Media creative information generation method and device, computer equipment and storage medium

Abstract

The invention relates to a media creative information generation method and device, computer equipment, a storage medium and a computer program product. The method can be applied to the field of artificial intelligence and intelligent traffic scenes, and comprises the following steps: obtaining seed creative information, media exploration information and media attribute information; generating media guidance information based on the media attribute information and the seed creative information; performing information processing on the seed creative information and the media exploration information according to the media guidance information to obtain candidate media creative information; putting the candidate media creative information into a media putting system to obtain media evaluation information; and in the candidate media creativity information, selecting media creativity information based on the media evaluation information and an exploration target in the media exploration information. By adopting the method, the media creative information meeting the business requirements can be generated, so that the interaction rate of the media creative information is effectively improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of artificial intelligence, in particular to a media creative information generation method and device, computer equipment, storage medium and computer program product. BACKGROUND

[0002] With the development of computer technology, media creative information is widely used in media promotion (such as advertising), so as to improve the interaction rate of media creative information, and thus effectively improve the conversion rate of products.

[0003] In the traditional scheme, the AIGC (Artificial Intelligence Generated Content) model is usually directly used to produce media creative information, and then the produced media creative information is put into use. However, this way cannot effectively meet the business needs, so as to truly and effectively improve the interaction rate of media creative information. SUMMARY

[0004] Therefore, it is necessary to provide a media creative information generation method, device, computer equipment, computer readable storage medium and computer program product, which can generate media creative information that meets business needs, so as to effectively improve the interaction rate of media creative information.

[0005] In a first aspect, the present application provides a media creative information generation method, which comprises:

[0006] obtaining seed creative information, media exploration information and media attribute information; the media exploration information comprises an exploration target and an exploration budget value of a target object when requesting creative exploration, and the media attribute information comprises an industry and an audience group corresponding to the media;

[0007] generating media guide information based on the media attribute information and the seed creative information;

[0008] performing information processing on the seed creative information and the media exploration information according to the media guide information, to obtain candidate media creative information;

[0009] putting the candidate media creative information into a media putting system to obtain media evaluation information;

[0010] selecting media creative information from the candidate media creative information based on the media evaluation information and the exploration target in the media exploration information.

[0011] In a second aspect, the present application provides another media creative information generation method, which comprises:

[0012] sending a creative information generation request carrying seed creative information and media exploration information to a server; wherein the creative information generation request is used to instruct the server to generate media guide information based on the seed creative information and media attribute information, to perform information processing on the seed creative information and the media exploration information according to the media guide information, to obtain candidate media creative information, to put the candidate media creative information into a media putting system to obtain media evaluation information, and to select media creative information based on the media evaluation information and an exploration target in the media exploration information from the candidate media creative information; the media exploration information comprises an exploration target and an exploration budget value of a target object when requesting creative exploration, and the media attribute information comprises an industry and an audience group corresponding to a media;

[0013] receiving the media creative information returned by the server in response to the creative information generation request;

[0014] displaying the media creative information.

[0015] In a third aspect, the present application further provides a media creative information generation device, comprising:

[0016] an acquisition module configured to acquire seed creative information, media exploration information and media attribute information; the media exploration information comprises an exploration target and an exploration budget value of a target object when requesting creative exploration, and the media attribute information comprises an industry and an audience group corresponding to a media;

[0017] a first generation module configured to generate media guide information based on the media attribute information and the seed creative information;

[0018] a second generation module configured to perform information processing on the seed creative information and the media exploration information according to the media guide information to obtain candidate media creative information;

[0019] a putting module configured to put the candidate media creative information into a media putting system to obtain media evaluation information;

[0020] a selection module configured to select media creative information based on the media evaluation information and an exploration target in the media exploration information from the candidate media creative information.

[0021] In one embodiment, the first generation module is further configured to perform information processing on the media attribute information and the seed creative information by a guide generator to obtain a first creative style, a first creative quality and a first creative similarity corresponding to the seed creative information; and to splice the first creative style, the first creative quality and the first creative similarity to obtain media guide information.

[0022] In one of the embodiments, the device further comprises:

[0023] a training module, configured to acquire sample media attribute information, sample seed creative information and label information from a data source; perform information processing on the sample media attribute information and the sample seed creative information by an initial guide generator to obtain a second creative style, a second creative quality and a second creative similarity corresponding to the sample seed creative information; determine a loss value according to the second creative style, the second creative quality, the second creative similarity and the label information; and adjust parameters in the initial guide generator based on the loss value to obtain the guide generator.

[0024] In one of the embodiments, the candidate media creative information is at least one candidate creative information in a candidate media creative image or a candidate media creative script; and the device further comprises:

[0025] a verification module, configured to perform quality verification on the at least one creative information to obtain a verification result;

[0026] The delivery module is further configured to, in a case where the verification result indicates that the at least one creative information meets a delivery condition, deliver the at least one candidate creative information to a media delivery system.

[0027] In one of the embodiments, the verification module is further configured to sequentially perform image quality checking and compliance checking on the candidate media creative image to obtain an image checking result; and sequentially perform semantic detection and privacy information detection on the candidate media creative script to obtain a script detection result.

[0028] In one of the embodiments, the device further comprises:

[0029] The acquisition module is further configured to receive a creative information generation request initiated by a terminal; read seed creative information and media exploration information from the creative information generation request; and acquire media attribute information based on the creative information generation request;

[0030] a sending module, configured to send the media creative information to the terminal for display.

[0031] In one of the embodiments, the selection module is further configured to sort the candidate media creative information based on an exploration target in the media evaluation information and the media exploration information to obtain a creative information sequence; and select media creative information meeting a sequence condition in the creative information sequence.

[0032] In one of the embodiments, the device further comprises:

[0033] determining a residual budget value according to an exploration budget value and an exploration consumption value in the media exploration information; and determining a media exploration effect according to the media evaluation information and the media exploration information;

[0034] The sending module is further configured to send the media creative information, the residual budget value, and the media exploration effect to the terminal for display.

[0035] In one of the embodiments, the apparatus further comprises:

[0036] The sending module is further configured to send the media creative information, the residual budget value, and the media exploration effect to the terminal for display when the media exploration effect meets the exploration target, or the media exploration effect does not meet the exploration target and the residual budget value is less than a preset threshold value.

[0037] The processing module is configured to select at least one creative information in the creative information sequence as a new seed creative information when the media exploration effect does not meet the exploration target and the residual budget value is greater than or equal to the preset threshold value, perform a next iteration round of creative information generation based on the new seed creative information, and stop the iteration process until the media exploration effect meets the exploration target or the media exploration effect does not meet the exploration target and the residual budget value is less than the preset threshold value, thereby obtaining the media creative information.

[0038] In one of the embodiments, the candidate media creative information comprises k creative information.

[0039] The determining module is configured to determine a number of creative information to be explored in each iteration round and an iteration number in a preset total time period, determine a product value between the iteration number and the number of creative information, and take a quotient value between the exploration budget value and the product value as a budget allocation value of each creative information in the candidate media creative information, or take a product value between a target cost in the exploration target and a preset parameter as the budget allocation value of each creative information in the candidate media creative information.

[0040] In a fourth aspect, the present application further provides a media creative information generation apparatus, which comprises:

[0041] The sending module sends a creative information generation request carrying seed creative information and media exploration information to a server; wherein the creative information generation request is used to instruct the server to generate media guide information based on the seed creative information and media attribute information, to perform information processing on the seed creative information and the media exploration information according to the media guide information, to obtain candidate media creative information, to put the candidate media creative information into a media putting system to obtain media evaluation information, and to select media creative information based on the media evaluation information and an exploration target in the media exploration information among the candidate media creative information; the media exploration information comprises an exploration target and an exploration budget value of a target object when requesting creative exploration, and the media attribute information comprises an industry and an audience group corresponding to a media;

[0042] The receiving module is configured to receive the media creative information returned by the server in response to the creative information generation request.

[0043] The display module is configured to display the media creative information.

[0044] In a fifth aspect, the present application further provides a computer device, which comprises a memory and a processor, the memory stores a computer program, and the processor implements the steps of the media creative information generation method when executing the computer program.

[0045] In a sixth aspect, the present application further provides a computer readable storage medium, which stores a computer program, and the computer program implements the steps of the media creative information generation method when executed by a processor.

[0046] In a seventh aspect, the present application further provides a computer program product, which comprises a computer program, and the computer program implements the steps of the media creative information generation method when executed by a processor.

[0047] The media creative information generation method, device, computer device, storage medium and computer program product have the advantages that the media guide information is generated based on the media attribute information and the seed creative information, and then the candidate media creative information is generated according to the seed creative information, the media guide information and the media exploration information, and the guidance of the media guide information is utilized in the process of generating the candidate media creative information, which can help to improve the putting effect of the media creative information. In addition, the candidate media creative information is put into the media putting system to obtain the media evaluation information, and therefore the media creative information is selected by combining the media evaluation information and the exploration target in the media exploration information, which can select the media creative information with high putting effect and meeting the exploration target, so that the obtained media creative information meets the business requirement and can effectively improve the interaction rate of the media creative information. Attached Figure Description

[0048] Figure 1 This is an application environment diagram of a method for generating media creative information in one embodiment;

[0049] Figure 2 This is a flowchart illustrating a method for generating media creative information in one embodiment;

[0050] Figure 3 This is a schematic diagram illustrating how an advertiser inputs ad exploration information and seed ad creatives via a terminal in one embodiment;

[0051] Figure 4 This is a schematic diagram of the exploration results received by the advertiser in one embodiment;

[0052] Figure 5 This is a schematic diagram of the exploration results received by the advertiser in another embodiment;

[0053] Figure 6 This is a flowchart illustrating a method for generating media creative information in another embodiment;

[0054] Figure 7 This is a schematic diagram of the framework of a method for generating media creative information in one embodiment;

[0055] Figure 8 This is a flowchart illustrating a method for generating media creative information in another embodiment;

[0056] Figure 9 This is a structural block diagram of a media creative information generation device in one embodiment;

[0057] Figure 10 This is a structural block diagram of a media creative information generation device in another embodiment;

[0058] Figure 11 This is a structural block diagram of a media creative information generation device in another embodiment;

[0059] Figure 12 This is an internal structural diagram of a computer device in one embodiment;

[0060] Figure 13 This is an internal structural diagram of a computer device in one embodiment. Detailed Implementation

[0061] 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.

[0062] It should be noted that in the following description, the terms "first", "second", and "third" are merely used to distinguish similar objects, and do not represent a specific order of the objects. Understandably, "first", "second", and "third" can be interchanged in a specific order or sequence as allowed, so that the embodiments described herein can be implemented in an order other than that illustrated or described herein.

[0063] Before describing the embodiments of the present application, the technologies involved in the present application are introduced as follows:

[0064] Artificial intelligence (AI) is the use of digital computers or digital computer controlled machines to simulate, extend and expand human intelligence, perceive the environment, acquire knowledge and use knowledge to obtain the best results. In other words, artificial intelligence is a comprehensive technology of computer science, which aims to understand the essence of intelligence and produce a new intelligent machine that can react in a similar way to human intelligence. Artificial intelligence is to study the design principles and implementation methods of various intelligent machines, so that machines have the functions of perception, reasoning and decision-making.

[0065] Artificial intelligence technology is a comprehensive discipline, involving a wide range of fields, both hardware and software technologies. Artificial intelligence basic technologies generally include sensors, special artificial intelligence chips, cloud computing, distributed storage, big data processing technology, pre-training model technology, operation / interaction system, mechatronics, etc. Among them, the pre-training model is also called large model, basic model, which can be widely applied to downstream tasks in various directions of artificial intelligence after fine-tuning. Artificial intelligence software technology mainly includes computer vision technology, speech processing technology, natural language processing technology and machine learning / deep learning, etc.

[0066] Creative: refers to the production of different visual and audio effects through novel methods to maximize the attraction of consumers. In the present application, the creative can be an advertisement image and text placed by the advertiser in the advertisement placement system, or an advertisement image or text generated by an AIGC (Artificial Intelligence Generated Content) model, so the creative can also refer to a creative advertisement.

[0067] Optimization target: the main optimization target of the advertisement in the advertisement placement system, such as order, payment and activation, etc.

[0068] Optimization target cost: one of the indicators for measuring the effectiveness of the launch, together with the optimization target conversion quantity, to measure the effectiveness of the ad creative. Among them, the calculation method of the optimization target cost is: the consumption value of the launch / total conversion quantity. For example, an ad creative with the goal of placing an order, after the launch ends, the consumption value is 100 yuan, and 5 orders are placed, then the average cost of placing an order is 20 yuan.

[0069] Optimization target conversion quantity: refers to the number of optimization targets achieved after the launch; this optimization target conversion quantity is one of the indicators for measuring the effectiveness of the launch, together with the optimization target cost, to measure the effectiveness of the creative.

[0070] Launch effect: used to describe the revenue that an ad creative can bring to the advertiser, mainly composed of two parts: running capacity and cost, measured by conversion quantity and cost. For example, when the cost of ad creative A is lower than that of creative B, it cannot be simply concluded that ad creative A is better than ad creative B. Only when the running capacity, targeting, and launch environment are consistent, and the cost of ad creative A is lower than that of ad creative B, it indicates that ad creative A is better than ad creative B.

[0071] Exploration target: the launch effect expressed by the advertiser to the creative producer. In this application, the creative producer sets the launch according to the targeting and optimization target given by the advertiser, and can bring the conversion quantity and conversion cost to the advertiser after one day of launch. For example, the advertiser promotes yoga classes, and a completed search target is "targeting: A region-male, optimization target: order, cost: 100 yuan / single, conversion quantity: 50 orders".

[0072] Exploration budget: the resource value (such as cost) that the advertiser is willing to spend to produce an ad creative that meets the exploration target, so the exploration budget can refer to the exploration budget value.

[0073] AIGC model: a network model that combines artificial intelligence and graphics computing technology, aiming to enhance the effectiveness and experience of artificial intelligence applications by utilizing the power of graphics computing, with the ability to generate text and images.

[0074] Prompt (media guidance information): text or instructions input to the AIGC model to guide the AIGC model to generate corresponding replies, texts or images, which can effectively improve the output quality and accuracy of the AIGC model, and in this application, media creative information (such as creative ads in image or copy form) can be generated. In addition, the prompt generator can be a model that generates prompts.

[0075] Bandit algorithm: a type of reinforcement learning algorithm used to solve multi-armed bandit problems, which constantly tries different arms and updates the selection strategy according to the reward results to maximize cumulative rewards.

[0076] nobid: is the smooth spending of the budget by the advertising delivery system within a given time, as much as possible to bring more conversions for the advertiser, that is, the delivery mode of maximizing conversion.

[0077] The media creative information generation method provided by the embodiments of the present application can be applied to the application environment as shown in the figure. Figure 1 The terminal 102 communicates with the server 104 through the network. The data storage system can store the data required by the server 104 for processing. The data storage system can be integrated on the server 104, or placed on the cloud or other network servers.

[0078] The terminal 102 sends a creative information generation request to the server, and the server 104 obtains seed creative information, media exploration information and media attribute information from the creative information generation request; generates media guide information based on the media attribute information and the seed creative information; performs information processing on the seed creative information and the media exploration information according to the media guide information to obtain candidate media creative information; puts the candidate media creative information into a media delivery system to obtain media evaluation information; selects media creative information based on the media evaluation information and the exploration target in the media exploration information among the candidate media creative information, and then returns the media creative information to the terminal 102, and the terminal 102 displays the media creative information.

[0079] The terminal 102 can be a smart phone, a tablet computer, a notebook computer, a desktop computer, an Internet of Things device, and a portable wearable device. The Internet of Things device can be a smart speaker, a smart television, a smart air conditioner, and a smart vehicle device. The portable wearable device can be a smart watch, a smart bracelet, and a head-mounted device.

[0080] The server 104 can be a standalone physical server, or a service node in a blockchain system. The service nodes in the blockchain system form a peer-to-peer (Peer To Peer) network between each other. The peer-to-peer protocol is an application layer protocol running on the transmission control protocol (TCP, Transmission Control Protocol) protocol. In addition, the server 104 can also be a server cluster composed of multiple physical servers, which can be a cloud server providing cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communication, middleware services, domain name services, security services, content delivery networks (Content Delivery Network, CDN), and big data and artificial intelligence platforms. Basic cloud computing services such as platform.

[0081] The terminal 102 and the server 104 can be connected through a communication connection mode such as Bluetooth, a universal serial bus (USB), or a network, and the present application does not limit this.

[0082] In one embodiment, as shown in Figure 2 , a media creative information generation method is provided, which can be executed by a terminal or a server in Figure 1 , or can be executed by a terminal and a server in Figure 1 , for example, the method executed by the server in Figure 1 , including the following steps:

[0083] S202, seed creative information, media exploration information, and media attribute information are obtained.

[0084] The seed creative information can refer to creative information that generates new media creative information as a seed (which can be referred to as a seed creative or a creative seed). In some specific application scenarios, the seed creative information can be a seed advertisement creative, that is, an advertisement that generates a new creative advertisement as a seed. The seed advertisement creative can also be referred to as a seed creative advertisement or a seed advertisement creative.

[0085] The media exploration information can refer to the constraint of the generation of media creative information in the media exploration (i.e., media generation) process, so that the generated media creative information meets the actual business needs. The media exploration information can include an exploration target and an exploration budget value of a target object (such as an advertiser) when requesting creative exploration; the exploration target can be the delivery effect expressed by the advertiser to the creative producer, including media targeting, optimization target, target cost (i.e., exploration cost), and conversion volume (such as daily conversion volume), etc., and the exploration budget value is the resource value that the advertiser is willing to spend to produce media creative information that meets the exploration target.

[0086] The media attribute information can be the attribute information of the media, which can include the industry and the audience group corresponding to the media, and can further include the industry type. The media attribute information can be extracted from the media creative information that meets the preset cost condition (such as a decrease of more than 5% in the optimization cost).

[0087] In one embodiment, the server receives a creative information generation request initiated by the terminal; reads the seed creative information and the media exploration information from the creative information generation request, and generates the media attribute information based on the creative information generation request. For example, when an advertiser needs to explore a creative advertisement that meets the exploration target, the advertiser inputs the advertisement exploration information (including the exploration target and the exploration budget) on the advertiser terminal, and uploads the seed advertisement creative, such as Figure 3The server can obtain the advertisement exploration information and the seed creative advertisement from the creative information generation request, and obtain the advertisement attribute information from the advertisement that meets the preset cost condition (such as a cost optimization of more than 5%) according to the creative information generation request.

[0088] In S204, media guidance information is generated based on the media attribute information and the seed creative information.

[0089] The media guidance information is text or instructions (such as media guidance words) input to the AIGC model, which is used to guide the AIGC model to generate media creative information (such as creative advertisements in the form of images, texts, or a combination of images and texts).

[0090] In one embodiment, the server inputs the media attribute information and the seed creative information to a guidance generator, which processes the media attribute information and the seed creative information to obtain the media guidance information.

[0091] The guidance generator can be a prompt generator that generates media guidance information as instructions to guide the AIGC model to generate media creative information (such as creative advertisements in the form of images, texts, or a combination of images and texts).

[0092] For example, after the server inputs the media attribute information and the seed creative information to the prompt generator, the prompt generator pre-processes the media attribute information and the seed creative information and extracts key information; then generates media guidance information according to the key information. The pre-processing can include word segmentation, part-of-speech tagging, and named entity recognition, etc.

[0093] In one embodiment, the server can process the media attribute information and the seed creative information through the guidance generator to obtain a first creative style, a first creative quality, and a first creative similarity corresponding to the seed creative information; and concatenate the first creative style, the first creative quality, and the first creative similarity to obtain the media guidance information.

[0094] The first creative style can be a creative style or a media style of the seed creative information. For example, assuming that the seed creative information is a seed advertisement creative image, the first creative style can be an advertisement style or an image style of the seed advertisement creative image, such as an animation style and a realistic style, etc.

[0095] The first creative quality can be a quality of the seed creative information. For example, assuming that the seed creative information is a seed advertisement creative image, the first creative quality can be an image quality (such as a resolution of the image and a realism of a texture); assuming that the seed creative information is a seed advertisement creative script, the first creative quality can be a clarity of semantics and logic of the advertisement creative script.

[0096] The first creative similarity can be a similarity between the generated media creative information and the seed creative information.

[0097] For example, after the server inputs the media attribute information and the seed creative information into the prompt generator, the prompt generator can pre-process the media attribute information and the seed creative information, and extract key information; then generate the first creative style, the first creative quality and the first creative similarity corresponding to the seed creative information according to the key information, and splice the first creative style, the first creative quality and the first creative similarity to obtain the media guidance information.

[0098] In one embodiment, the server can also generate the media guidance information based on the media exploration information, the media attribute information and the seed creative information, so as to improve the accuracy and stability of the media guidance information, and facilitate the generated candidate media creative information to have a high delivery effect.

[0099] In the process of generating the media guidance information, the server can input the media exploration information, the media attribute information and the seed creative information into the guidance generator, so that the guidance generator performs information processing on the media exploration information, the media attribute information and the seed creative information to obtain the first creative style, the first creative quality and the first creative similarity corresponding to the seed creative information, and splice the first creative style, the first creative quality and the first creative similarity to obtain the media guidance information.

[0100] For example, the server can input the advertisement exploration information, the advertisement attribute information and the seed advertisement creative into the guidance generator, so that the guidance generator performs information processing on the advertisement exploration information, the advertisement attribute information and the seed advertisement creative to obtain the first creative style, the first creative quality and the first creative similarity corresponding to the seed advertisement creative, and splice the first creative style, the first creative quality and the first creative similarity to obtain the advertisement guidance information.

[0101] In addition, in the process of generating the media guidance information, the server can also perform information processing on the media exploration information, the media attribute information and the seed creative information through the guidance generator to obtain the first creative style, the first creative quality, the first creative similarity and the first creative scene type corresponding to the seed creative information, and splice the first creative style, the first creative quality, the first creative similarity and the first creative scene type to obtain the media guidance information.

[0102] The first creative scene type can be an application scene of the seed creative information, such as an indoor scene or an outdoor scene.

[0103] In one embodiment, the guidance generator is a generator obtained by training an initial guidance generator, and the training process includes: the server obtains sample media attribute information, sample seed creative information and label information from a data source; the initial guidance generator is used to process the sample media attribute information and the sample seed creative information to obtain a second creative style, a second creative quality and a second creative similarity corresponding to the sample seed creative information; a loss value is determined according to the second creative style, the second creative quality, the second creative similarity and the label information; and the parameters in the initial guidance generator are adjusted based on the loss value to obtain the guidance generator, so that the guidance generator can learn how to understand the semantics and context of the problem (such as the input media creative information and media attribute information) and can extract useful information.

[0104] The sample media attribute information can be extracted from media creative information whose optimization cost meets a preset cost condition (such as a decrease of more than 5% in the optimization cost).

[0105] The sample seed creative information refers to creative information that can be used to generate new media creative information as a seed in the training process. In some specific application scenarios, the sample seed creative information can be a seed advertisement creative used for training. It should be noted that the sample media attribute information and the sample seed creative information can be regarded as input features of the initial guidance generator in the training process.

[0106] The label information can include the similarity, the creative quality and the creative style of the sample seed creative information. In addition, it can also include an application scene, such as an indoor scene or an outdoor scene.

[0107] In S206, the seed creative information and the media exploration information are processed according to the media guidance information to obtain candidate media creative information.

[0108] The candidate media creative information can be a plurality of media creative information generated in the exploration process.

[0109] In one embodiment, the server inputs the seed creative information, the media guidance information and the media exploration information into the AIGC model. The media guidance information can control the generation effect of the AIGC model, that is, the AIGC model can use the media guidance information as the generation effect of the candidate creative media information to extract and process the seed creative information and the media exploration information, so as to obtain the candidate creative media information meeting the generation effect.

[0110] For example, assuming that the seed creative information is a creative advertisement in the form of an image, and the corresponding media guide information is advertisement guide information, the AIGC model can perform feature extraction on the creative advertisement and the media exploration information to obtain creative advertisement image features and media exploration text features, take the advertisement guide information as a generation effect of the candidate creative advertisement, and generate a candidate creative advertisement matching the generation effect according to the creative advertisement image features and the media exploration text features.

[0111] In view of the fact that in the conventional scheme, the AIGC model is usually directly used to generate candidate media creative information according to seed creative information (such as a seed creative image or a seed creative script), the actual delivery effect may be poor. Therefore, the media guide information generated by the guide generator can guide the AIGC model to generate candidate media creative information, which can effectively control the generation effect of the candidate media creative information and is conducive to improving the actual delivery effect of the media creative information.

[0112] S208, delivering the candidate media creative information to a media delivery system to obtain media evaluation information.

[0113] The media delivery system can be a system or platform for delivering media creative information and displaying it to a large number of consumers. In some specific application scenarios, the media delivery system can be a professional advertisement delivery platform, a shopping platform, a takeout platform, a short video platform, a live broadcast platform, a game platform, or other business platforms.

[0114] The media evaluation information can be evaluation information for the candidate media creative information, such as at least one of a click rate, a conversion volume (or conversion rate), a conversion cost, or an optimization target of the candidate media creative information.

[0115] In one embodiment, after the server delivers the candidate media creative information to the media delivery system, the consumer can interact with the candidate media creative information through the terminal, such as clicking on a certain media creative information in the candidate media creative information, and then entering a product page to place an order, or entering an activation page to perform an activation operation (such as activating a bank card or activating a game account), or entering a download page to download, or entering a payment page to perform a payment operation, etc. After the candidate media creative information is delivered to the media delivery system, the media delivery system can count the media evaluation information of the candidate media creative information. The server can request the media evaluation information from the media delivery system at regular intervals, or the server can receive the media evaluation information pushed by the media delivery system at regular intervals.

[0116] In an embodiment, the candidate media creative information is at least one of a candidate media creative image or a candidate media creative script; the method further comprises: performing quality verification on the at least one creative information by the server to obtain a verification result. S206 can comprise: in a case where the verification result indicates that the at least one creative information meets the delivery condition, delivering the at least one candidate creative information to the media delivery system.

[0117] The at least one candidate creative information can be a candidate media creative image or a candidate media creative script, or a candidate media creative image and a candidate media creative script. The number of the at least one candidate creative information can be multiple. For example, when the at least one candidate creative information is a candidate media creative image, the number of the candidate media creative images can be multiple. For another example, when the at least one candidate creative information is a candidate media creative script, the number of the candidate media creative scripts can be multiple. For another example, when the at least one candidate creative information is a candidate media creative script and a candidate media creative image, the number of the candidate media creative script and the number of the candidate media creative image can be multiple respectively.

[0118] For delivery of the candidate media creative information, specifically: in a case where the verification result indicates that the at least one creative information meets the delivery condition, sorting media creative information in the at least one candidate creative information to obtain a candidate creative information sequence, selecting k creative information in the candidate creative information sequence, and then delivering the k creative information to the media delivery system.

[0119] In an embodiment, for the process of quality verification, specifically can comprise: sequentially performing image quality verification and compliance verification on the candidate media creative image by the server to obtain an image verification result; sequentially performing semantic detection and privacy information detection on the candidate media creative script to obtain a script detection result.

[0120] The image quality verification can be to verify texture authenticity, image resolution (or definition), etc. of the candidate media creative image. The compliance verification can be to verify whether the candidate media creative image involves abnormal behavior or abnormal performance that is not in compliance.

[0121] The semantic detection can be to detect whether the semantic expression of the candidate media creative script is clear, smooth and coherent. The privacy information detection can be to detect whether there is private information of an advertiser or a consumer in the candidate media creative script.

[0122] S210, in the candidate media creative information, selecting media creative information based on media evaluation information and an exploration target in media exploration information.

[0123] The media creative information can be creative information in the candidate media creative information that satisfies a selection condition, such as media evaluation information of selected media creative information satisfying a media evaluation condition, and an exploration target of the selected media creative information satisfying an exploration target condition. In addition, the media creative information can also be creative information that is selected according to ranking of the candidate media creative information according to the media evaluation information and the exploration target, and satisfies a ranking condition (such as topk).

[0124] In one embodiment, after S210, the server can send the selected media creative information to the terminal for display, so that the advertiser can view the result of the media creative information exploration; in addition, the server can also send the remaining budget value and the media exploration effect to the terminal. For example, the server selects one or more advertisement creatives from the candidate advertisement creatives according to the advertisement evaluation information and the exploration target in the advertisement exploration information, and then sends the selected advertisement creatives to the terminal for display; in addition, the server can also send the remaining budget value and the advertisement exploration effect to the terminal for display, as shown in Figure 4

[0125] In one embodiment, for selection of the media creative information, the server can sort the candidate media creative information according to the media evaluation information and the exploration target in the media exploration information to obtain a creative information sequence; in the creative information sequence, the media creative information that satisfies a sequence condition is selected.

[0126] For example, the server can sort the candidate media creative information according to the target cost (or conversion amount) of the candidate media creative information in the delivery process and the target cost (or conversion amount) set by the target object when requesting creative exploration to obtain a creative information sequence; in the creative information sequence, the media creative information that satisfies the first ranking or the top ranking is selected.

[0127] In one embodiment, the server can determine the remaining budget value according to the exploration budget value and the exploration consumption value in the media exploration information; determine the media exploration effect according to the media evaluation information and the media exploration information; and send the media creative information, the remaining budget value and the media exploration effect to the terminal for display.

[0128] In one embodiment, the server sends the media creative information, the remaining budget value and the media exploration effect to the terminal for display when the media exploration effect meets the exploration target, or the media exploration effect does not meet the exploration target and the remaining budget value is less than a preset threshold. For example, when the advertisement exploration effect meets the exploration target, the server can also send the remaining budget value and the advertisement exploration effect to the terminal for display, as shown in Figure 4 ​As shown; for example, when the media exploration effect does not meet the exploration target, and the remaining budget value is less than the preset threshold, the server can also send the remaining budget value and the advertisement exploration effect to the terminal for display, as shown in Figure 5

[0129] In addition, when the media exploration effect does not meet the exploration target, and the remaining budget value is greater than or equal to the preset threshold, at least one creative information in the creative information sequence is selected as a new seed creative information, and a creative information generation method of a next iteration round is executed based on the new seed creative information, until the media exploration effect meets the exploration target, or the media exploration effect does not meet the exploration target, and the remaining budget value is less than the preset threshold, the iteration process is stopped, and the media creative information is obtained.

[0130] The step of executing the creative information generation method of the next iteration round based on the new seed creative information can include: performing information processing on the seed creative information and the media exploration information according to the media guide information to obtain candidate media creative information; putting the candidate media creative information into the media putting system to obtain media evaluation information; and selecting the media creative information in the candidate media creative information based on the media evaluation information and the exploration target in the media exploration information. It should be noted that the creative information generation method of each round is the same, except that the seed creative information is different in different iteration rounds.

[0131] It should be noted that S202-S210 is a media creative information exploration process of one iteration round, if the media exploration effect does not meet the exploration target, and the remaining budget value is greater than or equal to the preset threshold, a suitable media creative information is selected from the candidate media creative information as a seed creative information for the next iteration round of media creative information exploration.

[0132] In the above embodiment, the media guide information is generated based on the media attribute information and the seed creative information, and then the candidate media creative information is generated according to the seed creative information, the media guide information and the media exploration information. In the process of generating the candidate media creative information, the guidance of the media guide information is used, which can be beneficial to improve the putting effect of the media creative information. In addition, the candidate media creative information is put into the media putting system to obtain the media evaluation information, so that when the media creative information is selected, the media evaluation information and the exploration target in the media exploration information are combined for selection, and the media creative information with high putting effect and meeting the exploration target can be selected, so that the obtained media creative information meets the business demand, and the interaction rate of the media creative information can be effectively improved.

[0133] ​In an embodiment, the candidate media creative information includes k creative information; the method further includes: the server determining the number of creative information to be explored in each iteration round and the number of iterations in the preset total period; determining the product value between the number of iterations and the number of creative information; taking the quotient value between the exploration budget value and the product value as the budget allocation value of each creative information in the candidate media creative information; or taking the product value between the target cost in the exploration target and the preset parameter as the budget allocation value of each creative information in the candidate media creative information.

[0134] The preset total period can be the total time length of all iteration rounds, such as the number of iterations in one day. The target cost can refer to the exploration cost set before the exploration of creative information.

[0135] For budget allocation: the budget for this iteration is divided from the total exploration budget value, which needs to consider the confidence of the effect of each iteration, the total exploration budget value and the exploration cost, etc., so the budget allocated to each media creative in this iteration can refer to the following budget calculation method:

[0136]

[0137] Wherein, B is the total exploration budget value expressed by the advertiser for this exploration; C is the exploration cost given by the advertiser; k is the average number of creative to be explored in each iteration; m is the number of iterations in the entire exploration process; r is a preset parameter (i.e. hyperparameter), which represents the minimum number of conversions for confidence of delivery effect, which can be 5 for example.

[0138] Suppose the media creative information is an advertisement creative, then the budget calculation formula can mean: the total budget of the advertiser is evenly allocated to the m*k creatives in m iterations, if the allocated exploration budget (i.e. budget allocation value) is too small to complete a confident exploration, then increase the exploration budget to r*C.

[0139] In another embodiment, the candidate media creative information includes k creative information, and the exploration target in the media exploration information includes a target cost and a conversion amount; the method further includes:

[0140] The server determines the number of creative information to be explored in each iteration round; determines the ratio between the traffic in the target exploration period and the traffic in the preset total period; determines the number of exploration periods according to the preset total period and the target exploration period; determines the number of iterations according to the preset total period; determines the budget allocation value of each creative information in the k creative information according to the number of iterations, the number of exploration periods, the ratio, the number of creative information, the target cost and the conversion amount.

[0141] The budget allocation value is used to determine the exploration consumption value.

[0142] For the calculation of the budget allocation value, specifically, the ratio between the number of iterations and the number of creative information can be calculated first, the total ratio in the preset exploration period is determined according to the number of exploration periods and the ratio, the product between the total ratio, the target cost and the conversion amount is calculated, and the budget allocation value of each creative information is determined according to the product and the ratio.

[0143] For example, the budget allocation value of each creative information can be calculated by using the following calculation formula:

[0144]

[0145] Wherein, i is the target exploration period (such as the current hour), is the budget allocation value, C is the target cost, T is the conversion amount, f(x) is the ratio between the traffic in the target exploration period and the traffic in the preset total period, is the number of exploration periods, k is the number of creative information (referred to as creative number) to be explored in each iteration round, and m is the number of iterations in the preset total period.

[0146] In the above embodiment, each creative information to be explored for each iteration round is allocated a corresponding budget allocation value, so that the creative information of each iteration round can be uniformly allocated to the corresponding budget value, so that the creative information of each iteration round can be fully put into the accurate and effective media evaluation information, which is beneficial to improve the accuracy and effectiveness of the media evaluation information.

[0147] In one embodiment, as shown in Figure 6 , a media creative information generation method is provided, which is executed by a terminal in Figure 1 , including the following steps:

[0148] S602, a creative information generation request carrying seed creative information and media exploration information is sent to the server; wherein the creative information generation request is used to instruct the server to generate media guide information based on the seed creative information and the media attribute information, to process the seed creative information and the media exploration information according to the media guide information, to obtain candidate media creative information, to put the candidate media creative information into a media putting system to obtain media evaluation information, and to select media creative information in the candidate media creative information based on the exploration target in the media evaluation information and the media exploration information.

[0149] Wherein, the seed creative information can refer to the creative information used as a seed to generate new media creative information. In some specific application scenarios, the seed creative information can be a seed advertisement creative, that is, an advertisement used as a seed to generate a new creative advertisement. The seed advertisement creative can also be referred to as a seed creative advertisement.

[0150] The media exploration information can refer to a constraint on generation of the media creative information in a media exploration (i.e., media generation) process, so that the generated media creative information meets actual business requirements. The media exploration information can include an exploration target and an exploration budget value of a target object (e.g., an advertiser) when requesting creative exploration; the exploration target can be a delivery effect expressed by the advertiser to a creative production party, including media targeting, optimization targets, target costs (i.e., exploration costs), and conversion amounts (e.g., daily conversion amounts), and the exploration budget value is a resource value that the advertiser is willing to spend to produce media creative information that meets the exploration target.

[0151] The media attribute information can be attribute information possessed by the media, such as an industry and an audience group of the media. The media attribute information can be extracted from media creative information in which an optimization cost meets a preset cost condition (e.g., a decrease of more than 5% in the optimization cost).

[0152] The candidate media creative information can be a plurality of media creative information generated in the exploration process.

[0153] In one embodiment, the terminal acquires input seed creative information and media exploration information in response to an input operation, and generates a creative information generation request carrying the seed creative information and the media exploration information in response to a confirmation operation, and sends the creative information generation request to the server.

[0154] In one embodiment, the terminal acquires input seed creative information and media exploration information in response to an input operation, and generates a creative information generation request carrying the seed creative information and the media exploration information in response to a confirmation operation, and sends the creative information generation request to the server.

[0155] For the media creative information generation process of the server, refer to the embodiments of Figure 2 .

[0156] S604, receiving media creative information returned by the server in response to the creative information generation request.

[0157] S606, displaying the media creative information.

[0158] In the above embodiments, the media guide information is generated based on the media attribute information and the seed creative information, and then the candidate media creative information is generated according to the seed creative information, the media guide information and the media exploration information. In the process of generating the candidate media creative information, the guidance of the media guide information is used, which can be beneficial to improve the delivery effect of the media creative information. In addition, the candidate media creative information is delivered to the media delivery system to obtain media evaluation information. Therefore, when selecting the media creative information, the exploration target in the media evaluation information and the media exploration information is fused for selection, and the media creative information with high delivery effect and meeting the exploration target can be selected. Therefore, the obtained media creative information meets the business demand and can effectively improve the interaction rate of the media creative information.

[0159] As an example, in order to make the technical solutions of the present application clearer, the present application is described in combination with an advertising scenario, Figure 7 and Figure 8 as follows:

[0160] Firstly, the advertising creative is very important to improve the advertising effect. The advertiser will invest considerable resources to optimize the advertising creative. Generally, designers create novel images, copy or image-text according to historical experience to attract target groups. However, there are problems such as long creative production cycle and unpredictable effect. In addition, the effect of an advertising creative will obviously decline after being delivered for a period of time. The present application mainly meets the demand of improving the advertising effect by optimizing the creative.

[0161] 1) Demand expression of the advertiser.

[0162] The advertiser uploads a seed creative through a terminal, and sets an exploration target and an exploration budget. For example, the advertiser promotes a course, the target is male in A area, the optimization target is order, the previous delivery cost of the advertiser is 80 yuan per time, and the daily conversion quantity is 100. The advertiser expects to generate at least one better advertising creative, the optimization cost C=50 yuan / time, and the daily conversion quantity T=100. In order to explore such an advertising creative, the advertiser is willing to pay an exploration budget B=1000 yuan.

[0163] 2) Candidate creative generation.

[0164] The server generates N candidate creatives (such as candidate creative images, candidate creative copy or candidate creative image-text) based on the prompt generator and the AIGC model according to the seed creative uploaded by the advertiser.

[0165] The candidate creative can refer to a candidate advertising creative, and the number can be multiple.

[0166] Considering that in the traditional scheme, the AIGC model is usually directly used to generate candidate creatives from seed ideas, the actual delivery effect may be poor. Therefore, the prompt generator is used to generate an advertisement guide word to guide the AIGC model to generate a candidate creative, which can effectively control the generation effect of the candidate creative and improve the delivery effect of the advertisement creative in practice. The prompt generator is a model trained according to historical data, which comprehensively considers the differences in effect of different industries, targeting, optimization targets and image types.

[0167] The prompt generator is composed of multiple multi-classification models, which outputs the similarity, image style and image quality of the original advertisement creative image or script according to the industry, targeting factors and advertisement creative image or script of the advertisement, and then splices them into an advertisement guide word (prompt) according to certain rules.

[0168] For the training of the prompt generator, the sample source, input features and labels are as follows:

[0169] Sample source: good effect prompts in historical generated advertisement creatives (such as optimization cost reduction of more than 5%);

[0170] Input features: industry attributes, targeting attributes and optimization targets of seed ideas, as shown in Table 1;

[0171] Label: label of generated prompt, such as similarity, image style and image quality, as shown in Table 1.

[0172] Table 1

[0173]

[0174] 3) Quality audit.

[0175] In order to ensure the quality of the candidate creative, filter those obviously poor effect candidate creatives to avoid affecting the experience of the advertiser or wasting the delivery resources, such as: candidate creative images that are not in line with regulations, candidate creative images that are not clear in details, candidate creative images that are not real enough in texture; in addition, candidate creative scripts that do not conform to common sense, such as object collocation, script sentences that are not smooth in meaning consistency; candidate creative scripts involving personal data or privacy leakage; candidate creative images that are obviously poor in promotion quality, such as not meeting the aesthetic standards of the public, etc.

[0176] The quality audit process can be divided into model discrimination and manual screening. The model part improves accuracy and recall through continuous iteration, and is expected to eventually liberate manpower. After quality audit, candidate creatives that meet the audit requirements are obtained.

[0177] 4) Ad creative delivery to the advertisement delivery system.

[0178] After the candidate creatives are audited by the quality audit, the next step is to enter the delivery stage. If there are many images in the delivery stage, considering the limited budget and the confidence of the delivery effect, K ad creatives can be selected from the candidate creatives after the candidate creative images and candidate creative texts are arranged and combined for delivery.

[0179] The delivery is in the nobid mode, that is, the given budget is spent at a constant speed within a given delta hours. It should be noted that the K ad creatives need to be completely consistent in the settings of the ad delivery system, including targeting, budget, and delivery time, etc., and only the ad creatives are different.

[0180] Budget allocation: the total budget is divided into the budget for this iteration, which needs to consider the confidence of the effect of each iteration, the total exploration budget, the exploration target, etc. The budget allocated to each creative in this iteration can be calculated according to the budget calculation formula, which is as follows:

[0181]

[0182] Where B is the total exploration budget expressed by the advertiser; C is the exploration cost given by the advertiser; k is the average number of creatives to be explored in each iteration; m is the number of iterations in the entire exploration process; r is the minimum number of conversions for the delivery effect confidence, which is a hyperparameter, such as 5.

[0183] The budget calculation formula can mean that the total budget of the advertiser is evenly distributed to the m*k creatives in m iterations. If the allocated exploration budget (i.e., the budget allocation value) is very small, so as to be unable to complete a confident exploration, then increase the exploration budget to r*C.

[0184] 5) Effect evaluation.

[0185] After one iteration is completed, the delivery effect of each ad creative is collected, the conversion cost of each ad creative is sorted according to the optimization target set by the advertiser, the topk with the best effect is selected, and the delivery effect in the mode set in this paper refers to the conversion cost. Considering that the budget of each ad creative is consistent and the delivery mode is nobid, the conversion amount can also directly reflect the conversion cost.

[0186] 6) Determine whether the cost meets the standard.

[0187] If the cost meets the standard, it means that the ad creative has met the needs of the advertiser, and the loop exploration mechanism is exited. If there is remaining budget, it means that the needs of the advertiser have been met within the budget range of the advertiser. The remaining budget, the ad creative with the cost meeting the standard, and the actual conversion cost are returned to the terminal.

[0188] 7) Determine if the budget has been exhausted.

[0189] If the conversion cost does not meet expectations, but the advertising budget has reached the advertiser's exploration budget limit, no further exploration will be conducted. Instead, the currently explored ad creatives and conversion costs will be returned for the advertiser's reference to decide whether to conduct further exploration.

[0190] 8) Screening and iterating seed ideas.

[0191] If there is still a budget and the campaign performance does not meet the advertiser's standards, based on the Bandit algorithm, the actual online campaign performance will be used as a reward. One or several of the best-performing creatives will be selected from k creatives as seed creatives for the next iteration of creative generation, deployment, and evaluation.

[0192] like Figure 3 As shown, the advertiser expects to explore a new ad creative based on the uploaded seed creative, and hopes that the actual ad cost of this ad creative, under the setting of targeting and optimization goals, will not exceed 20 yuan per order and will have a conversion rate of not less than 50. The advertiser is willing to spend 1,000 yuan for this exploration.

[0193] like Figure 4 As shown, based on the AIGC model and advertiser feedback, the server ultimately developed a creative for the advertiser with a cost of 20 yuan per order and a daily conversion rate of 51 orders, without spending the entire budget. This demonstrates that the creative met the advertiser's expectations, and the advertiser can choose to download and view it.

[0194] like Figure 5 As shown, based on the AIGC model and advertiser feedback, the server ultimately developed a creative for the advertiser at 22 yuan per order with a daily conversion rate of 51 orders after the budget was exhausted. It is evident that this did not meet the advertiser's expectations, but the advertiser can choose to download and view it.

[0195] If advertisers believe that the current exploration results meet their expectations, they can choose "Start Launch with One Click". The server will then start the launch directly according to the current targeting, optimization goals and budget (daily conversions * target cost).

[0196] This application proposes a systematic creative generation scheme that clarifies the contract between the platform and advertisers, generating ad creatives under the constraints of advertisers' exploration results. Furthermore, it not only uses manual screening and model recognition but also incorporates real online feedback to evaluate the effectiveness of the creatives, making online campaigns more reliable. Finally, traditional solutions often struggle to guarantee the effectiveness of ad creatives generated directly based on AIGC models, wasting exploration costs. In contrast, this application's prompt generator generates prompts based on historical data, better aligning with advertisers' optimization goals.

[0197] It should be understood that although each step in the flowchart involved in the embodiments described above is shown in sequence according to the arrow, the steps are not necessarily executed in the order indicated by the arrow. Unless otherwise specified herein, the execution of the steps is not strictly limited in sequence, and the steps can be executed in other orders. Moreover, at least part of the steps in the flowchart involved in the embodiments described above can include multiple steps or multiple stages, which are not necessarily executed at the same time, but can be executed at different times, and the execution order of the steps or stages is not necessarily sequential, but can be alternately executed with other steps or steps or stages in other steps.

[0198] Based on the same inventive concept, the embodiments of the present application also provide a media creative information generation device for implementing the media creative information generation method described above. The problem-solving implementation scheme provided by the device is similar to the implementation scheme described in the above method, so the specific limitations in one or more media creative information generation device embodiments provided below can refer to the limitations of the media creative information generation method described above, and will not be repeated here.

[0199] In one embodiment, as shown in Figure 9 A media creative information generation device is provided, comprising: an acquisition module 902, a first generation module 904, a second generation module 906, a delivery module 908, and a selection module 910, wherein:

[0200] The acquisition module 902 is configured to acquire seed creative information, media exploration information, and media attribute information.

[0201] The first generation module 904 is configured to generate media guide information based on the media attribute information and the seed creative information.

[0202] The second generation module 906 is configured to perform information processing on the seed creative information and the media exploration information according to the media guide information to obtain candidate media creative information.

[0203] The delivery module 908 is configured to deliver the candidate media creative information to a media delivery system to obtain media evaluation information.

[0204] The selection module 910 is configured to select media creative information based on the media evaluation information and an exploration target in the media exploration information from the candidate media creative information.

[0205] In one of the embodiments, the first generation module 904 is further configured to: perform information processing on the media attribute information and the seed creative information by the guide generator to obtain a first creative style, a first creative quality and a first creative similarity corresponding to the seed creative information; and splice the first creative style, the first creative quality and the first creative similarity to obtain the media guide information.

[0206] In one of the embodiments, as shown in Figure 10 the apparatus further comprises:

[0207] The training module 912 is configured to: obtain sample media attribute information, sample seed creative information and label information from a data source; perform information processing on the sample media attribute information and the sample seed creative information by an initial guide generator to obtain a second creative style, a second creative quality and a second creative similarity corresponding to the sample seed creative information; determine a loss value according to the second creative style, the second creative quality, the second creative similarity and the label information; and adjust parameters in the initial guide generator based on the loss value to obtain the guide generator.

[0208] In one of the embodiments, the candidate media creative information is at least one candidate creative information in a candidate media creative image or a candidate media creative script; as shown in Figure 10 the apparatus further comprises:

[0209] The verification module 914 is configured to perform quality verification on the at least one creative information to obtain a verification result.

[0210] The delivery module 908 is further configured to, in a case where the verification result indicates that the at least one creative information meets the delivery condition, deliver the at least one candidate creative information to the media delivery system.

[0211] In one of the embodiments, the verification module 914 is further configured to sequentially perform image quality checking and compliance checking on the candidate media creative image to obtain an image checking result; and sequentially perform semantic detection and privacy information detection on the candidate media creative script to obtain a script detection result.

[0212] In one of the embodiments, as shown in Figure 10 the apparatus further comprises:

[0213] The acquisition module 902 is further configured to: receive a creative information generation request initiated by a terminal; read the seed creative information and the media exploration information from the creative information generation request; and acquire the media attribute information based on the creative information generation request.

[0214] The sending module 916 is configured to send the media creative information to the terminal for display.

[0215] In one of the embodiments, the selecting module 910 is further configured to sort the candidate media creative information based on the media evaluation information and the exploration target in the media exploration information to obtain a creative information sequence; and select media creative information satisfying an order condition from the creative information sequence.

[0216] In one of the embodiments, as shown in Figure 10 the apparatus further comprises:

[0217] The determining module 918 is configured to determine a residual budget value according to an exploration budget value and an exploration consumption value in the media exploration information; and determine a media exploration effect according to the media evaluation information and the media exploration information.

[0218] The sending module 916 is further configured to send the media creative information, the residual budget value and the media exploration effect to the terminal for display.

[0219] In one of the embodiments, as shown in Figure 10 the apparatus further comprises:

[0220] The sending module 916 is further configured to send the media creative information, the residual budget value and the media exploration effect to the terminal for display when the media exploration effect meets the exploration target, or the media exploration effect does not meet the exploration target and the residual budget value is less than a preset threshold value.

[0221] The processing module 920 is configured to select at least one creative information in the creative information sequence as a new seed creative information when the media exploration effect does not meet the exploration target and the residual budget value is greater than or equal to the preset threshold value, execute the creative information generation method of the next iteration round based on the new seed creative information, and stop the iteration process until the media exploration effect meets the exploration target, or the media exploration effect does not meet the exploration target and the residual budget value is less than the preset threshold value, to obtain the media creative information.

[0222] In one of the embodiments, the candidate media creative information includes k creative information, and the exploration target in the media exploration information includes a target cost and a conversion amount.

[0223] The determining module is configured to determine a number of creative information to be explored in each iteration round; determine a ratio between a traffic in a target exploration period and a traffic in a preset total period; determine a number of exploration periods according to the preset total period and the target exploration period; determine a number of iterations in the preset total period; and determine a budget allocation value of each creative information in the k creative information according to the number of iterations, the number of exploration periods, the ratio, the number of creative information, the target cost and the conversion amount; wherein each budget allocation value is used to determine the exploration consumption value.

[0224] In the above embodiment, the creative information to be explored for each iteration round is allocated with a corresponding budget allocation value, so that the creative information of each iteration round can be evenly allocated to the corresponding budget value, so that the creative information of each iteration round can be fully put into accurate and effective media evaluation information, which is beneficial to improve the accuracy and effectiveness of the media evaluation information.

[0225] In one embodiment, as shown in Figure 11 A media creative information generation apparatus is provided, comprising: a sending module 1102, a receiving module 1104 and a display module 1106, wherein:

[0226] The sending module 1102 is configured to send a creative information generation request carrying seed creative information and media exploration information to a server; wherein the creative information generation request is used to instruct the server to generate media guide information based on the seed creative information and media attribute information, to perform information processing on the seed creative information and the media exploration information according to the media guide information, to obtain candidate media creative information, to put the candidate media creative information into a media putting system to obtain media evaluation information, and to select media creative information in the candidate media creative information based on the exploration target in the media evaluation information and the media exploration information; the media exploration information includes an exploration target and an exploration budget value of a target object when requesting creative exploration, and the media attribute information includes an industry and an audience group corresponding to the media;

[0227] The receiving module 1104 is configured to receive media creative information returned by the server in response to the creative information generation request;

[0228] The display module 1106 is configured to display the media creative information.

[0229] In the above embodiment, the creative information generation request is sent to the server, so that the server generates media guide information based on the media attribute information and the seed creative information, and then generates candidate media creative information according to the seed creative information, the media guide information and the media exploration information. In the process of generating the candidate media creative information, the guidance of the media guide information is used, which can be beneficial to improve the putting effect of the media creative information. In addition, the candidate media creative information is put into the media putting system to obtain the media evaluation information, so that when the media creative information is selected, the media evaluation information and the exploration target in the media exploration information are fused for selection, and the media creative information with high putting effect and meeting the exploration target can be selected, so that the obtained media creative information meets the business demand and can effectively improve the interaction rate of the media creative information.

[0230] The modules in the media creative information generation apparatus can be realized by software, hardware, or a combination thereof. The modules can be embedded in or independent of a processor in a computer device in hardware form, or stored in a memory in a computer device in software form, so as to be invoked and executed by a processor to perform operations corresponding to the modules.

[0231] In an embodiment, a computer device, which can be a server, has an internal structure as shown in Figure 12 The computer device includes a processor, a memory, an input / output interface, and a communication interface. The processor, the memory, and the input / output interface are connected through a system bus, and the communication interface is connected to the system bus through the input / output interface. The processor of the computer device is configured to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program, and a database. The internal memory provides an environment for running the operating system and the computer program in the non-volatile storage medium. The database of the computer device is configured to store creative information. The input / output interface of the computer device is configured to exchange information between the processor and external devices. The communication interface of the computer device is configured to communicate with external terminals through a network connection. The computer program is executed by the processor to implement a media creative information generation method.

[0232] In an embodiment, a computer device, which can be a terminal, has an internal structure as shown in Figure 13The computer device includes a processor, a memory, an input / output interface, a communication interface, a display unit and an input device. The processor, the memory and the input / output interface are connected through a system bus, and the communication interface, the display unit and the input device are connected to the system bus through the input / output interface. The processor of the computer device is configured to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system and a computer program. The internal memory provides an environment for running the operating system and the computer program in the non-volatile storage medium. The input / output interface of the computer device is configured to exchange information between the processor and external devices. The communication interface of the computer device is configured to perform wired or wireless communication with external terminals. The wireless communication can be achieved through WIFI, mobile cellular network, NFC (Near Field Communication) or other technologies. The computer program is executed by the processor to implement a media creative information generation method. The display unit of the computer device is configured to form a visually visible picture, which can be a display screen, a projection device or a virtual reality imaging device. The display screen can be a liquid crystal display screen or an electronic ink display screen. The input device of the computer device can be a touch layer overlaid on the display screen, or a key, trackball or touchpad arranged on the shell of the computer device, or an external keyboard, touchpad or mouse, etc.

[0233] Those skilled in the art can understand that, Figure 12 and Figure 13 The structure shown in the figure is only a block diagram of part of the structure related to the scheme of the present application, and does not constitute a limitation on the computer device to which the scheme of the present application is applied. The specific computer device can include more or fewer components than those shown in the figure, or combine certain components, or have a different arrangement of components.

[0234] In one embodiment, a computer device is provided, including a memory and a processor, the memory storing a computer program, and the processor executing the computer program to implement the steps of the media creative information generation method.

[0235] In one embodiment, a computer readable storage medium is provided, storing a computer program, and the computer program is executed by a processor to implement the steps of the media creative information generation method.

[0236] In one embodiment, a computer program product is provided, including a computer program, and the computer program is executed by a processor to implement the steps of the media creative information generation method.

[0237] It should be noted that the user information (including but not limited to user equipment information, user personal information, etc.) and data (including but not limited to data for analysis, stored data, displayed data, etc.) involved in the present application are all information and data authorized by the user or authorized by all parties, and the collection, use and processing of related data need to comply with relevant laws, regulations and standards of the country and region.

[0238] A person of ordinary skill in the art can understand that all or part of the processes in the above-mentioned embodiment methods can be completed by instructing related hardware through a computer program. The computer program can be stored in a non-volatile computer readable storage medium, and when the computer program is executed, the processes of the above-mentioned embodiments of each method can be included. In the embodiments provided in the present application, any reference to memory, database or other medium can include at least one of non-volatile and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical storage, high-density embedded non-volatile memory, resistive memory (ReRAM), magnetoresistive random access memory (MRAM), ferroelectric memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. As an illustration but not limitation, RAM can be in various forms, such as static random access memory (SRAM) or dynamic random access memory (DRAM), etc. The database involved in the embodiments provided in the present application can include at least one of a relational database and a non-relational database. The non-relational database can include a distributed database based on a block chain, etc., without being limited thereto. The processor involved in the embodiments provided in the present application can be a general-purpose processor, a central processing unit, a graphics processing unit, a digital signal processor, a programmable logic device, a data processing logic device based on quantum computing, etc., without being limited thereto.

[0239] The technical features of the above embodiments can be combined in any manner. To make the description concise, not all possible combinations of the technical features in the above embodiments are described, but as long as the combinations of the technical features do not exist contradictions, they should be considered as the scope of the present application.

[0240] The above-described embodiments are merely illustrative of several embodiments of the present application, and the description is relatively specific and detailed, but should not be understood as a limitation on the scope of the patent. It should be noted that for those skilled in the art, without departing from the concept of the present application, a number of modifications and improvements can be made, which are all within the scope of the present application. Therefore, the scope of protection of the present application should be subject to the appended claims.

Claims

1. A method for generating media creative information, characterized in that, The method includes: Acquire seed idea information, media exploration information, and media attribute information; the media exploration information includes the exploration goals and exploration budget of the target object when requesting idea exploration, and the media attribute information includes the industry and audience of the media; Based on the media attribute information and the seed creative information, media guidance information is generated; Based on the media guidance information, the seed creative information and the media exploration information are processed to obtain candidate media creative information; The candidate media creative information is submitted to the media delivery system to obtain media evaluation information; Among the candidate media creative information, media creative information is selected based on the exploration objectives in the media evaluation information and the media exploration information.

2. The method according to claim 1, characterized in that, The generation of media guidance information based on the media attribute information and the seed creative information includes: The guide generator processes the media attribute information and the seed creative information to obtain the first creative style, first creative quality and first creative similarity corresponding to the seed creative information. By combining the first creative style, the first creative quality, and the first creative similarity, media guidance information is obtained.

3. The method according to claim 2, characterized in that, The method further includes: Obtain sample media attribute information, sample seed creative information, and tag information from the data source; The initial guidance generator processes the sample media attribute information and the sample seed creative information to obtain the second creative style, second creative quality and second creative similarity corresponding to the sample seed creative information. The loss value is determined based on the second creative style, the second creative quality, the second creative similarity, and the tag information; The parameters in the initial guidance generator are adjusted based on the loss value to obtain the guidance generator.

4. The method according to claim 1, characterized in that, The candidate media creative information is at least one of the candidate media creative images or candidate media creative text. The method further includes: performing quality verification on the at least one creative information to obtain a verification result; The step of placing the candidate media creative information into the media delivery system includes: If the verification result indicates that the at least one creative information meets the placement conditions, the at least one candidate creative information will be placed in the media placement system.

5. The method according to claim 4, characterized in that, The quality verification of the at least one creative information, and the resulting verification results, include: The candidate media creative images are sequentially subjected to image quality verification and compliance verification to obtain image verification results; The candidate media creative copy is subjected to semantic detection and privacy information detection in sequence to obtain the copy detection results.

6. The method according to claim 1, characterized in that, The acquisition of seed idea information, media exploration information, and media attribute information includes: Receive creative information generation requests initiated by the terminal; From the creative information generation request, seed creative information and media exploration information are read; and media attribute information is obtained based on the creative information generation request. After selecting media creative information from the candidate media creative information based on the media evaluation information and the exploration targets in the media exploration information, the method further includes: The media creative information is sent to the terminal for display.

7. The method according to claim 6, characterized in that, The selection of media creative information based on the exploration objectives in the media evaluation information and the media exploration information from the candidate media creative information includes: Based on the media evaluation information and the exploration objectives in the media exploration information, the candidate media creative information is sorted to obtain a creative information sequence; In the sequence of creative information, media creative information that meets the order condition is selected.

8. The method according to claim 6 or 7, characterized in that, The method further includes: The remaining budget value is determined based on the exploration budget value and exploration consumption value in the media exploration information; The effectiveness of media exploration is determined based on the media evaluation information and the media exploration information. The step of sending the media creative information to the terminal for display includes: The media creative information, the remaining budget value, and the media exploration effect are sent to the terminal for display.

9. The method according to claim 8, characterized in that, The step of sending the media creative information, the remaining budget value, and the media exploration effect to the terminal for display includes: When the media exploration effect meets the exploration goal, or when the media exploration effect does not meet the exploration goal and the remaining budget value is less than a preset threshold, the media creative information, the remaining budget value, and the media exploration effect are sent to the terminal for display. The method further includes: when the media exploration effect does not meet the exploration target and the remaining budget value is greater than or equal to the preset threshold, selecting at least one creative information from the creative information sequence as a new seed creative information, and executing the creative information generation method of the next iteration based on the new seed creative information until the media exploration effect meets the exploration target, or the media exploration effect does not meet the exploration target and the remaining budget value is less than the preset threshold, and stopping the iteration process to obtain the media creative information.

10. The method according to claim 8, characterized in that, The candidate media creative information includes k creative information items; the method further includes: Determine the amount of creative information to be explored in each iteration round and the number of iterations within the preset total time period; Determine the product between the number of iterations and the amount of creative information; The quotient between the exploration budget value and the product value is used as the budget allocation value for each creative information in the candidate media creative information; or... The product of the target cost in the exploration objective and the preset parameters is used as the budget allocation value for each creative information in the candidate media creative information.

11. A method for generating media creative information, characterized in that, The method includes: A creative information generation request carrying seed creative information and media exploration information is sent to the server. The creative information generation request instructs the server to generate media guidance information based on the seed creative information and media attribute information. The seed creative information and media exploration information are then processed according to the media guidance information to obtain candidate media creative information. This candidate media creative information is then deployed to a media delivery system to obtain media evaluation information. From the candidate media creative information, media creative information is selected based on the exploration objectives in the media evaluation information and the media exploration information. The media exploration information includes the exploration objectives and budget value of the target object when requesting creative exploration, and the media attribute information includes the industry and target audience of the media. Receive the media creative information returned by the server in response to the creative information generation request; Display the media creative information.

12. A device for generating media creative information, characterized in that, The device includes: The acquisition module is used to acquire seed creative information, media exploration information, and media attribute information; the media exploration information includes the exploration goals and exploration budget of the target object when requesting creative exploration, and the media attribute information includes the industry and target audience of the media. The first generation module is used to generate media guidance information based on the media attribute information and the seed creative information; The second generation module is used to process the seed creative information and the media exploration information according to the media guidance information to obtain candidate media creative information; The delivery module is used to deliver the candidate media creative information to the media delivery system and obtain media evaluation information. The selection module is used to select media creative information from the candidate media creative information based on the media evaluation information and the exploration targets in the media exploration information.

13. A device for generating media creative information, characterized in that, The device includes: A sending module is used to send a creative information generation request carrying seed creative information and media exploration information to a server. The creative information generation request instructs the server to generate media guidance information based on the seed creative information and media attribute information, process the seed creative information and media exploration information according to the media guidance information to obtain candidate media creative information, and deploy the candidate media creative information to a media delivery system to obtain media evaluation information. From the candidate media creative information, media creative information is selected based on the exploration objectives in the media evaluation information and the media exploration information. The media exploration information includes the exploration objectives and exploration budget of the target object when requesting creative exploration, and the media attribute information includes the industry and target audience of the media. The receiving module is configured to receive the media creative information returned by the server in response to the creative information generation request; The display module is used to display the media creative information.

14. A computer device comprising a memory and a processor, wherein the memory stores a computer program, 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 11.

15. 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 11.

16. A computer program product, comprising a computer program, 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 11.

Images