System for generating marketing materials and marketing material generation method
By introducing the collaborative work of identifying AI agents and functional AI agents in the marketing material generation system, analyzing user needs and generating personalized marketing materials, the problem of insufficient automation and personalization in the existing technology is solved, and efficient and personalized marketing material generation is achieved.
Patent Information
- Application Number
- CN202411576329.2
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-11-06
- Publication Date
- 2025-03-28
Smart Images
Figure CN119720959A_ABST
Abstract
Description
Technical Field
[0001] The present application relates to the technical field of marketing material generation, and in particular to a system for generating marketing materials and a method for generating marketing materials. Background Art
[0002] At present, the existing marketing copy and picture generation technology has the following problems:
[0003] The generated content is not automated or personalized enough, and lacks precise matching for specific application scenarios (such as professional knowledge, product sales, life scenarios, entertainment scenarios, transaction recurrence scenarios, etc.).
[0004] Copywriting generation and image matching require switching and collaboration of multiple tools, which is a cumbersome and inefficient process.
[0005] With existing technologies, copywriting generation and image matching fail to achieve intelligent collaboration and are unable to automatically customize and output high-quality marketing content based on different user needs. Summary of the invention
[0006] The technical problem to be solved by the embodiments of the present application is to provide a system for generating marketing materials and a method for generating marketing materials, so as to effectively solve the problems of insufficient personalization, complicated operation procedures and low efficiency in the prior art.
[0007] In a first aspect, an embodiment of the present application provides a system for generating marketing materials, the system comprising: identifying an AI agent and N functional AI agents, where N is a positive integer greater than or equal to 1, wherein:
[0008] The identification AI agent is used to analyze the text input by the user, identify the user's needs, generate a marketing material generation instruction according to the user needs, and send the marketing material generation instruction to the target function AI agent corresponding to the user needs, wherein the target function AI agent is an AI agent among the N function AI agents;
[0009] The target function AI agent is used to obtain the target industry to which the user belongs, and process the resource information of the target industry according to the data stream of the target function AI agent to generate marketing materials belonging to the target industry and meeting the needs of the user;
[0010] Wherein, the marketing material includes: at least one of a marketing image and a marketing text.
[0011] Optionally, when the user demand is a demand for generating professional knowledge materials, the target function AI agent is an AI agent that matches the professional knowledge materials.
[0012] The target function AI agent is specifically used to generate a marketing theme corresponding to the target industry, use a preset text-image algorithm to process the text input by the user according to the marketing theme, and generate M initial marketing images; use a cutout plug-in to cut out the M initial marketing images to obtain M cutout images; use a drawing board plug-in to splice the M cutout images to generate marketing materials belonging to the target industry that meet the requirements of the professional knowledge materials;
[0013] M is a positive integer greater than or equal to 1.
[0014] Optionally, when the user demand is a demand for generating life scene materials, the target function AI agent is an AI agent that matches the life scene materials.
[0015] The target function AI agent is specifically used to use a random number algorithm to obtain industry knowledge information corresponding to the target industry from the knowledge base; use a topic selection network model to process the industry knowledge information to generate a marketing theme; use a network text search plug-in to search for text information related to the marketing theme, and use an embedded image stream to process the text information to generate an initial marketing image; use a copy editing model to process the text information to generate an initial marketing text; and generate marketing materials belonging to the target industry that meet the needs of the life scenario materials based on the initial marketing image and the initial marketing text.
[0016] Optionally, when the user demand is a demand for generating life scene materials, the target function AI agent is an AI agent that matches the life scene materials.
[0017] The target function AI agent is specifically used to obtain industry knowledge information corresponding to the target industry from the knowledge base using a random number algorithm; use a topic selection network model to process the industry knowledge information to generate a number of marketing topics; use a random number algorithm to screen out a target marketing topic from the several marketing topics, and screen out target knowledge information matching the target marketing topic from the knowledge base; use a copy editing model to process the target knowledge information to generate an initial marketing text; use a text optimization model to perform text optimization processing on the initial marketing text to obtain an intermediate marketing text; use an embedded image stream to process the intermediate marketing text to generate marketing materials belonging to the target industry that meet the needs of the life scene materials.
[0018] Optionally, when the user demand is a demand for generating entertainment scene materials, the target function AI agent is an AI agent that matches the entertainment scene materials.
[0019] The target function AI agent is specifically used to generate a marketing theme corresponding to the target industry, and use a random number algorithm to obtain knowledge information corresponding to the marketing theme from a knowledge base; use a copy editing model to process the knowledge information to obtain an initial marketing copy; use an embedded workflow to process the initial marketing copy to generate marketing materials belonging to the target industry that meet the needs of the entertainment scene materials.
[0020] Optionally, when the user demand is a demand for generating material according to the user input image, the target function AI agent is an AI agent that matches the demand for generating material according to the user input image.
[0021] The target function AI agent is specifically used to obtain the image provided by the user; use an image understanding plug-in to process the image to obtain image understanding information corresponding to the image; use a copy editing model to perform copy editing on the image understanding information to generate marketing materials belonging to the target industry that meet the needs of generating entertainment scene materials.
[0022] Optionally, when the user demand is a demand for generating materials according to user input information, the target function AI agent is an AI agent that matches the demand for generating materials according to user input information.
[0023] The target function AI agent is specifically used to obtain the information of the preset dimension input by the user according to the prompt word; use the copy editing model to process the information of the preset dimension, and generate marketing materials belonging to the target industry that meet the needs of generating materials based on the user input information.
[0024] Optionally, when the user demand is a transaction scenario reproduction demand, the target function AI agent is an AI agent that matches the transaction scenario reproduction demand.
[0025] The target function AI agent is specifically used to obtain the user's historical transaction information within a preset time period from the current time; use the copy editing model to process the historical transaction information to generate marketing materials belonging to the target industry that meet the transaction scenario reproduction requirements.
[0026] Optionally, the system further includes: an AI agent jump module,
[0027] The AI agent jump module is used to obtain the target demand information input by the user when the marketing material generated by the target functional AI agent does not meet the needs of the user, and jump from the target functional AI agent to a functional AI agent among the N functional AI agents that matches the target demand information based on the target demand information, so that the functional AI agent that matches the target demand information can generate marketing materials belonging to the target industry and meeting the target demand information.
[0028] In a second aspect, an embodiment of the present application provides a marketing material generation method, which is applied to any of the above-mentioned systems, wherein the system comprises: identifying an AI agent and N functional AI agents, where N is a positive integer greater than or equal to 1, wherein:
[0029] Calling the recognition AI agent to analyze the text input by the user, identify the user's needs, generate a marketing material generation instruction according to the user needs, and send the marketing material generation instruction to the target function AI agent corresponding to the user needs, the target function AI agent is an AI agent among the N function AI agents;
[0030] Calling the target function AI agent to obtain the target industry to which the user belongs, and processing the resource information of the target industry according to the data stream of the target function AI agent to generate marketing materials belonging to the target industry and meeting the needs of the user;
[0031] Wherein, the marketing material includes: at least one of a marketing image and a marketing text.
[0032] Compared with the prior art, the embodiments of the present application have the following advantages:
[0033] The system provided by the embodiment of the present application includes: an identification AI agent and N functional AI agents, N is a positive integer greater than or equal to 1, wherein the identification AI agent is used to analyze the text input by the user, identify the user's user needs, and generate marketing material generation instructions according to the user needs, and send the marketing material generation instructions to the target functional AI agent corresponding to the user needs, and the target functional AI agent is an AI agent among the N functional AI agents. The target functional AI agent is used to obtain the target industry to which the user belongs, and process the resource information of the target industry according to the data stream of the target functional AI agent, and generate marketing materials belonging to the target industry that meet the user's needs. Among them, the marketing materials include: at least one of marketing images and marketing texts. The embodiment of the present application automatically generates texts and illustrations that are highly matched with user needs through the collaborative work of different AI agents, covering product professional knowledge texts, marketing scene texts, life scene narrations and entertainment scene descriptions, etc., effectively solving the problems of insufficient personalization, complicated operation procedures and low efficiency in the prior art.
[0034] It should be understood that the foregoing general description and the following detailed description are exemplary and explanatory only and are not restrictive of the present application. BRIEF DESCRIPTION OF THE DRAWINGS
[0035] Figure 1 A schematic diagram of the structure of a system for generating marketing materials provided in an embodiment of the present application;
[0036] Figure 2 A schematic diagram of a marketing material generation process provided in an embodiment of the present application;
[0037] Figure 3 A schematic diagram of another marketing material generation process provided in an embodiment of the present application;
[0038] Figure 4 A schematic diagram of another marketing material generation process provided in an embodiment of the present application;
[0039] Figure 5 A schematic diagram of an AI text graph process provided in an embodiment of the present application;
[0040] Figure 6 A schematic diagram of another marketing material generation process provided in an embodiment of the present application;
[0041] Figure 7 A schematic diagram of an embedded workflow provided in an embodiment of the present application;
[0042] Figure 8 A schematic diagram of another embedded workflow provided in an embodiment of the present application;
[0043] Fig. 9A schematic diagram of another embedded workflow provided in an embodiment of the present application;
[0044] Fig.10 A schematic diagram of a process for generating marketing materials based on images provided in an embodiment of the present application;
[0045] Fig.11 A schematic diagram of a process for generating marketing materials according to user input information provided in an embodiment of the present application;
[0046] Fig.12 A flowchart of the steps of a method for generating marketing materials provided in an embodiment of the present application. DETAILED DESCRIPTION
[0047] In order to make the above-mentioned objects, features and advantages of the present application more obvious and easy to understand, the present application is further described in detail below in conjunction with the accompanying drawings and specific implementation methods.
[0048] The terms used in the embodiments of the present application are only for the purpose of describing specific embodiments, and are not intended to limit the present application. The singular forms "a", "said" and "the" used in the embodiments of the present application and the appended claims are also intended to include plural forms, unless the context clearly indicates other meanings.
[0049] Reference Figure 1 , shows a schematic diagram of the structure of a system for generating marketing materials provided by an embodiment of the present application. Figure 1 As shown, the system 100 for generating marketing materials may include: identifying an AI agent 110 and N functional AI agents (as shown in the figure, functional AI agent 1, functional AI agent 2, ..., functional AI agent N), where N is a positive integer greater than or equal to 1.
[0050] Among them, the recognition AI agent 110 can be used to analyze the text input by the user, identify the user's needs, generate marketing material generation instructions based on the user needs, and send the marketing material generation instructions to the target function AI agent corresponding to the user needs. The target function AI agent is an AI agent among N functional AI agents.
[0051] The target function AI agent can be used to obtain the target industry to which the user belongs, and process the resource information of the target industry according to the data flow of the target function AI agent to generate marketing materials belonging to the target industry that meet the user's needs.
[0052] In this embodiment, the system for generating marketing materials is a multi-agent system (MAS), that is, a system composed of multiple AI agents that can perform tasks independently and complete complex tasks together through communication and collaboration. In this embodiment, the system is composed of multiple AI agents, each of which focuses on different subtasks, such as generating copy, generating illustrations, collecting user information, etc., and finally jointly generates complete social media content.
[0053] AI agents (i.e., AI (Artificial Intelligence) AI agents) refer to autonomous systems that have the ability to perceive the environment, make decisions, and take actions. AI agents are able to self-adjust through data input and dynamically adjust strategies according to changes in the external environment. In this embodiment, each AI agent is responsible for handling specific tasks, such as copywriting, image generation, entertainment scenes, or life scenes. Multiple AI agents collaborate to complete the generation of marketing content.
[0054] In specific implementations, AI agents can be mainly divided into the following categories:
[0055] 1. Physical AI agents: such as robots, which can operate and interact in actual physical environments.
[0056] 2. Virtual AI agents: including various software agents, such as virtual assistants, game characters, etc., which interact in the network or virtual environment.
[0057] In this embodiment, the user can start the recognition AI agent by inputting a character (or text, etc.), at which time, the recognition AI agent 110 can interact with the user to prompt the user to input text, or directly obtain the text input by the user. And analyze the text input by the user to identify the user's user needs. In this example, the text input by the user can be a description, a question, or a request. For example, the user can enter his own demand text or click a corresponding button, and the recognition AI agent can analyze the demand text input by the user to identify the user's needs.
[0058] Of course, after the user starts the recognition AI agent by inputting a character (or text, etc.), the recognition AI agent can provide the user with several demand options (such as professional knowledge / life scene soft advertisements / entertainment scene soft advertisements / transaction scene reproduction / matching text according to pictures, etc.), and the user selects the specified option from these demand options to determine the user's needs.
[0059] Afterwards, the recognition AI agent 110 can generate a marketing material generation instruction according to the user's needs, and send the marketing material generation instruction to the target function AI agent corresponding to the user's needs. The target function AI agent is an AI agent among the N function AI agents, that is, the target function AI agent can be Figure 1 The functional AI agent 1, functional AI agent 2, ..., functional AI agent N shown are functional AI agents that can generate marketing materials that meet user needs.
[0060] The target function AI agent can be used to obtain the target industry to which the user belongs. The target industry can be, but is not limited to, the vehicle sales industry, the real estate sales industry, the electronic product sales industry, the scientific research product sales industry, the chemical product sales industry, the health product sales industry, the daily necessities sales industry, the skin care product sales industry, the pet product sales industry, the clothing sales industry, the furniture sales industry, the art sales industry, the food sales industry, the knowledge payment industry, etc.
[0061] In this example, when a user logs in to the system for generating marketing materials provided by this embodiment, the user's account can be obtained. The correspondence between the user's industry and his account can be pre-stored in the system, and the target industry to which the user belongs can be obtained through this correspondence.
[0062] Of course, when using the system for generating marketing materials provided by this embodiment, the user may manually input the target industry to which the user belongs or provide the user with an option to select the target industry to which the user belongs.
[0063] Afterwards, the target function AI agent can process the resource information of the target industry (such as text or images, etc.) according to its own data stream to generate marketing materials belonging to the target industry and meeting user needs. The marketing materials may include: at least one of marketing images and marketing texts.
[0064] The embodiments of the present application automatically generate text and illustrations that are highly matched with user needs through the collaborative work of different AI agents, covering product professional knowledge text, marketing scenario text, life scenario narration and entertainment scenario description, transaction scenario reproduction, etc., effectively solving the problems of insufficient personalization, complicated operating procedures and low efficiency in the prior art.
[0065] In a specific implementation of the present application, when the user demand is to generate professional knowledge materials, the target function AI agent can be an AI agent that matches the professional knowledge materials. The generation process of marketing materials can be as follows: Figure 2 shown.
[0066] The target function AI agent can be specifically used to generate marketing topics corresponding to the target industry. In this example, the target function AI agent can generate marketing topics related to the target industry based on the characteristics of the target industry and the marketing goals (i.e., user needs) using built-in algorithms or databases.
[0067] Afterwards, a preset text-to-image algorithm can be used to process the text input by the user according to the marketing theme to generate M initial marketing images. Figure 2 As shown, two AI text-generating graph algorithms can be used to process the text input by the user to obtain two elements, namely element 1 and element 2. These two elements are the images generated by the AI model.
[0068] In this example, the preset image generation algorithm can be an image generation technology based on deep learning, such as diffusion models or variational autoencoders (VAE) and Transformers. Diffusion models are a generation method of step-by-step denoising, which first gradually adds noise to the input image to make it pure noise, and then gradually "reverse denoise" from the noise to restore it to the target image. This type of model is particularly good at generating images with rich details and high quality.
[0069] The process of AI Wensheng graph can be as follows Figure 5 As shown, the AI text-to-image algorithm is used to generate a corresponding image for the text input by the user or the text recognized in the image input by the user.
[0070] Then, a cutout plug-in may be used to perform cutout processing on M (M is a positive integer greater than or equal to 1) initial marketing images to obtain M cutout images. Figure 2 As shown, two cutout plug-ins can be used to perform cutout processing on element 1 and element 2 respectively to obtain corresponding cutout element 1 and cutout element 2.
[0071] Finally, a drawing board plug-in can be used to stitch the M cutout images to generate marketing materials that belong to the target industry and meet the needs of professional knowledge materials. Specifically, the target function AI agent can use a cutout plug-in (such as Photoshop's cutout tool, online cutout service, etc.) to cutout the M initial marketing images. The cutout plug-in can identify and separate the foreground elements (such as products, people, etc.) in the image, remove the background, and obtain M cutout images. Then use a drawing board plug-in (such as Canvas, GIMP's drawing board function plug-in, etc.) to stitch the M cutout images. This step involves adjustments to the image layout, color matching, element size, etc., to ensure the overall aesthetics and coordination of the final marketing materials. Figure 2As shown, the drawing board plug-in can splice cutout element 1 and cutout element 2 to obtain the final marketing material.
[0072] After the above steps, the target function AI agent can generate the final marketing materials. These materials not only meet the professional characteristics of the target industry, but also meet the requirements of the marketing goals.
[0073] In another specific implementation of the present application, when the user demand is to generate life scene materials, the target function AI agent can be an AI agent that matches the life scene materials. Figure 3 This is described in detail as follows.
[0074] The target function AI agent can be specifically used to obtain industry knowledge information corresponding to the target industry from the knowledge base using a random number algorithm. In this example, the knowledge base refers to a storage body where the system stores and manages structured information, usually including knowledge items, concepts and their interrelationships about a certain field. In this patent, the knowledge base is used to store information related to products, customers or specific fields, and provide a reference for copy generation. The AI agent can extract relevant data from the knowledge base, such as product models, application scenarios, etc., to generate more accurate and personalized copy content. The random number algorithm can ensure that the marketing materials generated each time have a certain degree of novelty and diversity.
[0075] Afterwards, the target function AI agent can use the topic selection network model to process the industry knowledge information and generate marketing topics. That is, the target function AI agent uses the topic selection network model to process the acquired industry knowledge information. The topic selection network model can analyze the key points and hot spots in the knowledge information to generate attractive and relevant marketing topics. Marketing topics can arouse the interest of the target audience and are closely connected with their life scenes to improve the attractiveness and conversion rate of marketing materials. Figure 3 As shown, a random number algorithm is used to randomly select industry knowledge information from the knowledge base, and then it is processed by a large language model (i.e., topic selection model) to generate a marketing topic.
[0076] Then, the target function AI agent can use the network text search plug-in to search for text information related to the marketing theme, and use the embedded image stream to process the text information to generate an initial marketing image. That is, the target function AI agent can use the network text search plug-in to search for relevant text information on the Internet based on the generated marketing theme. The search plug-in can efficiently filter out text information that is highly relevant to the marketing theme and of high quality, providing material for subsequent processing. The searched text information can be processed using an embedded image stream. Image stream technology can analyze keywords and descriptions in the text and generate images that match them. Figure 3 As shown, two text network search plug-ins can be used to search for text information related to the marketing topic. At the same time, the text information can be processed through the image stream embedded in the target function AI agent to generate an initial marketing image.
[0077] The copy editing model is used to process the text information and generate the initial marketing text. Based on the initial marketing image and the initial marketing text, marketing materials that meet the needs of life scene materials belonging to the target industry are generated. The copy editing model is responsible for processing the searched text information and generating the initial marketing text that matches the marketing theme. The copy editing model can extract the key information in the text and express it in concise and powerful language to attract the attention of the target audience. Finally, the initial marketing image and the initial marketing text can be combined to generate marketing materials that meet the needs of life scene materials belonging to the target industry. Figure 3 As shown, the large language model can edit the marketing copy based on the searched text information and the generated initial marketing image. Finally, the initial marketing text and the initial marketing image can be spliced or directly stored to output the final marketing material.
[0078] In another specific implementation of the present application, when the user demand is to generate life scene materials, the target function AI agent can be an AI agent that matches the life scene materials. Figure 4 This is described in detail as follows.
[0079] The target function AI agent can be used to obtain the industry knowledge information corresponding to the target industry from the knowledge base using a random number algorithm. The topic selection network model is used to process the industry knowledge information and generate several marketing topics. Figure 4 As shown in the figure, we can first use the random number algorithm to obtain the industry knowledge information of the target industry from the knowledge base, and then use the topic selection network model to process the obtained industry knowledge information. The topic selection network model can analyze the key points and hot spots in the knowledge information and generate several marketing topics related to the target industry. These topics should be attractive and relevant to arouse the interest of the target audience.
[0080] A random number algorithm is used to select a target marketing topic from a number of marketing topics, and target knowledge information matching the target marketing topic is selected from the knowledge base. Figure 4 As shown, the target function AI agent can use the random number algorithm again to select the target marketing theme from the generated marketing themes. This step is to ensure that the marketing materials finally generated have a unique perspective and appeal. At the same time, the target knowledge information matching the target marketing theme can be selected from the knowledge base, which will be used as the material for subsequent copy editing and image generation.
[0081] Afterwards, the copywriting editing model can be used to process the target knowledge information and generate the initial marketing text. Figure 4 As shown, the large language model (copy editing model) can be used to process the filtered target knowledge information. The copy editing model can analyze the key points and highlights in the knowledge information and express them in concise and powerful language to generate the initial marketing text.
[0082] Then, the text optimization model can be used to optimize the initial marketing text to obtain the intermediate marketing text. Figure 4 As shown, in order to improve the quality and attractiveness of the marketing text, a large language model (text optimization model) can be used to further optimize the initial marketing text. The text optimization model can identify and improve grammatical errors, unclear expressions or lack of attractiveness in the text to obtain an intermediate marketing text.
[0083] Finally, the embedded image stream can be used to process the intermediate marketing text to generate marketing materials that belong to the target industry and meet the needs of life scene materials. The embedded image stream can analyze the keywords and descriptions in the text and generate images that match them. These images should be able to intuitively display the marketing theme and related information while meeting the aesthetics and preferences of the target audience. By combining the optimized text with the generated images, marketing materials that belong to the target industry and meet the needs of life scene materials are generated.
[0084] In another specific implementation of the present application, when the user demand is to generate entertainment scene materials, the target function AI agent can be an AI agent that matches the entertainment scene materials. The generation process of marketing materials can be as follows: Figure 6 shown.
[0085] The target function AI agent can be used to generate marketing topics corresponding to the target industry, and use a random number algorithm to obtain knowledge information corresponding to the marketing topic from the knowledge base. That is, based on the characteristics of the target industry, market trends, and audience preferences, a built-in algorithm or database can be used to generate marketing topics closely related to the target industry. These topics should be innovative, attractive, and targeted, and can resonate with and interest the target audience. Then, Figure 6 As shown in the figure, after the marketing theme is determined, the random number algorithm can be used to randomly select knowledge information corresponding to the marketing theme from the knowledge base. The knowledge base should contain various data, facts, trends, insights, and interesting stories or cases related to the target industry, so as to provide rich materials for subsequent copy editing.
[0086] Afterwards, the copy editing model can be used to process the knowledge information to obtain the initial marketing copy. Figure 6 As shown in the figure, after obtaining the knowledge information, the large language model (i.e., the copy editing model) can be used to process the information. The copy editing model can analyze the key points and highlights in the knowledge information and express them in concise and powerful language to generate attractive and appealing initial marketing copy.
[0087] Finally, the embedded workflow can be used to process the initial marketing copy to generate marketing materials that belong to the target industry and meet the needs of entertainment scene materials. Figure 6 As shown, the target intelligent function body can use three embedded workflows to process the initial marketing copy. The embedded workflow is a comprehensive tool that integrates multiple image processing, video production, audio editing and other functions. It can transform the initial marketing copy into marketing materials that meet the needs of entertainment scene materials. During the processing, appropriate images, videos, audio and other materials can be selected according to the marketing theme and the preferences of the target audience, and organically combined with the initial marketing copy to generate marketing materials that are interesting, interactive and communicative.
[0088] In this example, for Figure 6 The three embedded workflows shown can be combined separately Figure 7 , Figure 8 and Fig. 9 This is described in detail as follows.
[0089] The embedded workflow 1 can be Figure 7 As shown. First, the initial marketing copy output by the large language model can be used as the output of the embedded workflow 1, and character processing can be used to obtain the text after character processing. Then, two character processing methods are used to perform character processing on the processed text twice to obtain two character-processed texts. At the same time, two search plug-ins are used to search for relevant information of the target industry according to the character-processed text, and then the searched content is handed over to the large language model (i.e., the copy editing model). The marketing copy can be edited through the large language model to obtain the written marketing copy. Furthermore, the image search plug-in can be used to search for marketing images according to the written marketing copy. The written marketing copy and the searched marketing image are used to generate the final marketing material, and the marketing material is stored and output as the final marketing material.
[0090] Embedded workflow 2 can be Figure 8As shown. First, the initial marketing copy output by the large language model can be used as the output of the embedded workflow 1, and character processing can be used to obtain the text after character processing. Then, a search plug-in can be used to search for relevant information according to the character-processed text, and the searched text can be handed over to the large language model for editing of the marketing copy to obtain the written marketing copy. Furthermore, an image search plug-in can be used to search for marketing images according to the written marketing copy. The written marketing copy and the searched marketing image are used to generate the final marketing material, and the marketing material is stored and output as the final marketing material.
[0091] Embedded workflow 3 can be Fig. 9 As shown. First, the initial marketing copy output by the large language model can be used as the output of the embedded workflow 1, and character processing can be used to obtain the text after character processing. Then, two search plug-ins can be used to search for relevant information based on the text after character processing, and the text searched by the two search plug-ins is handed over to the large language model (i.e., the copy editing model) for editing to obtain the written marketing copy. Then, the large language model (i.e., the model used to identify key information) can be used to identify keywords in the marketing copy. Then, the image search plug-in is used to search for marketing images based on the identified keywords. Finally, the written marketing copy and the searched marketing image can be used to generate the final marketing material and store the marketing material as the final marketing material for output.
[0092] In another specific implementation of the present application, when the user demand is to generate materials according to the user input image, the target function AI agent can be an AI agent that matches the demand for generating materials according to the user input image. Fig.10 This is described in detail as follows.
[0093] The target function AI agent can be used to obtain images provided by the user. Fig.10 As shown in FIG, the target function AI agent can first receive images uploaded by users. These images can be taken by users themselves or selected from the Internet, and they are usually related to the user's interests, needs or specific scenes of the target industry.
[0094] Afterwards, the image is processed using the image understanding plug-in to obtain the image understanding information corresponding to the image. Fig.10As shown in the figure, after receiving the image, the built-in image understanding plug-in can be used to perform in-depth analysis on the image. The image understanding plug-in can identify key elements such as objects, scenes, colors, textures, etc. in the image, and extract the main information and emotional color conveyed by the image. Through the processing of the image understanding plug-in, the image understanding information corresponding to the image can be generated. This information includes the key elements, overall atmosphere, emotional tendency, etc. in the image, providing an important basis for subsequent copy editing.
[0095] Finally, the copy editing model can be used to edit the image understanding information to generate marketing materials that belong to the target industry and meet the needs of generating entertainment scene materials. Fig.10 As shown in the figure, after obtaining the image understanding information, the copy editing model can be used to process this information. The copy editing model can analyze the key points and highlights in the image understanding information, and generate creative, attractive and targeted marketing copy in combination with the characteristics of the target industry and the needs of the entertainment scene. The copy editing model will consider factors such as the language style, word choice, and sentence structure of the copy to ensure that the generated copy is in line with the tonality of the target industry and can resonate with and interest the audience. Finally, the generated marketing copy can be combined with the original image or the processed image to generate marketing materials that meet the needs of the target industry and entertainment scene.
[0096] In another specific implementation of the present application, when the user demand is to generate materials according to the user input information, the target function AI agent can be an AI agent that matches the demand for generating materials according to the user input information. Fig.11 Describe in detail.
[0097] The target function AI agent can be used to obtain information of preset dimensions input by the user according to the prompt word. Fig.11 As shown, the preset dimension information input by the user according to the prompt words can be received first. These prompt words may include product features, target audiences, marketing scenarios, brand tone and other aspects, and the user needs to provide corresponding information according to these prompt words. The preset dimension information may include text descriptions, digital data, keywords, etc., which together constitute the basic content of marketing materials.
[0098] After receiving the information input by the user, the target function AI agent will pre-process the information. This may include data cleaning, format conversion, keyword extraction, etc. to ensure the accuracy and consistency of the information.
[0099] Finally, the copywriting editing model can be used to process the information of the preset dimensions to generate marketing materials belonging to the target industry that meet the needs of generating materials based on user input information. Fig.11As shown in the figure, after obtaining the information of the preset dimensions, the large language model (i.e., the copy editing model) can be used to create the copy. The copy editing model can automatically generate creative, attractive, and targeted marketing copy based on the information input by the user and the characteristics of the target industry. The copy editing model will consider multiple aspects such as the language style, word choice, sentence structure, and emotional color of the copy to ensure that the generated copy is in line with the tonality of the target industry and can effectively convey the marketing information. Finally, the generated marketing copy is combined with the information of the preset dimensions to generate marketing materials that meet the needs of generating materials based on the user input information.
[0100] In another specific implementation of the present application, when the user demand is a transaction scene reproduction demand, the target function AI agent is an AI agent that matches the transaction scene reproduction demand.
[0101] The target function AI agent can be used to obtain the user's historical transaction information within a preset time from the current time, and use the copy editing model to process the historical transaction information to generate marketing materials belonging to the target industry that meet the transaction scenario reproduction requirements.
[0102] Specifically, the target function AI agent can guide the user to input historical order information of recent transactions. After obtaining the historical order information of recent transactions entered by the user (such as age, gender, city, occupation, etc.), the copy editing model can write marketing materials corresponding to the transaction scenario reproduction needs based on this information. For example, taking car sales as an example, a transaction story copy can be written based on the process of customers looking for marketing personnel to buy a car. This transaction story copy can be used as marketing material for reproducing car sales scenarios.
[0103] In another specific implementation of the present application, the system for generating marketing materials may further include: an AI agent jump module,
[0104] The AI agent jump module can be used to obtain the target demand information input by the user when the marketing materials generated by the target functional AI agent do not meet the user's needs, and jump from the target functional AI agent to a functional AI agent among N functional AI agents that matches the target demand information based on the target demand information, so that the functional AI agent that matches the target demand information can generate marketing materials that belong to the target industry and meet the target demand information.
[0105] In this embodiment, when the marketing material generated by the target function AI agent is reported by the user as not meeting the needs, the AI agent jump system can provide the user with several demand options (such as professional knowledge / soft advertisements for life scenarios / soft advertisements for entertainment scenarios / reproduction of transaction scenarios / matching texts according to pictures, etc.). Then, the user selects his own needs from these demand options (which can be the same as the previous needs), and the AI agent jump system can jump to the functional AI agent corresponding to the needs selected by the user to execute the process of generating marketing materials.
[0106] It is understandable that the demand selected by the user this time may be the same as the previous demand, so the target function body may generate different marketing materials again.
[0107] In this embodiment, the system for generating marketing materials can adopt a single AI agent design, that is, design a more complex and comprehensive single AI agent, so that it can independently complete copy generation, image generation and user interaction by processing different task modules in parallel. This method can simplify the system architecture and reduce communication costs.
[0108] The embodiment of the present application completes the generation of text and images through the collaborative work of multiple independent AI agents, and each AI agent is responsible for a specific task module, including automatic topic selection, text generation, image generation, scene writing, etc. Different from the existing single model generation system, this embodiment realizes the improvement of the efficiency and personalization of content generation through the division of labor of AI agents. Multiple AI agents perform tasks independently and collaborate through a unified scheduling mechanism to complete the process of text and image generation.
[0109] The system provided by the embodiment of the present application includes: an identification AI agent and N functional AI agents, N is a positive integer greater than or equal to 1, wherein the identification AI agent is used to analyze the text input by the user, identify the user's user needs, and generate marketing material generation instructions according to the user needs, and send the marketing material generation instructions to the target functional AI agent corresponding to the user needs, and the target functional AI agent is an AI agent among the N functional AI agents. The target functional AI agent is used to obtain the target industry to which the user belongs, and process the resource information of the target industry according to the data stream of the target functional AI agent, and generate marketing materials belonging to the target industry that meet the user's needs. Among them, the marketing materials include: at least one of marketing images and marketing texts. The embodiment of the present application automatically generates texts and illustrations that are highly matched with user needs through the collaborative work of different AI agents, covering product professional knowledge texts, marketing scene texts, life scene narrations and entertainment scene descriptions, etc., effectively solving the problems of insufficient personalization, complicated operation procedures and low efficiency in the prior art.
[0110] Reference Fig.12, shows a flowchart of the steps of a marketing material generation method provided in an embodiment of the present application, and the marketing material generation method can be applied to the system provided in the above embodiment. Fig.12 As shown, the marketing material generation method may include: step 1201 and step 1202.
[0111] Step 1201: Call the recognition AI agent to analyze the text input by the user, identify the user's needs, generate marketing material generation instructions based on the user needs, and send the marketing material generation instructions to the target functional AI agent corresponding to the user needs. The target functional AI agent is an AI agent among the N functional AI agents.
[0112] The embodiment of the present application can be applied to the system for generating marketing materials provided in the above embodiment. The system may include: identifying an AI agent and N functional AI agents, where N is a positive integer greater than or equal to 1.
[0113] In a specific implementation, the recognition AI agent can be called to analyze the text input by the user, identify the user's needs, generate marketing material generation instructions based on the user needs, and send the marketing material generation instructions to the target functional AI agent corresponding to the user needs. The target functional AI agent is an AI agent among the N functional AI agents.
[0114] Step 1202: Call the target function AI agent to obtain the target industry to which the user belongs, and process the resource information of the target industry according to the data flow of the target function AI agent to generate marketing materials belonging to the target industry that meet the needs of the user.
[0115] Then, the target function AI agent can be called to obtain the target industry to which the user belongs, and the resource information of the target industry can be processed according to the data flow of the target function AI agent to generate marketing materials belonging to the target industry that meet the needs of the user. The marketing materials can include: at least one of marketing images and marketing texts.
[0116] The target function AI agent can be any of: an AI agent that matches the need to generate materials based on user input information, an AI agent that matches the need to generate materials based on user input images, an AI agent that matches entertainment scene materials, an AI agent that matches life scene materials, an AI agent that matches professional knowledge materials, and an AI agent that matches the need to reproduce transaction scenes. The implementation process of the target function AI agent generating marketing materials for these functions can refer to the description of the above embodiment, and this embodiment will not be repeated here.
[0117] The marketing material generation method provided in the embodiment of the present application calls the recognition AI agent, analyzes the text input by the user, identifies the user's needs, generates marketing material generation instructions according to the user's needs, and sends the marketing material generation instructions to the target function AI agent corresponding to the user's needs. The target function AI agent is an AI agent among N functional AI agents. The target function AI agent is called to obtain the target industry to which the user belongs, and the resource information of the target industry is processed according to the data stream of the target function AI agent to generate marketing materials belonging to the target industry that meet the user's needs, wherein the marketing materials include: at least one of marketing images and marketing texts. The embodiment of the present application automatically generates texts and illustrations that are highly matched with user needs through the collaborative work of different AI agents, covering product professional knowledge texts, marketing scene texts, life scene narrations, and entertainment scene descriptions, etc., effectively solving the problems of insufficient personalization, complicated operation procedures, and low efficiency in the prior art.
[0118] Although the preferred embodiments of the present application have been described, those skilled in the art may make additional changes and modifications to these embodiments once they have learned the basic creative concept. Therefore, the appended claims are intended to be interpreted as including the preferred embodiments and all changes and modifications that fall within the scope of the embodiments of the present application.
[0119] Finally, it should be noted that, in this article, relational terms such as first and second, etc. are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or terminal including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or terminal. In the absence of further restrictions, the elements defined by the statement "comprise a ..." do not exclude the presence of other identical elements in the process, method, article or terminal including the elements.
[0120] The above is a detailed introduction to a system for generating marketing materials and a method for generating marketing materials provided by the present application. Specific examples are used in this article to illustrate the principles and implementation methods of the present application. The description of the above embodiments is only used to help understand the method of the present application and its core idea; at the same time, for general technical personnel in this field, according to the idea of the present application, there will be changes in the specific implementation method and application scope. In summary, the content of this specification should not be understood as a limitation on the present application.
Claims
1. A system for generating marketing materials, characterized in that: The system includes: identifying an AI agent and N functional AI agents, where N is a positive integer greater than or equal to 1, wherein: The identification AI agent is used to analyze the text input by the user, identify the user's needs, generate a marketing material generation instruction according to the user needs, and send the marketing material generation instruction to the target function AI agent corresponding to the user needs, wherein the target function AI agent is an AI agent among the N function AI agents; The target function AI agent is used to obtain the target industry to which the user belongs, and process the resource information of the target industry according to the data stream of the target function AI agent to generate marketing materials belonging to the target industry and meeting the needs of the user; Wherein, the marketing material includes: at least one of a marketing image and a marketing text.
2. The system according to claim 1, characterized in that When the user demand is a demand for generating professional knowledge materials, the target function AI agent is an AI agent that matches the professional knowledge materials. The target function AI agent is specifically used to generate a marketing theme corresponding to the target industry, use a preset text-image algorithm to process the text input by the user according to the marketing theme, and generate M initial marketing images; use a cutout plug-in to cut out the M initial marketing images to obtain M cutout images; use a drawing board plug-in to splice the M cutout images to generate marketing materials belonging to the target industry that meet the requirements of the professional knowledge materials; M is a positive integer greater than or equal to 1.
3. The system according to claim 1, characterized in that When the user demand is a demand for generating life scene materials, the target function AI agent is an AI agent that matches the life scene materials. The target function AI agent is specifically used to use a random number algorithm to obtain industry knowledge information corresponding to the target industry from the knowledge base; use a topic selection network model to process the industry knowledge information to generate a marketing theme; use a network text search plug-in to search for text information related to the marketing theme, and use an embedded image stream to process the text information to generate an initial marketing image; use a copy editing model to process the text information to generate an initial marketing text; and generate marketing materials belonging to the target industry that meet the needs of the life scenario materials based on the initial marketing image and the initial marketing text.
4. The system according to claim 1, characterized in that When the user demand is a demand for generating life scene materials, the target function AI agent is an AI agent that matches the life scene materials. The target function AI agent is specifically used to obtain industry knowledge information corresponding to the target industry from the knowledge base using a random number algorithm; process the industry knowledge information using a topic selection network model to generate a number of marketing topics; filter out a target marketing topic from the several marketing topics using a random number algorithm, and filter out target knowledge information matching the target marketing topic from the knowledge base; process the target knowledge information using a copywriting editing model to generate an initial marketing text; The initial marketing text is optimized by using a text optimization model to obtain an intermediate marketing text; the intermediate marketing text is processed by using an embedded image stream to generate marketing materials belonging to the target industry and meeting the needs of the life scenario materials.
5. The system according to claim 1, characterized in that When the user demand is a demand for generating entertainment scene materials, the target function AI agent is an AI agent that matches the entertainment scene materials. The target function AI agent is specifically used to generate a marketing theme corresponding to the target industry, and use a random number algorithm to obtain knowledge information corresponding to the marketing theme from a knowledge base; use a copy editing model to process the knowledge information to obtain an initial marketing copy; use an embedded workflow to process the initial marketing copy to generate marketing materials belonging to the target industry that meet the needs of the entertainment scene materials.
6. The system according to claim 1, characterized in that When the user demand is a demand for generating material according to the user input image, the target function AI agent is an AI agent that matches the demand for generating material according to the user input image. The target function AI agent is specifically used to obtain the image provided by the user; use the image understanding plug-in to process the image to obtain image understanding information corresponding to the image; The image understanding information is processed by copy editing using a copy editing model to generate marketing materials belonging to the target industry and meeting the requirements for generating entertainment scene materials.
7. The system according to claim 1, characterized in that When the user demand is a demand for generating materials according to user input information, the target function AI agent is an AI agent that matches the demand for generating materials according to user input information. The target function AI agent is specifically used to obtain the information of the preset dimension input by the user according to the prompt word; use the copy editing model to process the information of the preset dimension, and generate marketing materials belonging to the target industry that meet the needs of generating materials based on the user input information.
8. The system according to claim 1, characterized in that When the user demand is a transaction scenario reproduction demand, the target function AI agent is an AI agent that matches the transaction scenario reproduction demand, The target function AI agent is specifically used to obtain the user's historical transaction information within a preset time period from the current time; use the copy editing model to process the historical transaction information to generate marketing materials belonging to the target industry that meet the transaction scenario reproduction requirements.
9. The system according to claim 1, characterized in that The system also includes: an AI agent jump module, The AI agent jump module is used to obtain the target demand information input by the user when the marketing material generated by the target functional AI agent does not meet the needs of the user, and jump from the target functional AI agent to a functional AI agent among the N functional AI agents that matches the target demand information based on the target demand information, so that the functional AI agent that matches the target demand information can generate marketing materials belonging to the target industry and meeting the target demand information.
10. A method for generating marketing materials, applied to the system according to any one of claims 1 to 9, characterized in that: The system includes: identifying an AI agent and N functional AI agents, where N is a positive integer greater than or equal to 1, wherein: Calling the recognition AI agent to analyze the text input by the user, identify the user's needs, generate a marketing material generation instruction according to the user needs, and send the marketing material generation instruction to the target function AI agent corresponding to the user needs, the target function AI agent is an AI agent among the N function AI agents; Calling the target function AI agent to obtain the target industry to which the user belongs, and processing the resource information of the target industry according to the data stream of the target function AI agent to generate marketing materials belonging to the target industry and meeting the needs of the user; Wherein, the marketing material includes: at least one of a marketing image and a marketing text.