisition module 2, a historical execution policy data acquisition module 3, a policy rule adjustment judgment module 4, a real-time RTA policy rule acquisition module 5, and a real-time RTA policy rule output execution module 6.

[0134] The current RTA strategy rule acquisition module 1 is used to acquire all channels currently executing RTA strategy rules and all RTA strategy rules corresponding to each channel.

[0135] The current execution strategy data acquisition module 2 is used to obtain the channel advertising consumption amount and available quota of the target channel within the first time period before the current time, as well as the corresponding RTA strategy rules, and input them into the convolutional neural network model.

[0136] The historical execution strategy data acquisition module 3 is used to acquire the channel advertising consumption amount and available channel quota, as well as the corresponding RTA strategy rules from the historical data of the target channel, and input them into the convolutional neural network model.

[0137] The strategy rule adjustment judgment module 4 is used by the convolutional neural network model to compare the channel advertising consumption amount and channel available quota in the first time period before the current time with the channel advertising consumption amount and channel available quota in the historical data, and to determine whether it is necessary to adjust the RTA strategy rule currently being executed by the target channel.

[0138] The real-time RTA strategy rule acquisition module 5 is used to determine the value of the fusion model score in the real-time RTA strategy rule based on historical data if it is necessary to adjust the RTA strategy rule currently being executed by the target channel; otherwise, it keeps the RTA strategy rule currently being executed by the target channel as the real-time RTA strategy rule.

[0139] The real-time RTA policy rule output execution module 6 is used to obtain and execute real-time RTA policy rules when the media corresponding to the target channel initiates an RTA service query.

[0140] In this embodiment, the step of obtaining all channels currently executing RTA policy rules and all RTA policy rules corresponding to each channel includes:

[0141] Get the names of all channels currently executing the RTA policy rules, including Tencent, Toutiao, and Baidu;

[0142] Obtain all RTA strategy rules corresponding to each channel and the value of the fusion model score in each RTA strategy rule; each RTA strategy rule has its own focus;

[0143] Establish a database by mapping the channel name to the values ​​of the fusion model points in the corresponding executed RTA strategy rules.

[0144] In this embodiment, the convolutional neural network model compares the channel advertising spending and available channel credit within the first time period before the current time with the channel advertising spending and available channel credit in historical data to determine whether the RTA strategy rule currently being executed by the target channel needs to be adjusted. This includes the following steps:

[0145] The ratio of channel advertising spending to available channel credit within the first hour preceding the current time is calculated as the real-time consumption rate parameter.

[0146] The ratio of channel advertising spending to available channel credit within a second time period prior to the current time in historical data is used as the first baseline spending rate parameter; the second time period is longer than the first time period.

[0147] When the real-time consumption rate parameter is greater than the consumption rate parameter, it is determined that the RTA strategy rule currently being executed by the target channel needs to be adjusted.

[0148] In this embodiment, when the real-time consumption rate parameter is less than or equal to the first baseline consumption rate parameter, the method further includes:

[0149] The ratio of channel advertising spending to available channel credit within a third time period prior to the current time in historical data is used as the second baseline consumption rate parameter; the third time period is longer than the second time period.

[0150] When the real-time consumption rate parameter is greater than the consumption rate parameter, it is determined that the RTA strategy rule currently being executed by the target channel needs to be adjusted.

[0151] When the real-time consumption rate parameter is less than or equal to the second baseline consumption rate parameter, it is determined that there is no need to adjust the RTA strategy rule currently being executed by the target channel.

[0152] In this embodiment, if it is necessary to adjust the RTA policy rule currently being executed by the target channel, the step of determining the value of the fusion model score in the real-time RTA policy rule based on historical data includes:

[0153] Obtain the value and level of the available quota for the target channel corresponding to the RTA policy rule currently being executed;

[0154] The available credit limit for the target channel is set to increase to the value within the next level;

[0155] Set the value of the fusion model score in the real-time RTA strategy rule based on the latest value of the available quota of the target channel.

[0156] In this embodiment, after obtaining the real-time RTA policy rules and executing the steps when the media corresponding to the target channel initiates an RTA service query, the method further includes:

[0157] Determine whether to participate in the bidding and respond to RTA service inquiries initiated by the media corresponding to the target channel;

[0158] Collect data on channel advertising spending and available credit after the target channel executes the real-time RTA strategy rules, and feed this data back to the convolutional neural network model;

[0159] The convolutional neural network model uses collected channel advertising spending and available channel credit data as historical data to iteratively update the real-time RTA strategy rules.

[0160] In this embodiment, the convolutional neural network model uses collected channel advertising spending and available channel credit data as historical data to iteratively update the real-time RTA strategy rules, including:

[0161] Obtain the available budget of the target channel before the value is increased, and the channel advertising consumption amount corresponding to the increase to the next level, and compare them;

[0162] The first quota value is the value of the available quota corresponding to the maximum advertising expenditure of the channel.

[0163] Select multiple second credit limits within the range of [first credit limit value - first threshold, first credit limit value + first threshold].

[0164] The available quota of the target channel is updated based on multiple second quota values. The values ​​are set as multiple experimental RTA strategy rules, and the value of the fusion model score is set in each experimental RTA strategy rule.

[0165] After executing all experimental RTA strategy rules on a weekly basis, the channel advertising consumption amount and available channel credit data after executing each experimental RTA strategy rule are collected and fed back to the convolutional neural network model.

[0166] The available credit for the channel corresponding to the maximum channel advertising expenditure in all experimental RTA strategy rules is compared and used as the credit value for the second iteration update;

[0167] This process is repeated until the available credit for the channel corresponding to the maximum channel advertising spending is obtained, which is then used as the final credit limit.

[0168] The aforementioned device for controlling the execution of RTA strategy rules incorporates a convolutional neural network model to achieve real-time adjustment of RTA strategy rules based on channel advertising spending and available channel credit. This reduces the workload of business personnel, minimizes the possibility of errors in manual strategy configuration, and ensures the accuracy of RTA service's judgment on whether a user has participated by adjusting RTA strategy rules in a timely manner.

[0169] Specific limitations regarding the apparatus for controlling the execution of RTA policy rules can be found in the limitations regarding the methods for controlling the execution of RTA policy rules described above, and will not be repeated here. Each module in the aforementioned apparatus for controlling the execution of RTA policy rules can be implemented entirely or partially through software, hardware, or a combination thereof. These modules can be embedded in or independent of the processor in a computer device in hardware form, or stored in the memory of a computer device in software form, so that the processor can call and execute the operations corresponding to each module.

[0170] In one embodiment, a computer device is provided, which may be a server, and its internal structure diagram may be as follows: Figure 10As shown, the computer device includes a processor, memory, network interface, and database connected via a system bus. The processor provides computing and control capabilities. The memory includes non-volatile storage media and internal memory. The non-volatile storage media stores the operating system, computer programs, and database. The internal memory provides an environment for the operation of the operating system and computer programs in the non-volatile storage media. The database stores data controlling the execution of RTA policy rules. The network interface communicates with external terminals via a network connection. When executed by the processor, the computer program implements a method for controlling the execution of RTA policy rules.

[0171] Those skilled in the art will understand that Figure 10 The structure shown is merely a block diagram of a portion of the structure related to the present application and does not constitute a limitation on the computer device to which the present application is applied. Specific computer devices may include more or fewer components than those shown in the figure, or combine certain components, or have different component arrangements.

[0172] 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 executes the computer program to perform the following steps:

[0173] Get all channels currently executing RTA policy rules and all RTA policy rules corresponding to each channel;

[0174] The system obtains the channel advertising spending amount and available credit for the first time period before the current time, as well as the corresponding RTA strategy rules, and inputs them into the convolutional neural network model.

[0175] The system obtains historical data on advertising spending and available credit for the target channel, along with the corresponding RTA strategy rules, and inputs them into a convolutional neural network model. The available credit for the channel includes the initial loan amount and / or the initial credit limit granted by the channel.

[0176] The convolutional neural network model compares the channel advertising consumption amount and available channel credit in the first time period before the current time with the channel advertising consumption amount and available channel credit in historical data to determine whether the RTA strategy rules currently being executed by the target channel need to be adjusted.

[0177] If it is necessary to adjust the RTA strategy rules currently being executed by the target channel, then the value of the fusion model score in the real-time RTA strategy rules is determined based on historical data; if it is not necessary to adjust the RTA strategy rules currently being executed by the target channel, then the RTA strategy rules currently being executed by the target channel are kept as real-time RTA strategy rules.

[0178] When the media corresponding to the target channel initiates an RTA service inquiry, the real-time RTA policy rules are obtained and executed.

[0179] In this embodiment, the step of obtaining all channels currently executing RTA policy rules and all RTA policy rules corresponding to each channel includes:

[0180] Get the names of all channels currently executing the RTA policy rules, including Tencent, Toutiao, and Baidu;

[0181] Obtain all RTA strategy rules corresponding to each channel and the value of the fusion model score in each RTA strategy rule; each RTA strategy rule has its own focus;

[0182] Establish a database by mapping the channel name to the values ​​of the fusion model points in the corresponding executed RTA strategy rules.

[0183] In this embodiment, the convolutional neural network model compares the channel advertising spending and available channel credit within the first time period before the current time with the channel advertising spending and available channel credit in historical data to determine whether the RTA strategy rule currently being executed by the target channel needs to be adjusted. This includes the following steps:

[0184] The ratio of channel advertising spending to available channel credit within the first hour preceding the current time is calculated as the real-time consumption rate parameter.

[0185] The ratio of channel advertising spending to available channel credit within a second time period prior to the current time in historical data is used as the first baseline spending rate parameter; the second time period is longer than the first time period.

[0186] When the real-time consumption rate parameter is greater than the consumption rate parameter, it is determined that the RTA strategy rule currently being executed by the target channel needs to be adjusted.

[0187] In this embodiment, when the real-time consumption rate parameter is less than or equal to the first baseline consumption rate parameter, the method further includes:

[0188] The ratio of channel advertising spending to available channel credit within a third time period prior to the current time in historical data is used as the second baseline consumption rate parameter; the third time period is longer than the second time period.

[0189] When the real-time consumption rate parameter is greater than the consumption rate parameter, it is determined that the RTA strategy rule currently being executed by the target channel needs to be adjusted.

[0190] When the real-time consumption rate parameter is less than or equal to the second baseline consumption rate parameter, it is determined that there is no need to adjust the RTA strategy rule currently being executed by the target channel.

[0191] In this embodiment, if it is necessary to adjust the RTA policy rule currently being executed by the target channel, the step of determining the value of the fusion model score in the real-time RTA policy rule based on historical data includes:

[0192] Obtain the value and level of the available quota for the target channel corresponding to the RTA policy rule currently being executed;

[0193] The available credit limit for the target channel is set to increase to the value within the next level;

[0194] Set the value of the fusion model score in the real-time RTA strategy rule based on the latest value of the available quota of the target channel.

[0195] In this embodiment, after obtaining the real-time RTA policy rules and executing the steps when the media corresponding to the target channel initiates an RTA service query, the method further includes:

[0196] Determine whether to participate in the bidding and respond to RTA service inquiries initiated by the media corresponding to the target channel;

[0197] Collect data on channel advertising spending and available credit after the target channel executes the real-time RTA strategy rules, and feed this data back to the convolutional neural network model;

[0198] The convolutional neural network model uses collected channel advertising spending and available channel credit data as historical data to iteratively update the real-time RTA strategy rules.

[0199] In this embodiment, the convolutional neural network model uses collected channel advertising spending and available channel credit data as historical data to iteratively update the real-time RTA strategy rules, including:

[0200] Obtain the available budget of the target channel before the value is increased, and the channel advertising consumption amount corresponding to the increase to the next level, and compare them;

[0201] The first quota value is the value of the available quota corresponding to the maximum advertising expenditure of the channel.

[0202] Select multiple second credit limits within the range of [first credit limit value - first threshold, first credit limit value + first threshold].

[0203] The available quota of the target channel is updated based on multiple second quota values. The values ​​are set as multiple experimental RTA strategy rules, and the value of the fusion model score is set in each experimental RTA strategy rule.

[0204] After executing all experimental RTA strategy rules on a weekly basis, the channel advertising consumption amount and available channel credit data after executing each experimental RTA strategy rule are collected and fed back to the convolutional neural network model.

[0205] The available credit for the channel corresponding to the maximum channel advertising expenditure in all experimental RTA strategy rules is compared and used as the credit value for the second iteration update;

[0206] This process is repeated until the available credit for the channel corresponding to the maximum channel advertising spending is obtained, which is then used as the final credit limit.

[0207] For specific limitations on the steps implemented by the processor when executing a computer program, please refer to the limitations on the methods for controlling the execution of RTA policy rules mentioned above, which will not be repeated here.

[0208] In one embodiment, a computer-readable storage medium is provided having a computer program stored thereon, the computer program performing the following steps when executed by a processor:

[0209] Get all channels currently executing RTA policy rules and all RTA policy rules corresponding to each channel;

[0210] The system obtains the channel advertising spending amount and available credit for the first time period before the current time, as well as the corresponding RTA strategy rules, and inputs them into the convolutional neural network model.

[0211] The system obtains historical data on advertising spending and available credit for the target channel, along with the corresponding RTA strategy rules, and inputs them into a convolutional neural network model. The available credit for the channel includes the initial loan amount and / or the initial credit limit granted by the channel.

[0212] The convolutional neural network model compares the channel advertising consumption amount and available channel credit in the first time period before the current time with the channel advertising consumption amount and available channel credit in historical data to determine whether the RTA strategy rules currently being executed by the target channel need to be adjusted.

[0213] If it is necessary to adjust the RTA strategy rules currently being executed by the target channel, then the value of the fusion model score in the real-time RTA strategy rules is determined based on historical data; if it is not necessary to adjust the RTA strategy rules currently being executed by the target channel, then the RTA strategy rules currently being executed by the target channel are kept as real-time RTA strategy rules.

[0214] When the media corresponding to the target channel initiates an RTA service inquiry, the real-time RTA policy rules are obtained and executed.

[0215] In this embodiment, the step of obtaining all channels currently executing RTA policy rules and all RTA policy rules corresponding to each channel includes:

[0216] Get the names of all channels currently executing the RTA policy rules, including Tencent, Toutiao, and Baidu;

[0217] Obtain all RTA strategy rules corresponding to each channel and the value of the fusion model score in each RTA strategy rule; each RTA strategy rule has its own focus;

[0218] Establish a database by mapping the channel name to the values ​​of the fusion model points in the corresponding executed RTA strategy rules.

[0219] In this embodiment, the convolutional neural network model compares the channel advertising spending and available channel credit within the first time period before the current time with the channel advertising spending and available channel credit in historical data to determine whether the RTA strategy rule currently being executed by the target channel needs to be adjusted. This includes the following steps:

[0220] The ratio of channel advertising spending to available channel credit within the first hour preceding the current time is calculated as the real-time consumption rate parameter.

[0221] The ratio of channel advertising spending to available channel credit within a second time period prior to the current time in historical data is used as the first baseline spending rate parameter; the second time period is longer than the first time period.

[0222] When the real-time consumption rate parameter is greater than the consumption rate parameter, it is determined that the RTA strategy rule currently being executed by the target channel needs to be adjusted.

[0223] In this embodiment, when the real-time consumption rate parameter is less than or equal to the first baseline consumption rate parameter, the method further includes:

[0224] The ratio of channel advertising spending to available channel credit within a third time period prior to the current time in historical data is used as the second baseline consumption rate parameter; the third time period is longer than the second time period.

[0225] When the real-time consumption rate parameter is greater than the consumption rate parameter, it is determined that the RTA strategy rule currently being executed by the target channel needs to be adjusted.

[0226] When the real-time consumption rate parameter is less than or equal to the second baseline consumption rate parameter, it is determined that there is no need to adjust the RTA strategy rule currently being executed by the target channel.

[0227] In this embodiment, if it is necessary to adjust the RTA policy rule currently being executed by the target channel, the step of determining the value of the fusion model score in the real-time RTA policy rule based on historical data includes:

[0228] Obtain the value and level of the available quota for the target channel corresponding to the RTA policy rule currently being executed;

[0229] The available credit limit for the target channel is set to increase to the value within the next level;

[0230] Set the value of the fusion model score in the real-time RTA strategy rule based on the latest value of the available quota of the target channel.

[0231] In this embodiment, after obtaining the real-time RTA policy rules and executing the steps when the media corresponding to the target channel initiates an RTA service query, the method further includes:

[0232] Determine whether to participate in the bidding and respond to RTA service inquiries initiated by the media corresponding to the target channel;

[0233] Collect data on channel advertising spending and available credit after the target channel executes the real-time RTA strategy rules, and feed this data back to the convolutional neural network model;

[0234] The convolutional neural network model uses collected channel advertising spending and available channel credit data as historical data to iteratively update the real-time RTA strategy rules.

[0235] In this embodiment, the convolutional neural network model uses collected channel advertising spending and available channel credit data as historical data to iteratively update the real-time RTA strategy rules, including:

[0236] Obtain the available budget of the target channel before the value is increased, and the channel advertising consumption amount corresponding to the increase to the next level, and compare them;

[0237] The first quota value is the value of the available quota corresponding to the maximum advertising expenditure of the channel.

[0238] Select multiple second credit limits within the range of [first credit limit value - first threshold, first credit limit value + first threshold].

[0239] The available quota of the target channel is updated based on multiple second quota values. The values ​​are set as multiple experimental RTA strategy rules, and the value of the fusion model score is set in each experimental RTA strategy rule.

[0240] After executing all experimental RTA strategy rules on a weekly basis, the channel advertising consumption amount and available channel credit data after executing each experimental RTA strategy rule are collected and fed back to the convolutional neural network model.

[0241] The available credit for the channel corresponding to the maximum channel advertising expenditure in all experimental RTA strategy rules is compared and used as the credit value for the second iteration update;

[0242] This process is repeated until the available credit for the channel corresponding to the maximum channel advertising spending is obtained, which is then used as the final credit limit.

[0243] For specific limitations on the steps implemented when a computer program is executed by a processor, please refer to the limitations on the methods for controlling the execution of RTA policy rules mentioned above, which will not be repeated here.

[0244] Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium, and when executed, it can include the processes of the embodiments of the above methods. Any references to memory, storage, databases, 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. By way of illustration and not limitation, RAM is available in various forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), dual data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link DRAM (SLDRAM), Rambus direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM), etc.

[0245] The technical features of the above embodiments can be combined in any way. For the sake of brevity, 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.

[0246] The embodiments described above are merely illustrative of several implementation methods of this application, and while the descriptions are relatively specific and detailed, they should not be construed as limiting the scope of the invention patent. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of this application, and these all fall within the protection scope of this application. Therefore, the protection scope of this patent application should be determined by the appended claims.

Claims

1. A method for controlling the execution of RTA policy rules, characterized in that, include: Get all channels currently executing RTA policy rules and all RTA policy rules corresponding to each channel; The system obtains the channel advertising spending amount and available credit for the first time period before the current time, as well as the corresponding RTA strategy rules, and inputs them into the convolutional neural network model. The system obtains historical data on advertising spending and available credit for the target channel, along with the corresponding RTA strategy rules, and inputs them into a convolutional neural network model. The available credit for the channel includes the initial loan amount and / or the initial credit limit granted by the channel. The convolutional neural network model compares the channel advertising consumption amount and available channel credit in the first time period before the current time with the channel advertising consumption amount and available channel credit in historical data to determine whether the RTA strategy rules currently being executed by the target channel need to be adjusted. If yes, then determine the value of the fusion model score in the real-time RTA strategy rule based on historical data; if no, then keep the RTA strategy rule currently executed by the target channel as the real-time RTA strategy rule. When the media corresponding to the target channel initiates an RTA service inquiry, the real-time RTA policy rules are obtained and executed. The convolutional neural network model compares the channel advertising spending and available channel credit within the first time period before the current time with the channel advertising spending and available channel credit in historical data to determine whether the RTA strategy rules currently being executed by the target channel need to be adjusted. This includes the following steps: The ratio of channel advertising spending to available channel credit within the first hour preceding the current time is calculated as the real-time consumption rate parameter. The ratio of channel advertising spending to available channel credit within a second time period prior to the current time in historical data is used as the first baseline spending rate parameter; the second time period is longer than the first time period. When the real-time consumption rate parameter is greater than the first baseline consumption rate parameter, it is determined that the RTA strategy rule currently being executed by the target channel needs to be adjusted. When the real-time consumption rate parameter is less than or equal to the first baseline consumption rate parameter, the method further includes: The ratio of channel advertising spending to available channel credit within a third time period prior to the current time in historical data is used as the second baseline consumption rate parameter; the third time period is longer than the second time period. When the real-time consumption rate parameter is greater than the second baseline consumption rate parameter, it is determined that the RTA strategy rule currently being executed by the target channel needs to be adjusted. When the real-time consumption rate parameter is less than or equal to the second baseline consumption rate parameter, it is determined that there is no need to adjust the RTA strategy rule currently being executed by the target channel.

2. The method for controlling the execution of RTA policy rules according to claim 1, characterized in that, If it is necessary to adjust the RTA strategy rules currently being implemented for the target channel, the steps for determining the value of the fusion model score in the real-time RTA strategy rules based on historical data include: Obtain the value and level of the available quota for the target channel corresponding to the RTA policy rule currently being executed; The available credit limit for the target channel is set to increase to the value within the next level; Set the value of the fusion model score in the real-time RTA strategy rule based on the latest value of the available quota of the target channel.

3. The method for controlling the execution of RTA policy rules according to claim 2, characterized in that, After obtaining the real-time RTA policy rules and executing the steps when the media corresponding to the target channel initiates an RTA service query, the process further includes: Determine whether to participate in the bidding and respond to RTA service inquiries initiated by the media corresponding to the target channel; Collect data on channel advertising spending and available credit after the target channel executes the real-time RTA strategy rules, and feed this data back to the convolutional neural network model; The convolutional neural network model uses collected channel advertising spending and available channel credit data as historical data to iteratively update the real-time RTA strategy rules.

4. The method for controlling the execution of RTA policy rules according to claim 3, characterized in that, The convolutional neural network model uses collected channel advertising spending and available channel credit data as historical data to iteratively update real-time RTA strategy rules, including: Obtain the available budget of the target channel before the value is increased, and the channel advertising consumption amount corresponding to the increase to the next level, and compare them; The first quota value is the value of the available quota corresponding to the maximum advertising expenditure of the channel. Select multiple second credit limits within the range of [first credit limit value - first threshold, first credit limit value + first threshold]. The available quota of the target channel is updated based on multiple second quota values. The values ​​are set as multiple experimental RTA strategy rules, and the value of the fusion model score is set in each experimental RTA strategy rule. After executing all experimental RTA strategy rules on a weekly basis, the channel advertising consumption amount and available channel credit data after executing each experimental RTA strategy rule are collected and fed back to the convolutional neural network model. The available credit for the channel corresponding to the maximum channel advertising expenditure in all experimental RTA strategy rules is compared and used as the credit value for the second iteration update; This process is repeated until the available credit for the channel corresponding to the maximum channel advertising spending is obtained, which is then used as the final credit limit.

5. The method for controlling the execution of RTA policy rules according to claim 1, characterized in that, The step of obtaining all channels currently executing RTA policy rules and all RTA policy rules corresponding to each channel includes: Get the names of all channels currently executing the RTA policy rules; Obtain all RTA strategy rules corresponding to each channel and the value of the fusion model score in each RTA strategy rule; Establish a database by mapping the channel name to the values ​​of the fusion model points in the corresponding executed RTA strategy rules.

6. An apparatus for controlling the execution of RTA policy rules, characterized in that, A method for implementing control RTA policy rule execution according to any one of claims 1 to 5, the apparatus comprising: The current RTA policy rule acquisition module is used to acquire all channels currently executing RTA policy rules and all RTA policy rules corresponding to each channel; The current execution strategy data acquisition module is used to obtain the channel advertising consumption amount and available quota of the target channel in the first time period before the current time, as well as the corresponding RTA strategy rules, and input them into the convolutional neural network model; The historical execution strategy data acquisition module is used to acquire the channel advertising consumption amount and available credit of the target channel from historical data, as well as the corresponding RTA strategy rules, and input them into the convolutional neural network model; the available credit of the channel includes the channel's first loan amount and / or the channel's initial credit limit. The strategy rule adjustment judgment module is used by the convolutional neural network model to compare the channel advertising consumption amount and channel available quota in the first time period before the current time with the channel advertising consumption amount and channel available quota in historical data, and to determine whether it is necessary to adjust the RTA strategy rule currently being executed by the target channel. The real-time RTA strategy rule acquisition module is used to determine the value of the fusion model part in the real-time RTA strategy rule based on historical data if it is necessary to adjust the RTA strategy rule currently being executed by the target channel; otherwise, it keeps the RTA strategy rule currently being executed by the target channel as the real-time RTA strategy rule. The Real-Time RTA Policy Rule Output and Execution Module is used to obtain and execute real-time RTA policy rules when the media corresponding to the target channel initiates an RTA service query.

7. 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, it implements the steps of the method according to any one of claims 1 to 5.

8. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by a processor, it implements the steps of the method according to any one of claims 1 to 5.

Images