Advertising data processing, model deployment method, device, electronic device and storage medium

Abstract

The embodiments of the present invention provide an advertising data processing and model deployment method, device, electronic device and storage medium. The advertising data processing method includes inputting the acquired usage data of the target platform into a pre-trained cache capacity prediction model to obtain the predicted cache capacity of the advertising cache space, wherein the advertising cache space is used to cache the advertisements to be displayed on the target platform; and using the predicted cache capacity, dynamically configuring the advertising cache space to display the advertisements to be displayed. In the scheme of the embodiment of the present invention, the pre-trained cache capacity prediction model can effectively and dynamically predict a reasonable cache capacity based on the usage data of the target platform, thereby improving the configuration of the advertising cache space using the cache capacity, realizing the utilization efficiency of the cache space, and thereby improving the display efficiency of advertisements.

Description

Technical Field

[0001] Embodiments of the present invention relate to the field of computer technology, and in particular to an advertisement data processing and model deployment method, device, electronic device, and storage medium. Background Art

[0002] In recent years, with the development of internet technology, traditional online advertising methods have become unsuitable for new media scenarios. Consequently, paid ranking advertising has emerged and continues to develop. This format allows for targeted advertising and generates substantial advertising revenue for media outlets. In the mobile internet landscape, mobile media developers can connect their mobile media ad spaces to mobile advertising platforms, and advertisers can obtain matching media ad spaces through paid ranking, thus enabling the distribution, placement, and commercialization of advertising content through mobile advertising platforms.

[0003] However, in this advertising content distribution model, there is still room for improvement in the advertising distribution efficiency of existing mobile advertising platforms. Summary of the Invention

[0004] In view of this, embodiments of the present invention provide an advertisement data processing and model deployment method, apparatus, electronic device, and storage medium to solve or alleviate the above-mentioned problems.

[0005] According to a first aspect of an embodiment of the present invention, there is provided an advertising data processing method, comprising: inputting the acquired usage data of a target platform into a pre-trained cache capacity prediction model to obtain a predicted cache capacity of an advertising cache space, wherein the advertising cache space is used to cache advertisements to be displayed on the target platform; and utilizing the predicted cache capacity to dynamically configure the advertising cache space to display the advertisements to be displayed.

[0006] According to a second aspect of an embodiment of the present invention, a model deployment method is provided, comprising: obtaining a training sample, the training sample comprising usage data obtained from a target platform and a label of the usage data, the label of the usage data corresponding to a labeled cache capacity of an advertising cache space configured on the target platform; performing classification training based on the training sample to obtain a cache capacity prediction model; and deploying the cache capacity prediction model on the target platform.

[0007] According to a third aspect of an embodiment of the present invention, an advertising data processing device is provided, comprising: a prediction module, which inputs the acquired usage data of the target platform into a pre-trained cache capacity prediction model to obtain a predicted cache capacity of the advertising cache space, wherein the advertising cache space is used to cache the advertisements to be displayed of the target platform; and a configuration module, which utilizes the predicted cache capacity to dynamically configure the advertising cache space to display the advertisements to be displayed.

[0008] According to a fourth aspect of an embodiment of the present invention, a model deployment device is provided, including: an acquisition module, which acquires training samples, wherein the training samples include usage data acquired from a target platform and a label of the usage data, and the label of the usage data corresponds to the annotated cache capacity of the advertising cache space configured on the target platform; a training module, which performs classification training based on the training samples to obtain a cache capacity prediction model; and a deployment module, which deploys the cache capacity prediction model on the target platform.

[0009] According to a fifth aspect of an embodiment of the present invention, an electronic device is provided, comprising: a processor, a memory, a communication interface and a communication bus, wherein the processor, the memory and the communication interface communicate with each other via the communication bus; the memory is used to store at least one executable instruction, and the executable instruction enables the processor to perform an operation corresponding to the method described in the first aspect or the second aspect.

[0010] According to a sixth aspect of an embodiment of the present invention, a storage medium is provided, on which a computer program is stored. When the program is executed by a processor, the method according to the first aspect or the second aspect is implemented.

[0011] In the solution of the embodiment of the present invention, the pre-trained cache capacity prediction model can effectively and dynamically predict the reasonable cache capacity based on the usage data of the target platform, thereby improving the configuration of advertising cache space using the cache capacity, realizing the utilization efficiency of the cache space, and further improving the display efficiency of advertisements. BRIEF DESCRIPTION OF THE DRAWINGS

[0012] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are only some embodiments recorded in the embodiments of the present invention. For ordinary technicians in this field, other drawings can also be obtained based on these drawings.

[0013] Figure 1 A schematic diagram of a software architecture for advertising data processing of an example mobile advertising platform;

[0014] Figure 2A is a schematic flow chart of an advertisement data processing method according to an embodiment of the present invention;

[0015] Figure 2B is a schematic diagram of an advertisement data processing method according to an embodiment of the present invention;

[0016] Figure 3 is a schematic diagram of an advertisement data processing method according to another embodiment of the present invention;

[0017] Figure 4 A schematic flowchart of a model deployment method according to another embodiment of the present invention;

[0018] Figure 5 is a schematic diagram of a cache capacity prediction model according to another embodiment of the present invention;

[0019] Figure 6 is a schematic block diagram of an advertisement data processing apparatus according to another embodiment of the present invention;

[0020] Figure 7 A schematic block diagram of a model deployment apparatus according to another embodiment of the present invention;

[0021] Figure 8 This is a hardware structure of an electronic device according to another embodiment of the present invention. DETAILED DESCRIPTION

[0022] In order to enable those skilled in the art to better understand the technical solutions in the embodiments of the present invention, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of them. All other embodiments obtained by those skilled in the art based on the embodiments in the embodiments of the present invention should fall within the scope of protection of the embodiments of the present invention.

[0023] The specific implementation of the embodiment of the present invention is further described below with reference to the accompanying drawings of the embodiment of the present invention. Figure 1 The following is a schematic diagram of a software architecture for advertising data processing in an exemplary mobile advertising platform. The specific implementation of the embodiment of the present invention will be further described below in conjunction with the accompanying drawings of the embodiment of the present invention. Figure 1This diagram illustrates the software architecture for advertising data processing on an example mobile advertising platform. User device 100 includes a network interface 110, an application 120, an advertising module 121, and a human interface 130. Advertising resource server 200 includes a network interface 210 and multiple advertising resources 1, 2, ..., and N for delivery. These advertising resources can be further sent to advertising module 121 on user device 100 through methods such as competitive bidding. The software module that provides advertising resources 1, 2, ..., and N can be referred to as an advertising return mechanism. Advertising module 121 can be installed in the operating system of user device 100 or in at least one application via a method such as an SDK (Software Development Kit). Advertising module 121 can directly request ads that have been successfully ranked by competitive bidding from the advertising return mechanism and cache them in a corresponding cache area. Advertising module 121 can also request at least one successfully ranked ad from advertising resource server 200 in response to an ad cache request issued by an application. In addition, the advertising module 121 can display the cached advertisements in the cache space to the advertisement slots in the application according to the advertisement display request of the application. The proportion of the displayed advertisement materials in the total advertisement materials in the cache pool can be referred to as the material utilization rate. A higher material utilization rate indicates a higher utilization of the advertisement materials and less waste. In addition, the proportion of the number of advertisements that can be returned each time the advertisement return device is requested to return to the total number of requests can be referred to as the advertisement fill rate. A higher advertisement fill rate indicates a higher advertisement replenishment efficiency. In other words, a higher advertisement fill rate indicates a higher SDK advertisement data processing performance.

[0024] For example, if a media ad slot is triggered 100 times during a certain period of time, then 100 ads are available. If an ad is not successfully displayed once, then the corresponding fill rate is (100-1) / 100, or 99%. Due to limitations such as the media's ad requests (supply) and advertisers' ad demands (demand) not always being equal, or due to limitations such as ad matching, not every ad request may return and display the corresponding ad material, which is an insufficient ad fill rate. Therefore, material utilization and ad fill rate are relatively important advertising content distribution efficiency indices for mobile advertising platforms.

[0025] To ensure a certain ad fill rate and material utilization, you can set the cache number. However, if the cache number is too short, the ad fill rate will be affected, and if it is too long, the utilization rate of the ad material will be affected.

[0026] Figure 2A The figure is a schematic flow chart of an advertisement data processing method according to an embodiment of the present invention. Figure 2AThe advertising data processing method is applicable to any appropriate electronic device with data processing capabilities, including but not limited to mobile terminals (such as mobile phones, PADs, etc.) and PCs, etc. It includes:

[0027] 210: Input the acquired usage data of the target platform into a pre-trained cache capacity prediction model to obtain a predicted cache capacity of the advertisement cache space, wherein the advertisement cache space is used to cache advertisements to be displayed on the target platform.

[0028] It should be understood that the target platform can be an operating system installed on a terminal device such as a mobile terminal, or it can be an application installed on the above-mentioned terminal device. The advertising cache space can be managed by a module in the target platform or a function of the target platform (for example, an installed advertising SKD), for example, managed by an advertising module. The advertising module can be responsible for the request, cache, rendering or dotting of advertisements, etc. For example, the advertising module can send an advertisement display request to request and read cached advertisements from the advertisement cache space for display. The advertisement display request can be triggered by any means such as page updates, operations at specific screen positions, arrival at specific time points, etc. In addition, the advertising module can also send a preloaded cached advertisement request to the backend or middle-end server, and the multiple advertisement return devices deployed by the server corresponding to multiple advertising delivery demanders can respond to the above-mentioned preloaded cached advertisement request to perform operations such as bidding ranking, and return the advertisements that have successfully bid to the advertisement cache space.

[0029] 220: Using the predicted cache capacity, dynamically configure the ad cache space to display the ads to be displayed.

[0030] It should be understood that the configured ad cache space has a predicted cache capacity. In other words, when the ad cache space has cached ads that meet the predicted cache capacity, the ad module stops sending cached ad preload requests and can send ad display requests. When the ad cache space has remaining cached ads, but the number of remaining cached ads is less than the predicted cache capacity, the ad module can send cached ad preload requests and ad display requests. When the ad cache space has no remaining cached ads, the ad module stops sending ad display requests and can send cached ad preload requests.

[0031] In the solution of the embodiment of the present invention, the pre-trained cache capacity prediction model can effectively and dynamically predict the reasonable cache capacity based on the usage data of the target platform, thereby improving the configuration of advertising cache space using the cache capacity, realizing the utilization efficiency of the cache space, and further improving the display efficiency of advertisements.

[0032] It should also be understood that the advertising data processing method can be applied to Internet of Things (IoT) devices, which can be installed with embedded operating systems or real-time operating systems. These operating systems can also be installed with applications or software-defined devices (SKDs). IoT devices interact with corresponding advertising resource servers and voice recognition servers.

[0033] In one example, an IoT device receives a voice command and, in response to the voice command, sends a voice recognition request to a voice recognition server. The IoT device receives a recognition result returned in response to the voice recognition request, triggering a request to display an ad that matches the recognition result. For example, the recognition result may be thematically or keyword-relevant to the ad.

[0034] In another example, the IoT device receives a voice command, and in response to the voice command, triggers a display request for an advertisement to be displayed, and sends a voice recognition request to a voice recognition server.

[0035] In another example, a speech recognition server deploys a pre-trained cache capacity prediction model. The speech recognition server inputs historical speech recognition data into the cache capacity prediction model to obtain the current predicted cache capacity. The speech recognition server then sends information indicating the current predicted cache capacity to an IoT device. The IoT device then configures ad cache space based on the current predicted cache capacity.

[0036] In another example, a speech recognition server sends a historical speech recognition request to an IoT device. The IoT device is deployed with a pre-trained cache capacity prediction model. The IoT device inputs the historical speech recognition data into the cache capacity prediction model to obtain the current predicted cache capacity. The IoT device configures the advertising cache space based on the current predicted cache capacity. In addition, the IoT device can receive a voice command and, in response to the voice command, send a speech recognition request to the speech recognition server. The IoT device receives the recognition result returned in response to the speech recognition request and triggers a display request for the advertisement to be displayed that matches the recognition result. For example, the recognition result has a topic relevance or a keyword relevance to the advertisement to be displayed. For another example, the IoT device can receive a voice command and, in response to the voice command, trigger a display request for the advertisement to be displayed and send a speech recognition request to the speech recognition server.

[0037] In another example, a speech recognition server sends historical speech recognition data to an advertising resource server, which is equipped with a pre-trained cache capacity prediction model. The advertising resource server inputs the historical speech recognition data into the cache capacity prediction model, obtains the current predicted cache capacity, and sends it to the IoT device. Figure 2BThis is a schematic diagram of an advertising data processing method according to an embodiment of the present invention. As shown in the figure, the application's advertising request module requests advertisements from the advertisement return device for cached advertising space. Multiple advertisement return devices 1, 2, ..., and N can be deployed in an advertising resource server such as a backend or a middle end, for example. Figure 1 The advertising resource service end 200 shown. Multiple advertising return devices can correspond to the resources of different advertising demand parties. The advertising request module and the advertising cache space of the application together form the advertising module of the application. The application is installed in a terminal device such as a mobile terminal. As shown in the figure, Advertisement 1 and Advertisement 2 already existing in the advertising cache space can be displayed in the advertising position of the application in response to the advertising request instruction sent by the advertising request module. In addition, the empty advertising cache space can preload the cached advertisements returned by the advertising return device in response to the advertising preloading instruction sent by the advertising request module. It should be understood that the number of cached advertisements and advertising return devices mentioned above are schematic, and this example does not limit this.

[0038] In another implementation of the present invention, the usage data of the target platform includes user data of users using the target platform and platform data of the target platform, wherein the acquired usage data of the target platform is input into a pre-trained cache capacity prediction model to obtain the predicted cache capacity of the advertising cache space, including: using the user data as the first input of the cache capacity prediction model, and using the platform data as the second input of the cache capacity prediction model, and outputting the predicted cache capacity.

[0039] Since the platform data of the target platform can reflect the platform characteristics related to the utilization rate of advertising materials, and the user data can reflect the user characteristics of the using users related to the utilization rate of advertising materials, the first input based on user data and the second input based on platform data can improve the accuracy of dynamic prediction of the cache capacity prediction model.

[0040] In another implementation of the present invention, the cache capacity prediction model includes a first sub-model, a second sub-model and a third sub-model, wherein the first input is used as the input of the first sub-model, the second input is used as the input of the second sub-model, and the output of the first sub-model and the output of the second sub-model are used as the input of the third sub-model, wherein user data is used as the first input of the cache capacity prediction model, and platform data is used as the second input of the cache capacity prediction model, and the predicted cache capacity is output, including: inputting user data into the first sub-model, inputting platform data into the second sub-model, and inputting the predicted cache capacity from the third sub-model.

[0041] Since the cache capacity prediction model includes a first sub-model, a second sub-model, and a third sub-model, and the first input serves as the input of the first sub-model and the second input serves as the input of the second sub-model, it is possible to achieve relatively independent training and prediction processes for platform data and user data, thereby improving the model deployment efficiency.

[0042] It should be understood that the IoT device can provide feedback to the advertising resource server regarding the percentage of time that the number of currently cached ads is less than the current predicted cache capacity. The advertising resource server can then update the cache capacity prediction model based on this percentage of time. Furthermore, for each of the above examples, the cache capacity prediction model can be updated on the server or IoT device where it is deployed.

[0043] In addition, the cache capacity prediction model can also use historical speech recognition data input by the speech recognition server to update the cache capacity prediction model.

[0044] In addition, alternatively, the voice recognition server may also be a gesture recognition server or other recognition server, which is not limited in the embodiment of the present invention.

[0045] Additionally, alternatively, the historical voice recognition data may be historical gesture recognition data or other operation data.

[0046] In addition, alternatively, the cache capacity prediction model can be any one of the first sub-model, the second sub-model or the third sub-model.

[0047] In another implementation of the present invention, the method further includes: reporting current usage data of the target platform to obtain an updated cache capacity prediction model, where the updated cache capacity prediction model is obtained by training with training samples based on the current usage data.

[0048] Since an updated cache capacity prediction model is generated using updated training samples based on current usage data of the target platform, real-time updating of the model is achieved.

[0049] It should be understood that the training labels for the first and second sub-models, as well as the cached ad capacity label for the third sub-model, can be determined based on the advertising material utilization rate. It should also be understood that the server can train and update the cache capacity prediction model as a whole. It should also be understood that the server can train and update the first, second, and third sub-models separately.

[0050] Figure 3 FIG. 4 is a schematic diagram of an advertisement data processing method according to another embodiment of the present invention.

[0051] 310: Input the acquired usage data of the target platform into a pre-trained cache capacity prediction model to obtain a predicted cache capacity of the advertisement cache space, wherein the advertisement cache space is used to cache advertisements to be displayed on the target platform.

[0052] 320: Using the predicted cache capacity, dynamically configure the ad cache space to display the ads to be displayed.

[0053] 330: Report the current ad display data and the current remaining ad data in the ad cache space to calculate the current ad utilization rate of the target platform.

[0054] 340: Obtain an updated cache capacity prediction model, where the updated cache capacity prediction model is obtained by training a training sample based on current advertisement utilization.

[0055] Since the cache capacity prediction model can be updated based on the reported current advertising display data and the current remaining advertising data in the advertising cache space, the current advertising utilization of the target platform, and the training samples based on the current advertising utilization, the dynamic prediction capability of the cache capacity prediction model is more accurate.

[0056] Figure 4 This is a schematic flowchart of a model deployment method according to another embodiment of the present invention. Figure 4 The model deployment method is applicable to any appropriate electronic device with data processing capabilities, including but not limited to servers such as private clouds, public clouds, or hybrid clouds. The model deployment method includes:

[0057] 410: Acquire training samples, where the training samples include usage data acquired from the target platform and labels of the usage data, where the labels of the usage data correspond to the annotated cache capacity of the advertisement cache space configured on the target platform.

[0058] 420: Perform classification training based on the training samples to obtain a cache capacity prediction model.

[0059] 430: Deploy the cache capacity prediction model on the target platform.

[0060] It should be understood that the server may include a model management module and an advertising resource management module (for example, an advertising resource database). The advertising resource management module may include the above-mentioned multiple advertising return devices. The model management module and the advertising resource module may be independent or may be configured in the same server. Among them, the deployment and training of the cache capacity prediction model can be managed by the model management module. When the server receives a cache advertisement preloading request from the advertising module in the application or operating system, the advertising resource management module is called to perform bidding ranking, and the advertisements that have successfully bid are returned to the end side and cached in the advertising cache space. In addition, the model management module can obtain bidding ranking data from the advertising resource management module, and determine the training samples of the cache capacity prediction model based on the bidding ranking data. For example, the sum of the advertising bid prices after bidding ranking within a certain period of time is maximized. And the cache capacity prediction model is trained based on this period.

[0061] In the solution of the embodiment of the present invention, the pre-trained cache capacity prediction model can effectively and dynamically predict the reasonable cache capacity based on the usage data of the target platform, thereby improving the configuration of advertising cache space using the cache capacity, realizing the utilization efficiency of the cache space, and further improving the display efficiency of advertisements.

[0062] In another implementation of the present invention, the method further includes: obtaining current advertising display data and current remaining advertising data of the advertising cache space from the target platform; and counting the current advertising utilization rate based on the current advertising display data and the current remaining advertising data, wherein obtaining training samples includes: obtaining the current usage data of the target platform and the current labeled cache capacity corresponding to the current advertising utilization rate, and generating updated training samples, wherein classification training is performed based on the training samples to obtain a cache capacity prediction model, including: updating the cache capacity prediction model using the updated training samples.

[0063] Since the cache capacity prediction model can be updated based on the reported current advertising display data and the current remaining advertising data in the advertising cache space, the current advertising utilization of the target platform, and the training samples based on the current advertising utilization, the dynamic prediction capability of the cache capacity prediction model is more accurate.

[0064] In another implementation of the present invention, the cache capacity prediction model includes a first sub-model, a second sub-model and a third sub-model, wherein the first input serves as the input of the first sub-model, the second input serves as the input of the second sub-model, the output of the first sub-model and the output of the second sub-model serve as the input of the third sub-model, and the updated training sample includes a first updated training sub-sample, which is generated based on the current user data of the user using the target platform, wherein updating the cache capacity prediction model using the updated training sample includes: updating the first sub-model using the first updated training sub-sample.

[0065] In another implementation of the present invention, the updated training sample also includes a second updated training sub-sample, which is generated based on the platform data of the target platform. wherein, updating the cache capacity prediction model using the updated training sample also includes: updating the second sub-model using the second updated training sub-sample.

[0066] It should be understood that the training labels for the first and second sub-models, as well as the cached ad capacity label for the third sub-model, can be determined based on the advertising material utilization rate. It should also be understood that the server can train and update the cache capacity prediction model as a whole. It should also be understood that the server can train and update the first, second, and third sub-models separately.

[0067] Since the cache capacity prediction model includes a first sub-model, a second sub-model, and a third sub-model, and the first input serves as the input of the first sub-model and the second input serves as the input of the second sub-model, it is possible to achieve relatively independent training and prediction processes for platform data and user data, thereby improving the model deployment efficiency.

[0068] In another implementation of the present invention, obtaining training samples includes: obtaining current usage data of the target platform; generating updated training samples based on the current usage data, wherein classification training is performed based on the training samples to obtain a cache capacity prediction model, including: using the updated training samples to train the cache capacity prediction model to obtain an updated cache capacity prediction model.

[0069] Since an updated cache capacity prediction model is generated using updated training samples based on current usage data of the target platform, real-time updating of the model is achieved.

[0070] Figure 5 FIG. 1 is a schematic diagram of a cache capacity prediction model according to another embodiment of the present invention. Figure 5The training and deployment methods of the model and its sub-models are described. As shown in the figure, the M1 device, the M2 device, and the M3 device correspond to the first sub-model, the second sub-model, and the third sub-model, respectively. The first sub-model, the second sub-model, and the third sub-model can be trained separately based on the same or different types of neural networks. The training data for the first sub-model can include user data (e.g., user identification, user history data, etc.) and its labels, which serve as input and output respectively during classification training. The user data can be the user identification, and its label can be the first feature vector extracted for the user identification. The training data for the second sub-model can include platform data (e.g., operating system data or application data, etc.) and its labels, which serve as input and output respectively during classification training. The platform data can be the application identification, and its label can be the second feature vector extracted for the application identification. The training data for the third sub-model can include the above-mentioned labels of the first and second sub-models and the cached advertising prediction capacity. During classification training, the first and second feature vectors can be used as input, and the advertising prediction capacity can be used as output. The advertising prediction capacity can be the number of advertisements. It should be understood that the first, second, and third sub-models can be trained for classification using a feedforward neural network, such as a multilayer perceptron, a convolutional neural network, or a recurrent neural network. Furthermore, the first, second, and third sub-models have different update cycles. The update cycle of the first sub-model can be shorter than that of the second sub-model, and the update cycle of the second sub-model can be shorter than that of the third sub-model.

[0071] The interactive process of updating the model and its sub-models is explained below.

[0072] In one example, the server obtains updated training data, which may include current usage data for the target platform. For example, the current usage data may be usage data for the current update cycle, or the usage data collected during the last model training session may include usage data for the current collection time. The updated training data may also include data tags. The tags for the current usage data may be determined based on the advertising material utilization rate, such that the corresponding advertising material utilization rate exceeds a preset threshold. For example, both user data and platform data may have corresponding tags.

[0073] Furthermore, the training labels for the first and second sub-models, as well as the cached ad capacity label for the third sub-model, can be determined based on the advertising material utilization rate. Furthermore, the server can train and update the cache capacity prediction model as a whole. Furthermore, the server can train and update the first, second, and third sub-models separately.

[0074] In another example, the server obtains updated training data, which may include current ad display data and the current remaining ad data in the ad cache. For example, the current ad display data and the current remaining ad data in the ad cache may be data within the current update cycle, or the collection time during the last model training may have corresponding data during the current collection time. The updated training data may also include data labels, which may be determined based on the advertising material utilization rate, such that the corresponding advertising material utilization rate is greater than a preset threshold.

[0075] Furthermore, the training labels for the first and second sub-models, as well as the cached ad capacity label for the third sub-model, can be determined based on the advertising material utilization rate. Furthermore, the server can train and update the cache capacity prediction model as a whole. Furthermore, the server can train and update the first, second, and third sub-models separately.

[0076] Figure 6 This is a schematic block diagram of an advertisement data processing device according to another embodiment of the present invention. Figure 6 The advertising data processing device is suitable for execution by any appropriate electronic device with data processing capabilities, including but not limited to mobile terminals (such as mobile phones, PADs, etc.) and PCs. The advertising data processing device includes:

[0077] The prediction module 610 inputs the acquired usage data of the target platform into a pre-trained cache capacity prediction model to obtain a predicted cache capacity of the advertisement cache space, wherein the advertisement cache space is used to cache advertisements to be displayed on the target platform.

[0078] The configuration module 620 uses the predicted cache capacity to dynamically configure the advertisement cache space to display the advertisement to be displayed.

[0079] In the solution of the embodiment of the present invention, the pre-trained cache capacity prediction model can effectively and dynamically predict the reasonable cache capacity based on the usage data of the target platform, thereby improving the configuration of advertising cache space using the cache capacity, realizing the utilization efficiency of the cache space, and further improving the display efficiency of advertisements.

[0080] In another implementation of the present invention, the usage data of the target platform includes user data of users of the target platform and platform data of the target platform, wherein the prediction module is specifically used to: use the user data as the first input of the cache capacity prediction model, and use the platform data as the second input of the cache capacity prediction model, and output the predicted cache capacity.

[0081] In another implementation of the present invention, the cache capacity prediction model includes a first sub-model, a second sub-model and a third sub-model, wherein the first input serves as the input of the first sub-model, the second input serves as the input of the second sub-model, and the output of the first sub-model and the output of the second sub-model serve as the input of the third sub-model, wherein the prediction module is specifically used to: input user data into the first sub-model, input platform data into the second sub-model, and predict cache capacity from the third sub-model input.

[0082] In another implementation of the present invention, the device also includes: a first network module, reporting current advertising display data and current remaining advertising data in the advertising cache space, so as to count the current advertising utilization rate of the target platform; obtaining an updated cache capacity prediction model, and the updated cache capacity prediction model is obtained by training with training samples based on the current advertising utilization rate.

[0083] In another implementation of the present invention, the device also includes: a second network module, which reports the current usage data of the target platform to obtain an updated cache capacity prediction model, and the updated cache capacity prediction model is obtained by training with training samples based on the current usage data.

[0084] The device of this embodiment is used to implement the corresponding methods in the aforementioned multiple method embodiments and has the beneficial effects of the corresponding method embodiments, which will not be described in detail here. In addition, the functional implementation of each module in the device of this embodiment can refer to the description of the corresponding parts in the aforementioned method embodiments, which will not be described in detail here.

[0085] Figure 7 This is a schematic block diagram of a model deployment apparatus according to another embodiment of the present invention. Figure 7 The model deployment device is applicable to any appropriate electronic device with data processing capabilities, including but not limited to servers such as private clouds, public clouds, or hybrid clouds. The model deployment device includes:

[0086] The acquisition module 710 acquires training samples, where the training samples include usage data acquired from the target platform and labels of the usage data. The labels of the usage data correspond to the annotated cache capacity of the advertisement cache space configured on the target platform.

[0087] The training module 720 performs classification training based on the training samples to obtain a cache capacity prediction model.

[0088] The deployment module 730 deploys the cache capacity prediction model on the target platform.

[0089] In the solution of the embodiment of the present invention, the pre-trained cache capacity prediction model can effectively and dynamically predict the reasonable cache capacity based on the usage data of the target platform, thereby improving the configuration of advertising cache space using the cache capacity, realizing the utilization efficiency of the cache space, and further improving the display efficiency of advertisements.

[0090] In another implementation of the present invention, the acquisition module is further configured to obtain current ad display data and currently remaining ad data in the ad cache from the target platform. Based on the current ad display data and the currently remaining ad data, current ad utilization is calculated. The acquisition module is specifically configured to obtain current usage data of the target platform and the current annotated cache capacity corresponding to the current ad utilization, and generate updated training samples. The training module is specifically configured to update the cache capacity prediction model using the updated training samples.

[0091] In another implementation of the present invention, the cache capacity prediction model includes a first sub-model, a second sub-model and a third sub-model, wherein the first input serves as the input of the first sub-model, the second input serves as the input of the second sub-model, the output of the first sub-model and the output of the second sub-model serve as the input of the third sub-model, and the updated training sample includes a first updated training sub-sample, which is generated based on the current user data of the user using the target platform, wherein the training module is specifically used to: use the first updated training sub-sample to update the first sub-model.

[0092] In another implementation of the present invention, the updated training sample further includes a second updated training subsample, which is generated based on the platform data of the target platform, wherein the training module is further configured to update the second sub-model using the second updated training subsample.

[0093] In another implementation of the present invention, the acquisition module is specifically used to: obtain current usage data of the target platform; generate updated training samples based on the current usage data, wherein classification training is performed based on the training samples to obtain a cache capacity prediction model, including: using the updated training samples to train the cache capacity prediction model to obtain an updated cache capacity prediction model.

[0094] The device of this embodiment is used to implement the corresponding methods in the aforementioned multiple method embodiments and has the beneficial effects of the corresponding method embodiments, which will not be described in detail here. In addition, the functional implementation of each module in the device of this embodiment can refer to the description of the corresponding parts in the aforementioned method embodiments, which will not be described in detail here.

[0095] Figure 8 The hardware structure of the electronic device of another embodiment of the present invention is as follows; Figure 8As shown, the hardware structure of the electronic device may include: a processor 801, a communication interface 802, a storage medium 803 and a communication bus 804;

[0096] The processor 801, the communication interface 802, and the storage medium 803 communicate with each other via the communication bus 804;

[0097] Optionally, the communication interface 802 may be an interface of a communication module;

[0098] The processor 801 may be specifically configured to: input the acquired usage data of the target platform into a pre-trained cache capacity prediction model to obtain a predicted cache capacity of an advertisement cache space, wherein the advertisement cache space is used to cache advertisements to be displayed on the target platform; and dynamically configure the advertisement cache space using the predicted cache capacity to display the advertisements to be displayed;

[0099] Alternatively, a training sample is obtained, wherein the training sample includes usage data obtained from a target platform and a label of the usage data, wherein the label of the usage data corresponds to the annotated cache capacity of the advertising cache space configured on the target platform; classification training is performed based on the training sample to obtain a cache capacity prediction model; and the cache capacity prediction model is deployed on the target platform.

[0100] Processor 801 can be a general-purpose processor, including a central processing unit (CPU), a network processor (NP), etc.; it can also be a digital signal processor (DSP), an application-specific integrated circuit (ASIC), a field-programmable gate array (FPGA), or other programmable logic devices, discrete gate or transistor logic devices, or discrete hardware components. It can implement or execute the various methods, steps, and logic block diagrams disclosed in the embodiments of the present invention. The general-purpose processor can be a microprocessor or any conventional processor.

[0101] The storage medium 803 can be, but is not limited to, a random access memory (RAM), a read-only memory (ROM), a programmable read-only memory (PROM), an erasable programmable read-only memory (EPROM), an electrically erasable programmable read-only memory (EEPROM), etc.

[0102] In particular, according to an embodiment of the present disclosure, the process described above with reference to the flowchart can be implemented as a computer software program. For example, an embodiment of the present disclosure includes a computer program product, which includes a computer program carried on a storage medium, and the computer program includes a program code configured to execute the method shown in the flowchart. In such an embodiment, the computer program can be downloaded and installed from the network through the communication part, and / or installed from a removable medium. When the computer program is executed by a central processing unit (CPU), the above-mentioned functions defined in the method of the present invention are executed. It should be noted that the storage medium described in the present invention can be a computer-readable signal medium or a computer-readable storage medium or any combination of the above two. The storage medium can be, for example, but not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, device or device, or any combination of the above. More specific examples of computer-readable storage media may include, but are not limited to, an electrical connection having one or more conductors, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage medium, a magnetic storage medium, or any suitable combination thereof. In the present invention, a computer-readable storage medium may be any tangible medium containing or storing a program that can be used by or in conjunction with an instruction execution system, apparatus, or device. In the present invention, a computer-readable signal medium may include a data signal propagated in baseband or as part of a carrier wave, which carries computer-readable program code. Such a propagated data signal may take a variety of forms, including, but not limited to, an electromagnetic signal, an optical signal, or any suitable combination thereof. A computer-readable signal medium may also be any storage medium other than a computer-readable storage medium that can transmit, propagate, or transfer a program configured for use by or in conjunction with an instruction execution system, apparatus, or device. The program code contained on the storage medium can be transmitted using any appropriate medium, including but not limited to wireless, wireline, optical cable, RF, etc., or any suitable combination of the foregoing.

[0103] Computer program code configured to perform the operations of the present invention may be written in one or more programming languages, or a combination thereof, including object-oriented programming languages ​​such as Java, Smalltalk, C++, and conventional procedural programming languages ​​such as "C" or similar programming languages. The program code may be executed entirely on the user's computer, partially on the user's computer, as a stand-alone software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In cases involving a remote computer, the remote computer may be connected to the user's computer through any type of network, including a local area network (LAN) or a wide area network (WAN), or may be connected to an external computer (e.g., through the Internet using an Internet service provider).

[0104] The flowcharts and block diagrams in the accompanying drawings illustrate the possible architectures, functions, and operations of the systems, methods, and computer program products according to various embodiments of the present invention. In this regard, each box in the flowchart or block diagram can represent a module, program segment, or portion of code, which contains one or more executable instructions configured to implement the specified logical function. The above-mentioned specific embodiments have specific sequential relationships, but these sequential relationships are merely exemplary. During the specific implementation, these steps may be fewer, more, or the execution order may be adjusted. In other words, in some alternative implementations, the functions marked in the boxes may also occur in an order different from that marked in the accompanying drawings. For example, two boxes shown in succession can actually be executed substantially in parallel, or they can sometimes be executed in the opposite order, depending on the functions involved. It should also be noted that each box in the block diagram and / or flowchart, as well as combinations of boxes in the block diagram and / or flowchart, can be implemented using a dedicated hardware-based system that performs the specified function or operation, or can be implemented using a combination of dedicated hardware and computer instructions.

[0105] The modules involved in the embodiments of the present invention may be implemented in software or hardware, and the names of these modules do not necessarily limit the modules themselves.

[0106] As another aspect, the present invention further provides a storage medium storing a computer program, which implements the method described in the above embodiment when executed by a processor.

[0107] As another aspect, the present invention further provides a storage medium, which may be included in the apparatus described in the above embodiment; or may exist independently and not be incorporated into the apparatus. The above storage medium carries one or more programs, and when the above one or more programs are executed by the apparatus, the apparatus: inputs the acquired usage data of the target platform into a pre-trained cache capacity prediction model to obtain a predicted cache capacity of an advertisement cache space, wherein the advertisement cache space is used to cache advertisements to be displayed on the target platform; and dynamically configures the advertisement cache space using the predicted cache capacity to display the advertisements to be displayed;

[0108] Alternatively, a training sample is obtained, wherein the training sample includes usage data obtained from a target platform and a label of the usage data, wherein the label of the usage data corresponds to the annotated cache capacity of the advertising cache space configured on the target platform; classification training is performed based on the training sample to obtain a cache capacity prediction model; and the cache capacity prediction model is deployed on the target platform.

[0109] The terms "first," "second," "the first," or "the second" used in various embodiments of the present disclosure may modify various components regardless of order and / or importance, but these terms do not limit the corresponding components. The above terms are configured solely for the purpose of distinguishing an element from other elements. For example, a first user device and a second user device represent different user devices, even though both are user devices. For example, a first element may be referred to as a second element, and similarly, a second element may be referred to as a first element without departing from the scope of the present disclosure.

[0110] When one element (for example, a first element) is referred to as being “(operably or communicably) coupled” or “(operably or communicably) coupled to” or “connected to” another element (for example, a second element), it should be understood that the one element is directly connected to the other element or that the one element is indirectly connected to the other element via yet another element (for example, a third element). Conversely, it should be understood that when an element (for example, a first element) is referred to as being “directly connected” or “directly coupled” to another element (the second element), there is no element (for example, a third element) interposed therebetween.

[0111] The above description is merely an illustration of the preferred embodiments of the present invention and the technical principles employed. Those skilled in the art should understand that the scope of the invention is not limited to the technical solutions formed by the specific combination of the above-mentioned technical features, but also encompasses other technical solutions formed by any combination of the above-mentioned technical features or their equivalents without departing from the above-mentioned inventive concept. For example, a technical solution formed by replacing the above-mentioned features with (but not limited to) technical features with similar functions disclosed in the present invention.

Claims

1. A method for processing advertising data, comprising: Inputting the acquired usage data of the target platform into a pre-trained cache capacity prediction model to obtain a predicted cache capacity of an advertisement cache space, wherein the advertisement cache space is used to cache advertisements to be displayed on the target platform; Dynamically configuring the advertisement cache space using the predicted cache capacity to display the advertisement to be displayed; Reporting current advertisement display data and current remaining advertisement data in the advertisement cache space so as to calculate the current advertisement utilization rate of the target platform; An updated cache capacity prediction model is obtained, where the updated cache capacity prediction model is obtained by training with training samples based on the current advertisement utilization rate.

2. The method according to claim 1, wherein The usage data of the target platform includes user data of users of the target platform and platform data of the target platform, wherein: Inputting the acquired usage data of the target platform into a pre-trained cache capacity prediction model to obtain the predicted cache capacity of the advertisement cache space includes: The user data is used as a first input of the cache capacity prediction model, and the platform data is used as a second input of the cache capacity prediction model, and the predicted cache capacity is output.

3. The method according to claim 2, wherein: The cache capacity prediction model includes a first sub-model, a second sub-model, and a third sub-model, wherein the first input serves as the input of the first sub-model, the second input serves as the input of the second sub-model, and the output of the first sub-model and the output of the second sub-model serve as the input of the third sub-model. The method of using the user data as the first input of the cache capacity prediction model and the platform data as the second input of the cache capacity prediction model to output the predicted cache capacity includes: The user data is input into the first sub-model, the platform data is input into the second sub-model, and the predicted cache capacity is input from the third sub-model.

4. The method according to claim 1, wherein Also includes: Reporting current usage data of the target platform to obtain an updated cache capacity prediction model, wherein the updated cache capacity prediction model is obtained by training with training samples based on the current usage data.

5. A model deployment method, comprising: Obtaining a training sample, the training sample including usage data obtained from a target platform and a label of the usage data, the label of the usage data corresponding to a marked cache capacity of an advertisement cache space configured on the target platform; Performing classification training based on the training samples to obtain a cache capacity prediction model; Deploying the cache capacity prediction model on the target platform; Acquire current advertisement display data and current remaining advertisement data in the advertisement cache space from the target platform; Based on the current advertisement display data and the current remaining advertisement data, current advertisement utilization rate is counted.

6. The method according to claim 5, wherein: The obtaining of training samples includes: Obtaining current usage data of the target platform; Generating updated training samples based on the current usage data, wherein performing classification training based on the training samples to obtain a cache capacity prediction model includes: The cache capacity prediction model is trained using the updated training samples to obtain an updated cache capacity prediction model.

7. The method according to claim 5, wherein: The obtaining of training samples includes: Obtaining current usage data of the target platform and a current annotated cache capacity corresponding to the current advertisement utilization rate, and generating updated training samples, wherein performing classification training based on the training samples to obtain a cache capacity prediction model includes: The cache capacity prediction model is updated using the updated training samples.

8. The method according to claim 7, wherein: The cache capacity prediction model includes a first sub-model, a second sub-model, and a third sub-model, wherein the first input serves as the input of the first sub-model, the second input serves as the input of the second sub-model, the output of the first sub-model and the output of the second sub-model serve as the input of the third sub-model, the updated training sample includes a first updated training sub-sample, and the first updated training sub-sample is generated based on current user data of users of the target platform, wherein updating the cache capacity prediction model using the updated training sample includes: The first sub-model is updated using the first updated training sub-sample.

9. The method according to claim 8, wherein The updated training sample further includes a second updated training subsample, where the second updated training subsample is generated based on the platform data of the target platform. The updating of the cache capacity prediction model using the updated training sample further includes: The second sub-model is updated using the second updated training sub-sample.

10. An advertising data processing device, comprising: a prediction module that inputs the acquired usage data of the target platform into a pre-trained cache capacity prediction model to obtain a predicted cache capacity of an advertisement cache space, wherein the advertisement cache space is used to cache advertisements to be displayed on the target platform; a configuration module, utilizing the predicted cache capacity to dynamically configure the advertisement cache space to display the advertisement to be displayed; The first network module is used to report the current advertising display data and the current remaining advertising data of the advertising cache space, so as to count the current advertising utilization of the target platform and obtain the updated cache capacity prediction model, and the updated cache capacity prediction model is obtained by training with training samples based on the current advertising utilization.

11. A model deployment device, comprising: an acquisition module, configured to acquire a training sample, wherein the training sample includes usage data acquired from a target platform and a label of the usage data, wherein the label of the usage data corresponds to a marked cache capacity of an advertisement cache space configured on the target platform; A training module, performing classification training based on the training samples to obtain a cache capacity prediction model; A deployment module deploys the cache capacity prediction model on the target platform; The acquisition module is further used for: Acquire current advertisement display data and current remaining advertisement data in the advertisement cache space from the target platform; Based on the current advertisement display data and the current remaining advertisement data, current advertisement utilization rate is counted.

12. An electronic device, comprising: A processor, a memory, a communication interface, and a communication bus, wherein the processor, the memory, and the communication interface communicate with each other via the communication bus; The memory is used to store at least one executable instruction, where the executable instruction enables the processor to perform an operation corresponding to the method according to any one of claims 1 to 9.

13. A storage medium having a computer program stored thereon, wherein when the program is executed by a processor, the method according to any one of claims 1 to 9 is implemented.

Images