Marketing material generation method and system, electronic equipment and storage medium

Abstract

The invention provides a marketing material generation method and system, electronic equipment and a storage medium. The method comprises the following steps: acquiring marketing materials and customer data; according to the marketing material and the customer data, determining an initial preference of the customer for the marketing material; performing emotion recognition on the customer data to obtain an emotion recognition result; determining marketing materials of the customer according to the emotion recognition result and the initial preference; different customers correspond to different marketing materials; namely, through a plurality of links of data collection, data analysis, emotion recognition, personalized marketing and the like, demands and preferences of customers can be better known, and more effective marketing strategies are formulated to attract and retain the customers, so that the efficiency of customizing marketing materials for the users is improved, all customers can be oriented, emotion connection with the customers is emphasized, and the marketing efficiency is improved. Marketing materials have infectivity and stimulate interests and resonance of customers.

Description

Technical Field

[0001] The present invention belongs to the field of data processing technology, and more specifically, particularly relates to a marketing material generation method and system, electronic equipment, and storage medium. Background Art

[0002] As competition in the banking industry becomes increasingly fierce, the effectiveness and efficiency of marketing activities have become one of the key factors that determine the market performance of banks. Among them, the rapid generation and personalized customization of marketing materials have become important issues that need to be solved in the field of bank marketing.

[0003] At present, the traditional method relies on the creativity and skills of professional designers, who carefully design and produce various marketing materials such as exquisite posters, detailed brochures and attractive video advertisements according to the bank's brand image, market strategy and customer needs. Although this method can ensure high quality and certain innovation of the design, its production efficiency is low and it is difficult to cope with large-scale and diversified marketing needs.

[0004] However, the manual designer production method is time-consuming and difficult to keep up with the rapidly changing market rhythm. Although manual design has certain innovations, it is difficult to cover all customers and often ignores the emotional connection with customers, resulting in marketing content that lacks appeal and is difficult to arouse customer interest and resonance. Summary of the invention

[0005] In view of this, the purpose of the present invention is to provide a marketing material generation method and system, an electronic device, and a storage medium, which can be used to better understand customer needs and preferences through multiple links such as data collection, data analysis, emotion recognition and personalized marketing, and formulate more effective marketing strategies to attract and retain customers.

[0006] The first aspect of the present application discloses a method for generating marketing materials, comprising:

[0007] Obtain marketing materials and customer data;

[0008] Determining the customer's initial preference for the marketing material based on the marketing material and the customer data;

[0009] Performing emotion recognition on the customer data to obtain an emotion recognition result;

[0010] The marketing material for the customer is determined based on the emotion recognition result and the initial preference; different customers have different corresponding marketing materials.

[0011] Optionally, determining the customer's initial preference for the marketing material based on the marketing material and the customer data includes:

[0012] The marketing material and the customer data are input into a prediction model, and the prediction model processes the marketing material and the customer data and outputs the initial preference.

[0013] Optionally, the process of constructing the prediction model includes:

[0014] Obtain historical marketing materials and historical customer data for preprocessing to obtain training samples;

[0015] Based on the training samples, a deep learning algorithm is used for training to obtain an initial prediction model;

[0016] The initial prediction model is optimized and iterated to obtain a final prediction model; the prediction model is used for marketing materials and customer data to determine the user's initial preference for the marketing materials.

[0017] Optionally, determining personalized marketing materials for customers based on the emotion recognition results and the initial preferences includes:

[0018] inputting the emotion recognition result and the initial preference into a material model;

[0019] The material model processes the emotion recognition result and the initial preference, and outputs the emotional tendency of the customer;

[0020] Based on the emotional tendency, marketing materials targeting the customer are determined.

[0021] Optionally, the process of constructing the material model includes:

[0022] Obtain historical emotion recognition results and historical initial preferences;

[0023] Using the historical emotion recognition results and historical initial preferences as target samples;

[0024] Using large model technology to train the target sample to obtain an initial material model;

[0025] The initial material model is iteratively optimized to obtain a final material model; the material model performs emotion recognition on the customer data to obtain an emotion recognition result.

[0026] Optionally, after determining the marketing material of the customer based on the emotion recognition result and the initial preference, the method further includes:

[0027] Preview the client's marketing materials.

[0028] Optionally, after determining the marketing material of the customer based on the emotion recognition result and the initial preference, the method further includes:

[0029] Get feedback from customers.

[0030] The second aspect of the present application discloses a marketing material generation system, comprising:

[0031] Acquisition module, used to obtain marketing materials and customer data;

[0032] A preference module, used to determine the customer's initial preference for the marketing material based on the marketing material and the customer data;

[0033] An emotion module, used to perform emotion recognition on the customer data to obtain an emotion recognition result;

[0034] The material module is used to determine the customer's marketing material based on the emotion recognition result and the initial preference; different customers have different corresponding marketing materials.

[0035] The third aspect of the present application discloses an electronic device, comprising:

[0036] one or more processors;

[0037] a storage device having one or more programs stored thereon;

[0038] When one or more programs are executed by one or more processors, the one or more processors implement the marketing material generation method as described in any one of the first aspects of the present application.

[0039] A fourth aspect of the present application discloses a storage medium having a computer program stored thereon, wherein when the computer program is executed by a processor, the marketing material generation method as described in any one of the first aspects of the present application is implemented.

[0040] It can be seen from the above technical scheme that a marketing material generation method provided by the present invention includes: obtaining marketing materials and customer data; determining the customer's initial preference for the marketing materials based on the marketing materials and the customer data; performing emotion recognition on the customer data to obtain emotion recognition results; determining the customer's marketing materials based on the emotion recognition results and the initial preferences; different customers have different corresponding marketing materials; that is, through multiple links such as data collection, data analysis, emotion recognition and personalized marketing, it is possible to better understand customer needs and preferences, and formulate more effective marketing strategies to attract and retain customers, improve the efficiency of customizing marketing materials for users, and can be oriented to all customers and focus on emotional connections with customers. The marketing materials are contagious and stimulate customer interest and resonance. BRIEF DESCRIPTION OF THE DRAWINGS

[0041] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0042] Figure 1 is a flow chart of a method for generating marketing materials provided by an embodiment of the present invention;

[0043] Figure 2 is a flow chart of another method for generating marketing materials provided by an embodiment of the present invention;

[0044] Figure 3 is a schematic diagram of a marketing material generation system provided by an embodiment of the present invention;

[0045] Figure 4 It is a schematic diagram of an electronic device provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0046] In order to make the purpose, technical solution and advantages of the embodiments of the present invention clearer, the technical solution in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0047] In the present application, the term "comprise", "include" or any other variant thereof is intended to cover non-exclusive inclusion, so that the process, method, article or equipment including a series of elements includes not only those elements, but also includes other elements not explicitly listed, or also includes elements inherent to such process, method, article or equipment. In the absence of more restrictions, the elements defined by the sentence "comprise one ..." do not exclude the existence of other identical elements in the process, method, article or equipment including the elements. In addition, the terms "first", "second", "third", "fourth", etc. (if present) in the specification and claims of the present application and the above-mentioned drawings are used to distinguish similar objects, and do not have to be used to describe a specific order or sequence. It should be understood that the data used in this way can be interchangeable in appropriate circumstances, so that the embodiments of the present application described here can be implemented in an order other than those illustrated or described here, for example.

[0048] The embodiments of the present application provide a method for generating marketing materials, which is used to solve the problem that the method of manual designer production in the prior art is time-consuming and difficult to keep up with the rapidly changing market rhythm. Although manual design has certain innovations, it is difficult to cover all customers and often ignores the emotional connection with customers, resulting in the marketing content lacking appeal and being difficult to arouse customer interest and resonance.

[0049] Explanation of relevant terms:

[0050] Big model technology. In the field of artificial intelligence, machine learning models are trained using ultra-large-scale parameters and data sets. These models usually have complex network structures and powerful computing capabilities, and are able to process massive amounts of data and learn complex patterns. In the fields of natural language processing, image processing, and speech recognition, big model technology has achieved remarkable results. In the application of automatic generation of marketing materials for banks, big model technology can learn and understand a large amount of historical marketing materials and customer data, thereby generating high-quality and diverse marketing content.

[0051] Affective computing. An AI technology that studies and develops the ability to identify, interpret, process and simulate human emotions. It extracts emotional information from multimodal data such as text, speech and images, and then understands and simulates human emotional states. In the field of banking marketing, affective computing technology can help banks gain a deeper understanding of customers' emotional needs and preferences, thereby generating marketing materials that better meet customers' psychological expectations and improve marketing effectiveness and customer satisfaction.

[0052] AI generation model. Refers to a model that is built using artificial intelligence technology and can automatically generate specific types of data. These models are usually based on deep learning algorithms, which learn the inherent laws and representations of data through a large amount of training data, so as to generate new data similar to the training data. In the application of automatic generation of bank marketing materials, AI generation models can automatically generate marketing materials such as pictures, texts, videos, etc. that meet the requirements according to customer needs and preferences.

[0053] Personalization. It means providing customized products or services according to customer needs and preferences. In the field of banking marketing, personalization means that banks can provide targeted marketing materials and strategies according to the characteristics and needs of different customers. By combining emotional computing technology and big model technology, banks can more accurately grasp the personalized needs of customers, realize personalized customization of marketing materials, and improve marketing effectiveness and customer satisfaction.

[0054] See also Figure 1 , marketing material generation method, including:

[0055] S101. Obtain marketing materials and customer data.

[0056] Marketing materials and customer data can be collected from multiple channels.

[0057] Marketing materials may include advertising copy, pictures, videos, product descriptions, etc., which are used to attract customer attention and convey product information.

[0058] Customer data may include transaction records (such as activity browsing, participation in transactions), personal information (age, gender, region), consumption habits (finance, entertainment, life, etc.), interaction history (entry from Douyin ads, purchase of financial products, WeChat fission sharing, etc.), preference settings (visual preferences, such as using the standard version or large font version, whether investment preferences are conservative or aggressive, etc.), behavioral data (browsing, clicking, purchasing, etc.), social media data (purchased data, such as the performance of third-party platforms), geographic location data (southeast coastal areas or sinking markets, etc.) and economic conditions (assets and liabilities, etc.).

[0059] This customer data can be used to deeply analyze customer behavior and preferences through big model technology to generate personalized and emotional marketing materials.

[0060] S102. Determine the customer's initial preference for the marketing material based on the marketing material and customer data.

[0061] After collecting marketing materials and customer data, companies need to use data analysis tools or algorithms to identify which marketing materials customers are more interested in. This may involve analyzing data such as customers' purchase history and browsing history to find out the similarities and differences in customer preferences. Through these analyses, companies can initially understand customers' preferences for marketing materials and lay the foundation for subsequent personalized marketing.

[0062] After successfully acquiring marketing materials and customer data, we can dig deeper into the information behind these data to accurately determine the degree of customer preference for different marketing materials. This step may include a detailed analysis of multi-dimensional information such as customer purchase history, online browsing behavior, click and interaction data. By comparing and summarizing these customer behavior data, we can reveal the general rules and individual differences in customer preferences, such as which types of marketing materials are more attractive to customers of a certain age group, or which content themes are more likely to resonate with specific interest groups.

[0063] Based on these detailed analysis results, companies can initially build a cognitive framework of customers' preferences for marketing materials, which not only helps companies grasp market trends, but also provides a solid basis for the subsequent formulation of personalized marketing strategies. By deeply understanding the unique preferences of each customer, companies can more specifically select or create marketing materials to ensure the accuracy and effectiveness of information transmission, thereby gaining an advantage in the fierce market competition.

[0064] S103: Perform emotion recognition on customer data to obtain emotion recognition results.

[0065] In addition to analyzing text data, we can also expand the boundaries of emotion recognition and use advanced emotion recognition technology to analyze customers' voice information and facial expressions. By analyzing the customer's voice intonation, we can capture the emotional fluctuations contained in the voice, such as joy, frustration or anger. At the same time, using facial recognition technology, we can further interpret the customer's non-verbal emotional expressions, such as smiling, frowning and other facial expressions, which can provide richer dimensions for emotion analysis.

[0066] The comprehensive application of these emotion recognition technologies can more comprehensively grasp the emotional state of customers and deeply understand their inner feelings and real needs. On this basis, more considerate and targeted marketing strategies can be formulated, which can not only effectively improve customer experience, but also enhance the emotional connection between brands and consumers, laying a solid emotional foundation for the long-term development of enterprises.

[0067] This step uses emotional computing technology to process customer data. Emotional computing technology is an advanced method that analyzes multiple modal data such as text, voice, and images of customers in detail, and ultimately extracts a series of quantitative emotional labels, scores, trends, and customer preferences and demand characteristics. These data are like a mirror that truly reflects the psychological expectations and emotional state of customers.

[0068] Taking a real-world scenario as an example, if the emotional computing technology identifies that the customer's emotional label is "joy", and the customer prefers conservativeness and is about 50 years old, then the activity page may be designed in the traditional form of opening red envelopes to cater to the customer's preferences. On the contrary, if the emotional label is also "joy", but the customer prefers entertainment and is around 20 years old, the activity page may adopt the idea of ​​opening blind boxes to attract the interest of young users.

[0069] As another core technology of this invention, emotional computing technology is responsible for accurately identifying, deeply interpreting and properly handling customers' emotional needs and preferences. With this technology, we can gain a deeper insight into customers' psychological expectations and emotional states, thus providing strong support for customizing marketing materials that meet customer needs.

[0070] Specifically, emotional computing technology covers the following key links:

[0071] Emotion recognition: By comprehensively analyzing the customer's text, voice, image and other multimodal data, we can accurately capture the emotional information and demand characteristics. In this process, a variety of methods such as voice analysis, face recognition, body language recognition (with the help of situational awareness technology) and behavior pattern analysis are used to ensure that the customer's emotional information can be fully and accurately captured when interacting with the mobile banking application.

[0072] Emotional understanding: After identifying emotional information, we will conduct in-depth analysis and understanding to extract the customer's emotional tendencies and demand preferences. This step is crucial for the subsequent generation of marketing materials that meet the customer's psychological expectations.

[0073] Emotional simulation: Based on the emotional needs and preferences of customers, we simulate human emotional states and carefully create marketing materials that can touch the hearts of customers. Compared with traditional marketing materials that are only generated based on users' explicit preferences and behavioral habits, emotional computing technology can more deeply understand users' emotional states and psychological expectations through multi-dimensional data analysis, thereby creating more emotionally resonant and personalized content.

[0074] Emotional feedback: We attach great importance to customer feedback and evaluation of generated materials. By collecting this information, we continuously optimize and improve the emotional computing model to ensure that the generated marketing materials can continuously improve the quality and customer satisfaction.

[0075] This approach not only enhances the appeal and appeal of marketing materials, but also deepens users’ emotional connection, effectively improving user engagement and conversion rates.

[0076] S104. Determine the customer's marketing materials based on the emotion recognition results and initial preferences.

[0077] Different customers have different corresponding marketing materials, which enables the generation of personalized marketing materials, and personalized materials are related to emotions and preferences.

[0078] After mastering the emotion recognition results and the preliminary preference information of customers, we have a solid foundation for tailoring personalized marketing materials for customers. This means that we can carefully plan and push exclusive advertising copy, images, videos and other forms of marketing content based on each customer's unique interests and emotional tendencies. These customized materials not only fit the customer's preferences, but also accurately touch their emotional resonance points, achieving in-depth communication and interaction.

[0079] In other words, by combining big model technology and emotional computing technology, the quality of automatic generation of bank marketing materials can be improved, and personalized and emotional marketing content creation can be achieved, thereby improving the bank's marketing effectiveness and customer satisfaction.

[0080] The charm of personalized marketing lies in its accuracy and differentiation, ensuring that every customer can receive the marketing information that best suits them. This not only significantly improves customer participation enthusiasm and satisfaction, making them feel the pleasure of being valued and understood, but also brings more direct sales conversion and customer loyalty. Through this strategy, not only the emotional connection with customers is deepened, but also the building of brand loyalty is effectively promoted, laying a solid foundation for continued business growth and market share expansion.

[0081] In this embodiment, marketing materials and customer data are obtained; based on the marketing materials and the customer data, the customer's initial preference for the marketing materials is determined; emotion recognition is performed on the customer data to obtain an emotion recognition result; based on the emotion recognition result and the initial preference, the customer's marketing materials are determined; different customers have different corresponding marketing materials; that is, through multiple links such as data collection, data analysis, emotion recognition and personalized marketing, the customer's needs and preferences can be better understood, and more effective marketing strategies can be formulated to attract and retain customers, and the efficiency of customizing marketing materials for users can be improved. It can be oriented to all customers and focus on emotional connections with customers. The marketing materials are contagious and stimulate the interest and resonance of customers.

[0082] Optionally, based on the marketing materials and customer data, determine the customer's initial preference for the marketing materials, including:

[0083] Marketing materials and customer data are input into the predictive model, which processes the marketing materials and customer data and outputs initial preferences.

[0084] The carefully planned marketing materials and detailed customer data are input into the prediction model. This model uses advanced algorithms and technologies to conduct in-depth analysis and processing of the input marketing materials (such as advertising copy, visual elements, promotion strategies, etc.) and customer data (including purchase history, browsing behavior, interests, hobbies, age, gender and other personal information). After a series of complex calculations and comparisons, the model will eventually generate and output a report or score on the customer's initial preference for marketing content, providing strong data support and reference for the subsequent formulation of personalized marketing strategies.

[0085] The predictive model can learn and understand the complex patterns and inherent laws in these marketing materials and customer data, which may include semantic understanding of text, extraction of visual features of images, etc. These features and patterns will be used for further processing of AI-generated models (material models).

[0086] Optionally, the process of building a predictive model includes:

[0087] Historical marketing materials and historical customer data are obtained for preprocessing to obtain training samples; based on the training samples, deep learning algorithms are used for training to obtain an initial prediction model; the initial prediction model is optimized and iterated to obtain a final prediction model; the prediction model is used for marketing materials and customer data to determine the user's initial preference for marketing materials.

[0088] Deep learning algorithms, especially the combination of deep neural networks (DNN) and convolutional neural networks (CNN), are used to process and analyze a large amount of historical marketing materials and customer data. Specifically, the input of the prediction model is mainly a large amount of historical marketing materials and customer data, including text descriptions, image designs, video ads, etc. These data are divided into training sets, validation sets, and test sets to support model training, validation, and testing.

[0089] First, we collected historical marketing materials (including advertising copy, pictures, videos, etc.) and historical customer data (including purchase records, browsing behavior, personal information, etc.) from company databases, third-party data sources, or internal systems. Then, we performed detailed preprocessing on these data, including data cleaning, deduplication, missing value processing, standardization or normalization, etc., to ensure the quality and accuracy of the data. After this series of processing, we successfully constructed a training sample set containing input features and output labels.

[0090] Next, based on these training samples, we selected a suitable deep learning algorithm for model training. During the training process, the model continuously learns how to predict customers’ preferences for marketing materials based on input features (such as the attributes of marketing materials and customer characteristics). After a certain number of training rounds, we obtained the initial prediction model.

[0091] In order to further improve the performance and accuracy of the model, the initial model was optimized and iterated. This included adjusting the model structure, changing the activation function, introducing regularization techniques to prevent overfitting, and using data augmentation techniques to increase the diversity of training samples. After multiple iterations of training, we finally obtained a prediction model with excellent performance.

[0092] Now, this carefully constructed prediction model is ready for practical application. When new marketing materials and customer data are input, the model can quickly analyze and determine the user's initial preference for marketing materials. This function provides strong support for the formulation of personalized marketing strategies, allowing us to more accurately push advertisements and product recommendations that meet user preferences, thereby improving marketing effectiveness and customer satisfaction.

[0093] In short, the process of building a predictive model is a systematic and complex task, from data collection and preprocessing to model training and optimization, and then to the final application. However, it is the rigor and scientific nature of this process that ensures that the model can accurately predict user preferences and provide strong support for the company's marketing activities.

[0094] Specifically, forecasting models analyze historical data and build mathematical models to predict future events. These models usually use statistical methods and machine learning algorithms to identify the relationship between variables and help decision makers make more informed choices by identifying patterns and relationships in the data. The following is a detailed explanation of how the forecasting model works:

[0095] First, data input and processing.

[0096] Collect marketing materials and customer data from various sources, such as advertising copy, pictures, videos, customer purchase records, browsing behavior, etc.

[0097] Clean the collected data to remove invalid, duplicate or missing data.

[0098] Standardize or normalize the data to ensure that different features have the same scale in the model.

[0099] Secondly, feature selection and extraction.

[0100] Select features from customer data that are relevant to the prediction target (such as customer preferences). Exclude features that are irrelevant or redundant to the prediction target.

[0101] Perform feature extraction on marketing materials, such as extracting keywords from advertising copy, colors and shapes from images, and converting these features into numerical forms that the model can understand.

[0102] Next, the model is trained and validated.

[0103] Choose an appropriate model, such as linear regression, decision tree, neural network, etc., based on the complexity of the problem and the characteristics of the data. Use historical data to train the model and adjust the model parameters to minimize the prediction error. During the training process, the model learns how to predict customer preferences based on the input features. Use the validation data set to validate the model and evaluate the model's performance. Adjust and optimize the model based on the validation results.

[0104] In addition, model prediction and application.

[0105] Input new marketing materials and customer data into the trained model. The model makes predictions based on the input data and outputs customer preferences for marketing content.

[0106] Explain and analyze the prediction results of the model. Develop personalized marketing strategies based on the prediction results, such as pushing advertisements that meet customer preferences and providing personalized product recommendations.

[0107] Finally, model update and optimization.

[0108] Continuously monitor the performance of the model to promptly identify and resolve potential problems; update and optimize the model based on new data or business needs; introduce new features or algorithms to improve the model's prediction accuracy and generalization capabilities.

[0109] Collect customer feedback data to further improve the model and enhance prediction results; continuously optimize the performance of the model through feedback loops to meet customer needs and expectations.

[0110] The model workflow is a process of continuous iteration and optimization, which requires continuous data collection, model training, result verification and model updating to adapt to the ever-changing market environment and customer needs.

[0111] Optionally, determine personalized marketing materials for customers based on emotion recognition results and initial preferences, including:

[0112] The emotion recognition results and initial preferences are input into the material model; the material model processes the emotion recognition results and initial preferences and outputs the customer's emotional tendency; based on the emotional tendency, the marketing materials for the customer are determined.

[0113] By using technologies such as deep learning algorithms and generative adversarial networks (GAN), models of bank marketing materials that meet the requirements are automatically generated based on the output results of large model technology and emotional computing technology.

[0114] Specifically, the emotion recognition results and initial preferences are input into the material model. This model has powerful processing capabilities and can comprehensively consider the emotion recognition results (i.e. the emotional response of customers when they come into contact with marketing materials) and initial preferences (the basic preferences of customers for marketing materials) to conduct in-depth analysis and calculation.

[0115] In the material model, complex algorithms and mechanisms will carefully process the input information, and finally output the customer's emotional tendency through a series of logical reasoning and pattern recognition. This result not only reflects the customer's emotional response to the marketing material, but also integrates their basic preferences, providing a key basis for the subsequent marketing strategy formulation.

[0116] With the guidance of emotional tendencies, marketing materials for customers can be determined more accurately. By matching customers' emotional needs and preferences, marketing content that better meets their expectations can be selected or designed, thereby more effectively attracting their attention and stimulating their desire to buy. This process not only improves the personalization of marketing activities, but also significantly enhances customer experience, laying a solid foundation for the company's marketing success.

[0117] Optional, the process of building a material model includes:

[0118] Obtain historical emotion recognition results and historical initial preferences; use historical emotion recognition results and historical initial preferences as target samples; use large model technology to train based on the target samples to obtain an initial material model; iteratively optimize the initial material model to obtain the final material model; use the material model to perform emotion recognition on customer data to obtain emotion recognition results.

[0119] The process of building a material model is similar to the process of building a prediction model mentioned above, and will not be described in detail here. It depends on the actual situation and is within the scope of protection of this application.

[0120] This material model uses big model technology to train a powerful and universal AI generation model. The model can learn and understand a large amount of data such as the bank's historical marketing materials, brand tone, market strategy, etc., so as to have the ability to automatically generate high-quality marketing materials. Through big model technology, marketing materials can be generated quickly while ensuring the diversity and creativity of the generated content.

[0121] Among them, adding an emotional computing module to the AI ​​generation model enables the model to identify and understand the emotional needs of customers. The emotional computing module extracts customers' emotional tendencies and demand characteristics by analyzing data such as customers' historical behaviors, preferences, and feedback. Then, these emotional characteristics are integrated into the generation process to make the generated marketing materials more targeted and appealing. Through emotional computing technology, personalized customization of marketing content can be achieved, improving customers' acceptance and response rate to marketing information.

[0122] That is, big model technology and emotional computing technology have realized the intelligent and personalized generation of bank marketing materials, effectively solved the problems existing in traditional marketing methods, and improved the bank's marketing effectiveness and market competitiveness.

[0123] The AI ​​generation model, as the core execution component of the present invention, undertakes the important task of converting the output results of large model technology and emotional computing technology into high-quality, personalized banking marketing materials. The model deeply integrates the essence of deep learning algorithms and generative adversarial networks (GANs), ensuring the diversity and innovation of marketing materials.

[0124] The AI ​​generation model, a key technical component, can intelligently generate the marketing materials needed by banks with its deep learning algorithm and GAN principles. It first draws inspiration from historical marketing materials and learns the inherent laws and patterns therein. Then, combined with the personalized needs and emotional tendencies of customers, the model can accurately generate text, image and video content that meets marketing goals.

[0125] Inside the model, the deep learning algorithm plays a crucial role. It is responsible for learning and extracting feature representations from historical marketing materials, which cover multiple dimensions such as text semantics, image visual features, and video dynamic information. These features are carefully stored in the model, providing a solid foundation for the subsequent generation of new materials.

[0126] GAN, with its unique adversarial training mechanism, further improves the quality of generated materials. GAN consists of two core parts: the generator and the discriminator. Through continuous confrontation and collaboration, they jointly drive the generator to generate more realistic and diverse marketing materials.

[0127] The workflow of AI generation model can be divided into the following three steps:

[0128] (1) Input processing.

[0129] In this stage, the model first receives the output results from the big model technology and emotional computing technology. These results contain key information such as customer demand characteristics, emotional tendencies, and marketing goals. The model then preprocesses this input information and converts it into numerical form so that the model can understand and process it.

[0130] Marketing goals can be set by the business department, and production materials can be achieved according to different goals, such as customer activation and customer acquisition. Customer activation means guiding users to participate or use, while customer acquisition means guiding new external customers to register and improve the certification level.

[0131] (2) Material generation.

[0132] In the material generation stage, the model will use deep learning algorithms to generate preliminary marketing materials based on the input information. These algorithms may include recurrent neural networks (RNN) for text generation, convolutional neural networks (CNN) for image generation, etc. Then, the preliminary generated materials will be sent to the generator part of GAN for further optimization and refinement. The generator will strive to generate more realistic and compliant marketing materials to deceive the discriminator. The discriminator will evaluate the authenticity of the generated materials and give feedback signals. The generator will continuously adjust itself based on these feedback signals to optimize the quality of the generated materials.

[0133] (3) Output display.

[0134] Finally, the generated marketing materials will be presented to customers or bank marketing staff. They can review and select these materials for subsequent publishing and promotion operations. If necessary, the generated materials can also be saved to the database or file system for subsequent use or analysis.

[0135] In AI generation models, the ingenious combination of deep learning algorithms and GAN provides powerful technical support for the generation of marketing materials. Recurrent neural networks (RNNs) are good at processing sequence data, such as text generation tasks; convolutional neural networks (CNNs) are good at processing image data, and can extract local features in images and gradually combine them into global feature representations; autoencoders can be used to learn the compressed representation and reconstruction of data, providing new ideas for generation tasks.

[0136] GAN, with its unique adversarial training mechanism, plays a crucial role in the model. Through the continuous confrontation and collaboration between the generator and the discriminator, GAN can drive the generator to generate more realistic marketing materials that are difficult to distinguish from real data. This feature makes the AI ​​generation model extremely practical and innovative in the generation of bank marketing materials.

[0137] In view of the personalized needs and preferences of customers, we have implemented in-depth customization and adjustment strategies for the generated marketing materials. With the help of the personalized customization module, we can accurately meet the unique requirements of customers, thereby significantly improving marketing effectiveness and customer satisfaction. The core functions of this strategy cover the following three aspects:

[0138] Accurately capture needs: comprehensively collect and analyze customers' personalized needs and preferences, including but not limited to key factors such as their preferred style, color matching, font selection, and overall layout.

[0139] Fine-tuning of materials: Based on a full understanding of customer needs, we make meticulous adjustments and optimizations to the generated marketing materials. This process is designed to ensure that the materials accurately reflect customer expectations and requirements, thereby enhancing the attractiveness and effectiveness of the materials.

[0140] Real-time preview and feedback: Provide customers with a convenient preview function so that they can instantly view the effects of customized marketing materials. At the same time, we actively collect customer feedback and suggestions for subsequent continuous improvement and optimization.

[0141] In summary, the technical solution of the present invention achieves efficient, personalized and emotionally rich automatic generation of bank marketing materials by deeply integrating large model technology and emotional computing technology. This solution not only significantly improves generation efficiency and quality, but also effectively reduces labor costs and perfectly meets customers' high-quality demand for customized content.

[0142] The above prediction model and material model can both use big model technology, that is, the advantages of deep neural network (DNN) and convolutional neural network (CNN) are integrated into the big model technology. DNN is good at processing text and structured data, and can learn complex data abstract representations; while CNN is proficient in processing image and video data, and is good at capturing local and global features.

[0143] DNN: Based on the expansion of multi-layer perceptron (MLP), it deeply mines data representation through multiple hidden layers and neurons. In the text processing of marketing materials, DNN effectively extracts keywords, phrases and sentence-level semantic information.

[0144] CNN: Combines convolutional layers, pooling layers, and fully connected layers to capture the spatial hierarchy and features of images. When processing marketing images and videos, CNN extracts visual features such as color, texture, shape, as well as motion and scene information.

[0145] Model training process: Data preparation: Collect historical marketing materials and customer data, and divide them into training, validation, and test sets. Preprocessing: Clean text (remove stop words, punctuation, etc.), scale and normalize images to improve training efficiency and performance. Feature engineering: For text and image data, extract features such as keyword frequency, TF-IDF weight, edge detection, color histogram, etc. Model configuration: Initialize DNN and CNN, set network structure and hyperparameters, which affect training results and generation quality. Training: Iterate the model, use backpropagation to update parameters, and minimize the loss function. Regularly evaluate the validation set performance and tune the model. Evaluation and optimization: Final test set evaluation, optimize the model based on indicators such as accuracy, recall, and F1 score.

[0146] By integrating DNN and CNN and combining them with a carefully designed training process, big model technology can effectively process and analyze historical data, provide strong support for the generation of marketing materials, improve efficiency and quality, and meet the preferences and needs of target customers.

[0147] In this embodiment, with the help of the dual assistance of large models and emotional computing technology, the AI ​​generation model was successfully trained to achieve the fully automated generation of bank marketing materials, while ensuring the diversity and high personalization of the materials. In this process, deep neural networks and convolutional neural network algorithms were used to build a powerful large model framework, and emotional computing technology was cleverly integrated. Through deep learning of the inherent laws and patterns of historical marketing materials, we accurately analyzed the emotional tendencies and demand preferences of customers. Subsequently, an AI generation model was carefully created using advanced deep learning algorithms and combined with generative adversarial network technology. The model uses the output results of large models and emotional computing technology as core inputs for in-depth analysis and precise judgment, thereby automatically generating high-quality, diverse and highly personalized marketing materials, covering a variety of content forms such as text, images and videos. It is worth mentioning that this application also has excellent feedback learning capabilities. It can continuously and automatically optimize and adjust marketing materials based on the feedback information provided by users to ensure that the quality of materials can be continuously improved to better meet customer needs.

[0148] The following are examples of the advantages of this application:

[0149] Significantly improve efficiency: The present invention uses AI generation models to achieve the automatic generation of marketing materials, a process that requires almost no human intervention or only a very small amount of manual review. This not only greatly accelerates the production process of marketing content, but also significantly improves work efficiency. Deep learning algorithms can mine and extract key features from massive amounts of historical data, thereby effectively avoiding the tedious manual design and material screening work of marketers.

[0150] Personalization and precision delivery: The application of emotional computing technology enables the present invention to accurately capture and analyze customers' emotional tendencies and demand characteristics. Based on this, the generated marketing materials can more closely fit the personalized preferences of target customers. This highly personalized marketing content is more likely to touch the hearts of customers, stimulate their interest and resonance, and thus improve the conversion rate and user participation of marketing activities.

[0151] Diverse innovation, eye-catching: AI generative models have powerful generation capabilities and can easily produce various types of marketing materials including text, images, videos, etc., perfectly adapting to the needs of various marketing channels and scenarios. At the same time, with the support of deep learning technologies such as generative adversarial networks (GAN), the generated materials show unprecedented innovation and diversity in style, design and presentation, which can quickly attract customers' attention.

[0152] Cost optimization and efficiency improvement: The automated material generation process significantly reduces the reliance on professional designers and marketers, thereby effectively controlling labor costs. In addition, through the implementation of precision marketing strategies, the placement of ineffective advertisements and waste of resources are reduced, further improving the efficiency of marketing budget utilization.

[0153] Continuous evolution and wide applicability: With the continuous influx of new data and continuous training of models, the performance of AI-generated models will continue to rise, thereby better adapting to changes in the market environment and customer needs. It is worth mentioning that although the implementation cases of this invention are mainly concentrated in the field of banking marketing, its core technology has extremely high flexibility and scalability, and can be easily applied to marketing activities in other industries, showing wide applicability and huge market potential.

[0154] It should be noted that in the current marketing practice of the banking industry, in order to address the challenges of low efficiency and lack of personalization in the production of marketing materials, the industry widely adopts manual design approaches and template-based automatic creation. However, both approaches are inefficient, lack personalization, and lack emotional connection.

[0155] In this embodiment, a powerful and versatile AI generation engine (material model) is constructed using large model technology to achieve the automated and rapid output of marketing materials. The engine can efficiently generate high-quality marketing content by deeply learning a large amount of historical materials and customer data; and an emotional computing module is embedded in the AI ​​generation engine to accurately capture and understand customers' emotional tendencies and preferences; by analyzing customer historical behavior, feedback and other data, personalized features are extracted and integrated into the generation process to achieve customized creation of marketing materials; at the same time, the emotional computing module can not only identify customer needs, but also simulate human emotions to make marketing materials more contagious and attractive. By building an emotional bridge with customers, the acceptance and response enthusiasm of marketing information can be improved.

[0156] Optionally, after determining the customer's marketing materials based on the emotion recognition results and initial preferences, it also includes:

[0157] Preview the client's marketing materials.

[0158] Before finalizing and applying marketing materials to clients, a preview session was set up; this session was crucial because it provided clients with an opportunity to view and evaluate the marketing materials in advance.

[0159] During the preview process, customers can carefully browse marketing materials, including advertising copy, pictures, videos and other forms. These materials are not only carefully planned and produced, but also fully consider the emotional needs and preferences of customers, aiming to bring them the best visual and emotional experience.

[0160] Through preview, customers can have a comprehensive understanding of the content, style, and presentation of marketing materials. They can provide valuable comments and suggestions on the materials based on their preferences and expectations. These feedbacks are crucial for further optimizing and adjusting marketing materials, as they can more accurately grasp customer needs and expectations, thereby producing marketing content that better meets customer needs.

[0161] In short, previewing the client's marketing materials is an indispensable step. It not only improves customer engagement and satisfaction, but also provides us with valuable feedback and improvement opportunities, laying a solid foundation for the ultimate marketing success.

[0162] Optionally, after determining the customer's marketing materials based on the emotion recognition results and initial preferences, it also includes:

[0163] Get feedback from customers.

[0164] In order to further improve and optimize the marketing strategy, actively seek and obtain customer feedback. This step is a crucial part of the entire marketing process because it is directly related to the degree of understanding of customer needs and the direction of improvement of subsequent services.

[0165] Collect customer feedback through a variety of channels, including but not limited to online surveys, customer service hotlines, social media interactions, and face-to-face customer interviews. These channels provide a variety of data sources to more fully understand customers' true feelings about marketing materials, products and services, and the overall shopping experience.

[0166] When collecting feedback information, we focus on the pertinence and depth of the questions, aiming to tap into the real thoughts and potential needs of customers. At the same time, we also encourage customers to put forward specific suggestions for improvement, so as to more accurately identify the problems and develop practical improvement measures.

[0167] After obtaining customer feedback, we will systematically organize and analyze it. Through data analysis and mining, we can discover common problems and personalized needs in customer feedback, and then optimize and adjust marketing strategies, product design, customer service and other aspects accordingly.

[0168] In short, obtaining customer feedback is a key step in continuously improving marketing effectiveness and service quality. We will continue to pay attention to customer feedback and changes in demand, and continuously optimize our products and services to provide a more personalized, efficient and satisfactory marketing experience.

[0169] It should be noted that in the field of automatic generation of advertising and marketing materials, one technical solution is an automated advertising generation system based on deep learning. With the help of deep learning models and training with a large amount of historical data, this system can automatically produce advertising materials that meet basic requirements, thereby achieving initial improvements in efficiency and quality. However, this technical solution still has several significant limitations, and it is difficult to meet the urgent needs of the banking marketing field for highly personalized and emotional marketing materials; specifically, lack of personalization: the automated advertising generation system based on deep learning mainly relies on general patterns extracted from historical data to generate advertisements. Although this method can produce advertising materials that meet certain standards, because it overemphasizes universality, the generated advertising content often lacks personalized elements for specific customer groups, making it difficult to stand out in the fiercely competitive market. Lack of emotional resonance: In the process of advertising generation, existing technical solutions often ignore the emotional needs of consumers. As a dual carrier of information and emotion, advertising that lacks emotional resonance is difficult to touch the hearts of consumers, thereby affecting the communication effect and conversion efficiency of advertising. Limited model adaptability: The performance of deep learning models is highly dependent on the quality and quantity of training data. When faced with new advertising types, styles or customer groups, existing technical solutions require re-collecting large amounts of data and retraining models, which is not only time-consuming and labor-intensive, but also limits the flexibility and adaptability of the system.

[0170] In view of the above shortcomings, this application proposes a new automatic generation solution for bank marketing materials by combining large model technology and emotional computing technology, aiming to achieve more efficient and personalized marketing material generation. Enhanced personalization: This application uses large model technology to enable the AI ​​generation model to learn and understand more historical data and market information, thereby generating more accurate and personalized advertising content. At the same time, the introduction of emotional computing technology enables the system to identify and understand consumers' emotional needs and preferences, integrate this information into the advertising generation process, and further improve the personalization level of advertising. Enhance emotional resonance: In the process of advertising generation, this application focuses on emotional communication with consumers. Through emotional computing technology, the system can analyze consumers' historical behavior, feedback and other data, extract consumers' emotional tendencies and demand characteristics, and integrate them into advertising content. In this way, advertising can not only convey information, but also establish emotional connections with consumers, enhancing the appeal and attractiveness of advertising. Improve model adaptability: The large model technology used in this application has stronger learning and adaptability. By training a powerful and general AI generation model, the system can adapt to various advertising types, styles and customer groups. When faced with new market or advertising demands, the system only needs to perform fine-tuning or incremental learning without retraining the entire model, which greatly improves the adaptability and flexibility of the model.

[0171] To sum up, this application successfully makes up for the shortcomings of existing technical solutions in personalization, emotional resonance and model adaptability by combining large model technology and emotional computing technology, and provides a more efficient and personalized marketing material automatic generation solution for the banking marketing field.

[0172] like Figure 2 As shown, the following is an example of a method for generating marketing materials:

[0173] 1. Data collection.

[0174] Collect historical marketing and customer data.

[0175] 2. Data preprocessing.

[0176] The collected data is cleaned, labeled, and feature extracted.

[0177] 3. Large model training.

[0178] Based on the preprocessed data, the model is trained using a deep learning algorithm to obtain a prediction model.

[0179] 4. Emotion recognition.

[0180] Identify customer data to obtain emotional information.

[0181] 5. Emotional understanding.

[0182] Deeply understand and analyze emotional tendencies.

[0183] 6. AI generates models.

[0184] Combine big model technology and emotional computing to generate marketing materials.

[0185] 7. Personalized customization.

[0186] Customize materials based on customer preferences.

[0187] 8. Preview and feedback.

[0188] Support customers to preview and provide feedback.

[0189] Another embodiment of the present application provides a marketing material generation system.

[0190] See also Figure 3 , the marketing material generation system includes:

[0191] The acquisition module 101 is used to acquire marketing materials and customer data.

[0192] The preference module 102 is used to determine the customer's initial preference for the marketing material based on the marketing material and the customer data.

[0193] The emotion module 103 is used to perform emotion recognition on customer data and obtain emotion recognition results;

[0194] The material module 104 is used to determine the marketing material of the customer based on the emotion recognition result and the initial preference; different customers have different corresponding marketing materials.

[0195] For details on the specific working process and principles of each of the above modules, please refer to the marketing material generation method provided in the above embodiment, which will not be described here one by one. It will depend on the actual situation and is within the scope of protection of this application.

[0196] In this embodiment, the acquisition module 101 acquires marketing materials and customer data; the preference module 102 determines the customer's initial preference for the marketing materials based on the marketing materials and customer data; the emotion module 103 performs emotion recognition on the customer data to obtain the emotion recognition result; the material model 104 determines the customer's marketing materials based on the emotion recognition result and the initial preference; different customers have different corresponding marketing materials; that is, through multiple links such as data collection, data analysis, emotion recognition and personalized marketing, we can better understand the needs and preferences of customers, and formulate more effective marketing strategies to attract and retain customers, improve the efficiency of customizing marketing materials for users, and can face all customers and focus on emotional connection with customers. The marketing materials are contagious and stimulate the interest and resonance of customers.

[0197] Another embodiment of the present application provides a storage medium having a computer program stored thereon, wherein when the computer program is executed by a processor, the method for generating marketing materials as described in any one of the above embodiments is implemented.

[0198] In the context of the present disclosure, a machine-readable medium may be a tangible medium that may contain or store a program for use by or in conjunction with an instruction execution system, device, or equipment. A machine-readable medium may be a machine-readable signal medium or a machine-readable storage medium. A machine-readable medium may include, but is not limited to, an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, device, or equipment, or any suitable combination of the foregoing. A more specific example of a machine-readable storage medium may include an electrical connection based on one or more lines, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.

[0199] It should be noted that the computer-readable medium disclosed above may be a computer-readable signal medium or a computer-readable storage medium or any combination of the above two. The computer-readable storage medium may be, for example, but not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, device or device, or any combination of the above. More specific examples of computer-readable storage media may include, but are not limited to: an electrical connection with one or more wires, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the above. In the present disclosure, a computer-readable storage medium may be any tangible medium containing or storing a program that may be used by or in combination with an instruction execution system, device or device. In the present disclosure, a computer-readable signal medium may include a data signal propagated in a baseband or as part of a carrier wave, in which a computer-readable program code is carried. This propagated data signal may take a variety of forms, including but not limited to an electromagnetic signal, an optical signal, or any suitable combination of the above. The computer readable signal medium may also be any computer readable medium other than a computer readable storage medium, which may send, propagate or transmit a program for use by or in conjunction with an instruction execution system, apparatus or device. The program code contained on the computer readable medium may be transmitted using any suitable medium, including but not limited to: wires, optical cables, RF (radio frequency), etc., or any suitable combination of the above.

[0200] The computer-readable medium may be included in the electronic device, or may exist independently without being incorporated into the electronic device.

[0201] Another embodiment of the present invention provides an electronic device, such as Figure 4 As shown, including:

[0202] One or more processors 201.

[0203] The storage device 202 stores one or more programs.

[0204] When one or more programs are executed by one or more processors 201 , the one or more processors 201 implement the marketing material generation method as described in any one of the above embodiments.

[0205] In particular, according to an embodiment of the present disclosure, the process described above with reference to the flowchart can be implemented as a computer software program. For example, an embodiment of the present disclosure includes a computer program product, which includes a computer program carried on a non-transitory computer-readable medium, and the computer program includes a program code for executing the method shown in the flowchart.

[0206] Although the subject matter has been described in language specific to structural features and / or methodological logical actions, it should be understood that the subject matter defined in the appended claims is not necessarily limited to the specific features or actions described above. On the contrary, the specific features and actions described above are merely example forms of implementing the claims.

[0207] Although several specific implementation details are included in the above discussion, these should not be interpreted as limiting the scope of the present disclosure. Certain features described in the context of a separate embodiment can also be implemented in a single embodiment in combination. On the contrary, the various features described in the context of a single embodiment can also be implemented in multiple embodiments individually or in any suitable sub-combination.

[0208] The above description is only a preferred embodiment of the present disclosure and an explanation of the technical principles used. Those skilled in the art should understand that the scope of disclosure involved in the present disclosure is not limited to the technical solutions formed by a specific combination of the above technical features, but should also cover other technical solutions formed by any combination of the above technical features or their equivalent features without departing from the above disclosed concept. For example, the above features are replaced with the technical features with similar functions disclosed in the present disclosure (but not limited to) by each other to form a technical solution.

[0209] The features recorded in the various embodiments in this specification can be replaced or combined with each other, and the same and similar parts between the various embodiments can refer to each other, and each embodiment focuses on the differences from other embodiments. In particular, for the system or system embodiment, since it is basically similar to the method embodiment, the description is relatively simple, and the relevant parts can refer to the partial description of the method embodiment. The system and system embodiments described above are only schematic, wherein the units described as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units, that is, they may be located in one place, or they may be distributed on multiple network units. Some or all of the modules can be selected according to actual needs to achieve the purpose of the scheme of this embodiment. Ordinary technicians in this field can understand and implement it without paying creative work.

[0210] Professionals may further appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, computer software, or a combination of the two. In order to clearly illustrate the interchangeability of hardware and software, the composition and steps of each example have been generally described in the above description according to function. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professionals and technicians may use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of the present invention.

[0211] The above description of the disclosed embodiments enables one skilled in the art to implement or use the present invention. Various modifications to these embodiments will be apparent to one skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present invention. Therefore, the present invention will not be limited to the embodiments shown herein, but rather to the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A method for generating marketing materials, characterized in that: include: Obtain marketing materials and customer data; Determining the customer's initial preference for the marketing material based on the marketing material and the customer data; Performing emotion recognition on the customer data to obtain an emotion recognition result; Determining the marketing material for the customer based on the emotion recognition result and the initial preference; The marketing materials corresponding to different customers are different.

2. The method for generating marketing materials according to claim 1, characterized in that: Determining the customer's initial preference for the marketing material based on the marketing material and the customer data includes: The marketing material and the customer data are input into a prediction model, and the prediction model processes the marketing material and the customer data and outputs the initial preference.

3. The method for generating marketing materials according to claim 2, characterized in that: The process of constructing the prediction model includes: Obtain historical marketing materials and historical customer data for preprocessing to obtain training samples; Based on the training samples, a deep learning algorithm is used for training to obtain an initial prediction model; The initial prediction model is optimized and iterated to obtain a final prediction model; the prediction model is used for marketing materials and customer data to determine the user's initial preference for the marketing materials.

4. The method for generating marketing materials according to claim 1, characterized in that: Determining personalized marketing materials for customers based on the emotion recognition results and the initial preferences includes: inputting the emotion recognition result and the initial preference into a material model; The material model processes the emotion recognition result and the initial preference, and outputs the emotional tendency of the customer; Based on the emotional tendency, marketing materials targeting the customer are determined.

5. The method for generating marketing materials according to claim 4, characterized in that: The process of constructing the material model includes: Obtain historical emotion recognition results and historical initial preferences; Using the historical emotion recognition results and historical initial preferences as target samples; Using large model technology to train the target sample to obtain an initial material model; The initial material model is iteratively optimized to obtain a final material model; the material model performs emotion recognition on the customer data to obtain an emotion recognition result.

6. The method for generating marketing materials according to claim 1, characterized in that: After determining the marketing material of the customer based on the emotion recognition result and the initial preference, the method further includes: A preview of the client's marketing materials is provided.

7. The method for generating marketing materials according to claim 1, characterized in that: After determining the marketing material of the customer based on the emotion recognition result and the initial preference, the method further includes: Get feedback from customers.

8. A marketing material generation system, characterized in that: include: Acquisition module, used to obtain marketing materials and customer data; A preference module, used to determine the customer's initial preference for the marketing material based on the marketing material and the customer data; An emotion module, used to perform emotion recognition on the customer data to obtain an emotion recognition result; The material module is used to determine the customer's marketing material based on the emotion recognition result and the initial preference; different customers have different corresponding marketing materials.

9. An electronic device, characterized in that: include: one or more processors; a storage device having one or more programs stored thereon; When one or more programs are executed by one or more processors, the one or more processors implement the marketing material generating method according to any one of claims 1 to 7.

10. A storage medium, characterized in that: A computer program is stored thereon, wherein when the computer program is executed by a processor, the marketing material generating method according to any one of claims 1 to 7 is implemented.

Images