A marketing-based image big data analysis method

Abstract

The present invention discloses a method for analyzing image big data based on marketing, which relates to the field of marketing technology and includes the following steps: extracting image features, constructing a dynamic analysis model, constructing a sentiment analysis model, constructing a model for the correlation between image features and consumer behavior, evaluating the impact of different images on marketing effects, constructing a marketing image recommendation model, and providing optimization solutions for multi-platform marketing strategies. This method for analyzing image big data based on marketing can accurately predict and recommend marketing images that are most suitable for target consumers through the deep integration of multi-dimensional data such as comprehensive image features, consumer behavior, and emotional reactions. By establishing a correlation model between image features and consumer behavior and a sentiment analysis model, the accuracy and market adaptability of image recommendations are greatly improved. Compared with traditional methods, the present invention can more accurately capture consumers' emotional needs and behavior patterns, and optimize marketing images in real time.

Description

Technical Field

[0001] The present invention relates to the field of marketing technology, and in particular to a marketing-based image big data analysis method. Background Art

[0002] With the rapid development of the internet and mobile internet, the importance of images in marketing has become increasingly prominent. Advertisers, brands, and e-commerce platforms are increasingly relying on visual content to attract consumers. Especially on social media and e-commerce platforms, images, as a crucial vehicle for advertising, directly influence consumer purchasing decisions and brand perception. However, traditional marketing image analysis methods often rely on simple image feature extraction and consumer behavior analysis, lacking a comprehensive understanding of the deeper correlation between image features and consumer behavior, and are inadequate in predicting consumer emotional responses.

[0003] Existing technologies typically focus solely on image features (such as color and shape) or consumer behavior (such as click-through rate and viewing time). This single-dimensional analysis limits personalized marketing image recommendations. Traditional methods fail to effectively integrate the multidimensional characteristics of image content with the dynamic changes in consumer behavior, and sentiment analysis often remains limited to a single level of feedback. Many existing technologies also fail to fully utilize the comprehensive score of image marketing effectiveness and lack precise control over scene and emotional memory effects. As a result, recommended images fail to fully align with consumer emotional preferences and market demand. Summary of the Invention

[0004] In view of the shortcomings of the existing technology, the present invention provides an image big data analysis method based on marketing to solve the problems raised in the above background technology.

[0005] To achieve the above object, the present invention provides the following technical solutions:

[0006] In a first aspect, an embodiment of the present invention provides a marketing-based image big data analysis method, comprising the following steps:

[0007] S1. Extract image features based on marketing images;

[0008] S2. Based on the extracted image features, a dynamic analysis model is constructed to conduct marketing scenario analysis;

[0009] S3. After conducting marketing scenario analysis, construct a sentiment analysis model to evaluate consumers’ emotional responses to different types of marketing images;

[0010] S4. Based on the results outputted from steps S2 and S3, a correlation model between image features and consumer behavior is constructed to further analyze the correlation between image features and consumer behavior;

[0011] S5. Based on the correlation analysis results, evaluate the impact of different images on marketing effectiveness and adjust the image elements in marketing activities in real time;

[0012] S6. Based on the correlation analysis results, a marketing image recommendation model is constructed to provide personalized marketing image recommendations for each consumer group;

[0013] S7. Evaluate the effectiveness of image marketing on different platforms through cross-platform analysis and provide optimization solutions for multi-platform marketing strategies.

[0014] To further optimize this technical solution, in step S1, image feature extraction includes:

[0015] Extract visual features from a large number of marketing-related images, including color distribution, texture features, shape information, brand logos, advertising elements, and consumer behavior;

[0016] Through deep convolutional neural network (CNN), images are converted into structured data features.

[0017] To further optimize this technical solution, in step S2, the dynamic analysis model combines the changes in time series, the dynamic evolution of image features, and the feedback mechanism of consumer behavior;

[0018] The dynamic analysis model sets the following parameters:

[0019] A comprehensive score representing the consumer behavior status at time t, including indicators such as click-through rate, conversion rate, and purchase intention;

[0020] It is the image features extracted from the image at time t, including color distribution, texture features, shape information, brand logo, and advertising elements;

[0021] is the change in image features, that is, the difference in image features between time t and time t-1;

[0022] It is the influencing factor of the marketing environment, indicating the interference or support of the external market environment on marketing activities at a certain moment, including holiday effects and competitor advertising;

[0023] It is a hyperparameter of the model, which represents the weight of image features, environmental factors, and historical behaviors in dynamic adjustment.

[0024] To further optimize this technical solution, the dynamic analysis model is as follows:

[0025] ;

[0026] in,

[0027] : Indicates the impact of current image features on consumer behavior at time t, It is the mapping function of image features to consumer behavior;

[0028] : Indicates the impact of changes in image features on consumer behavior, Reflects the contribution of feature changes to behavioral changes;

[0029] : Indicates the impact of external market environment factors on consumer behavior;

[0030] : Indicates the impact of consumer behavior at the previous moment, that is, the continued impact of consumers' historical behavior on current decisions.

[0031] To further optimize this technical solution, in step S3, the sentiment analysis model is as follows:

[0032] ;

[0033] in,

[0034] represents the consumer's emotional response to the image at time t, and the emotional categories include like, dislike, and excitement;

[0035] is the image feature at time t, including color distribution, texture features, shape information, and brand logo;

[0036] is the consumer behavior characteristic at time t;

[0037] is the consumer's emotional preference at time t;

[0038] It is the comprehensive score of consumer behavior status output by the dynamic analysis model;

[0039] are the weight parameters of this model, corresponding to image features, consumer behavior, emotional preferences, and comprehensive scores. and affective memory effects;

[0040] : Mapping function of the impact of image features on consumer emotional responses;

[0041] : Mapping function of the impact of consumer behavior data on emotional response;

[0042] : Mapping function of the impact of consumer emotional preferences on emotional responses;

[0043] : Comprehensive score Affect mapping function for emotional responses;

[0044] : Emotional memory effect, the continuity of consumers' previous emotional reactions on current evaluations.

[0045] To further optimize this technical solution, in step S4, the following parameters are set in the image feature and consumer behavior correlation model:

[0046] Represents the correlation score between image features and consumer behavior, ranging from [0,1];

[0047] is the image feature vector, including characteristics;

[0048] is the consumer behavior feature vector, including characteristics;

[0049] is the comprehensive score output by the dynamic analysis model;

[0050] is the weight of the image feature, which indicates the influence of each feature on consumer behavior;

[0051] is the weight of the consumer behavior, indicating the contribution of the behavior to the relevance;

[0052] The correlation model between image features and consumer behavior is as follows:

[0053] ;

[0054] in,

[0055] : Image features and behavioral characteristics The correlation function represents the statistical relationship between the two in historical data.

[0056] Further optimizing the technical solution, in step S5, the effects of different image designs are evaluated by continuously monitoring real-time data of marketing activities;

[0057] Feed consumer behavior status into dynamic analysis models to adjust image elements in marketing activities in real time.

[0058] Further optimizing this technical solution, in step S6, the marketing image recommendation model integrates image features , consumer behavior characteristics , emotional response , comprehensive score of consumer behavior status , the correlation score between image features and consumer behavior Build and recommend attractive and commercially valuable marketing images based on consumer preferences and situational needs;

[0059] The marketing image recommendation model sets the following parameters:

[0060] To consumers Image recommendation score;

[0061] Indicates the The feature vector of each candidate image;

[0062] Indicates consumers The weight vector is based on consumer behavior characteristics , emotional response and a comprehensive score of consumer behavior status ;

[0063] Indicates the The context-optimized weight of each image is used to dynamically adjust the recommendation priority.

[0064] To further optimize this technical solution, the marketing image recommendation model is as follows:

[0065] ;

[0066] in,

[0067] : The weighted inner product of consumer weight and image features, quantifying the consumer's preference for a specific image.

[0068] To further optimize this technical solution, in step S7, a comparative analysis is performed on the marketing effects of images published on different platforms, including social media, e-commerce platforms, and search engine advertisements;

[0069] Use A / B testing and cross-platform experiments to quantify the effectiveness of image marketing strategies on different platforms and evaluate the differences between platforms.

[0070] In a second aspect, an embodiment of the present invention provides a computer device comprising a memory and a processor, wherein the memory stores a computer program, wherein: when the computer program instructions are executed by the processor, the steps of a marketing-based image big data analysis method as described in the first aspect of the present invention are implemented.

[0071] In a third aspect, an embodiment of the present invention provides a computer-readable storage medium having a computer program stored thereon, wherein: when the computer program instructions are executed by a processor, the steps of a marketing-based image big data analysis method as described in the first aspect of the present invention are implemented.

[0072] Compared with the existing technology, the present invention provides a marketing-based image big data analysis method with the following beneficial effects:

[0073] This marketing-based image big data analysis method, through the deep integration of multi-dimensional data such as image features, consumer behavior, and emotional responses, can accurately predict and recommend marketing images that are most suitable for target consumers. By establishing a correlation model between image features and consumer behavior, as well as a sentiment analysis model, the accuracy and market adaptability of image recommendations are greatly improved. Compared with traditional methods, this invention can more accurately capture consumers' emotional needs and behavioral patterns, optimize marketing images in real time, improve advertising conversion rates and brand influence, and realize personalized and intelligent marketing strategies, ultimately achieving higher marketing benefits. BRIEF DESCRIPTION OF THE DRAWINGS

[0074] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0075] Figure 1 This is a flowchart of a marketing-based image big data analysis method proposed by the present invention;

[0076] Figure 2 This is a flow chart of a dynamic analysis model in a marketing-based image big data analysis method proposed by the present invention;

[0077] Figure 3 This is a flow chart of the sentiment analysis model in the marketing-based image big data analysis method proposed in the present invention. DETAILED DESCRIPTION

[0078] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, the specific embodiments of the present invention are described in detail below with reference to the accompanying drawings.

[0079] In the following description, many specific details are set forth to facilitate a full understanding of the present invention. However, the present invention may also be implemented in other ways different from those described herein. Those skilled in the art may make similar generalizations without violating the connotation of the present invention. Therefore, the present invention is not limited to the specific embodiments disclosed below.

[0080] Secondly, the term "one embodiment" or "embodiment" herein refers to a specific feature, structure, or characteristic that may be included in at least one implementation of the present invention. The phrase "in one embodiment" appearing in various places throughout this specification does not necessarily refer to the same embodiment, nor does it constitute a separate or selective embodiment that is mutually exclusive with other embodiments. Example 1

[0081] Reference Figures 1 to 3 , which is the first embodiment of the present invention, provides an image big data analysis method based on marketing, including the following steps:

[0082] S1. Extract image features based on marketing images.

[0083] In this embodiment, this step is the basis for the entire analysis, and subsequent image data processing and analysis rely on the key image features extracted in this step. Image feature extraction includes:

[0084] Extract visual features from a large number of marketing-related images, including color distribution, texture features, shape information, brand logos, advertising elements, and consumer behavior;

[0085] Using a deep convolutional neural network (CNN), images are converted into structured data features. This not only focuses on the image's pixel content but also considers contextual information, including elements such as the brand, product category, and consumer visual behavior. These features provide the necessary information foundation for subsequent data analysis and mining, ensuring the accurate reflection of the value of images in marketing.

[0086] S2. Based on the extracted image features, a dynamic analysis model is constructed to conduct marketing scenario analysis.

[0087] In this example, we combine extracted image features with marketing scenarios for dynamic analysis. "Marketing scenarios" here refer to actual market application environments such as advertising presentations, store displays, and online marketing campaigns. We analyze the dynamic changes in image features across different marketing campaigns to explore the short-term and long-term impact of images on consumer behavior. For example, color changes in advertising material, the frequency of brand logo exposure, and the layout of ad placements can all significantly influence consumer purchasing decisions.

[0088] The dynamic analysis model combines the changes in time series, the dynamic evolution of image features, and the feedback mechanism of consumer behavior;

[0089] The dynamic analysis model sets the following parameters:

[0090] A comprehensive score representing the consumer behavior status at time t, including indicators such as click-through rate, conversion rate, and purchase intention;

[0091] It is the image features extracted from the image at time t, including color distribution, texture features, shape information, brand logo, and advertising elements;

[0092] is the change in image features, that is, the difference in image features between time t and time t-1;

[0093] It is the influencing factor of the marketing environment, indicating the interference or support of the external market environment on marketing activities at a certain moment, including holiday effects and competitor advertising;

[0094] It is a hyperparameter of the model, which represents the weight of image features, environmental factors, and historical behaviors in dynamic adjustment, improving the prediction accuracy and adaptability of the model.

[0095] Furthermore, the dynamic analysis model is as follows:

[0096] ;

[0097] in,

[0098] : Indicates the impact of current image features on consumer behavior at time t, It is the mapping function of image features to consumer behavior, which can be fitted by machine learning models (such as neural networks, support vector machines, etc.).

[0099] : Indicates the impact of changes in image features on consumer behavior. Changes in image features may have an incentive or inhibitory effect on behavior, so they need to be modeled separately. Reflects the contribution of feature changes to behavioral changes.

[0100] : Represents the impact of external market factors on consumer behavior. This term can capture the time-varying effects of the market environment. For example, during a promotion, the visual effects of an ad may lead to a higher purchase conversion rate.

[0101] : represents the influence of consumer behavior at the previous moment, that is, the ongoing impact of a consumer's past behavior on their current decision. This term allows us to incorporate consumer behavioral inertia, reflecting the temporal transitivity of consumers' purchase intentions and decision-making patterns.

[0102] When used, the model includes:

[0103] The impact of dynamic changes in image features: At each time point t, we use the image features and feature variation To assess the immediate impact of an image on consumer behavior. For example, if the brand logo appears more frequently in an image, or if a change in the color tone of the ad produces a positive response from consumers, then The value of will increase, thereby increasing conversion rate.

[0104] Consideration of market environment factors: By introducing environmental factors We can embed external market variables, such as seasonal promotions and economic fluctuations, into the model to simulate the impact of different external conditions on marketing effectiveness. For example, during holiday promotions, advertising images may need to be more eye-catching to enhance consumer engagement.

[0105] Adaptability of historical behavior: The term captures the ongoing impact of a consumer's past behavior on current marketing campaigns. This is particularly important for optimizing long-term advertising campaigns. For example, if a consumer has shown a high click-through rate for images of a certain color in past ads, the model can use this historical behavior to predict their response to future ads.

[0106] Real-time adjustment and optimization: Through the feedback mechanism, we can continuously update and This allows the model to automatically adjust image design based on real-time market feedback in real-time. For example, a change in an ad element might initially spark consumer interest, but its effectiveness may decline over time. The model can capture this dynamic change and adjust image features to improve marketing effectiveness.

[0107] S3. After conducting marketing scenario analysis, construct a sentiment analysis model to evaluate consumers’ emotional responses to different types of marketing images.

[0108] In this example, consumers' emotions, such as likes, dislikes, and excitement, are analyzed when viewing different images. This analysis goes beyond superficial sentiment and incorporates consumer purchasing history and behavioral preferences to form a personalized sentiment analysis model. This step provides a more detailed consumer profile for subsequent marketing strategies, enabling more targeted marketing campaigns.

[0109] The sentiment analysis model looks like this:

[0110] ;

[0111] in,

[0112] represents the consumer's emotional response to the image at time t, and the emotional categories include like, dislike, and excitement;

[0113] is the image feature at time t, including color distribution, texture features, shape information, and brand logo;

[0114] is the consumer behavior characteristic at time t;

[0115] is the consumer's emotional preference at time t;

[0116] It is the comprehensive score of consumer behavior status output by the dynamic analysis model;

[0117] are the weight parameters of this model, corresponding to image features, consumer behavior, emotional preferences, and comprehensive scores. and affective memory effects;

[0118] : A mapping function of the impact of image features on consumer emotional responses, using a deep learning network (such as CNN) to train the relationship mapping between image features and emotional responses;

[0119] : A mapping function of the impact of consumer behavior data on emotional responses, based on regression analysis of behavioral data, establishes a statistical relationship between behavioral characteristics and emotional responses;

[0120] : A mapping function of the impact of consumer emotional preferences on emotional responses, using a personalized preference recommendation system (such as collaborative filtering) to calculate the impact weight of emotional preferences;

[0121] : Comprehensive score The mapping function of the impact of emotional response is used to fit the impact of the comprehensive score on emotional response through supervised learning of historical data of marketing effects. This function reflects how marketing effects directly or indirectly guide consumers' emotional responses. For example, high A value of may indicate a more attractive image design and placement strategy;

[0122] : Emotional memory effect, the continuity of consumers' previous emotional reactions on current evaluations.

[0123] When used, the model includes:

[0124] Data input and integration:

[0125] enter and emotional memory data ;

[0126] Ensuring data integrity, especially The exact calculation result of is the output of the previous model.

[0127] Weight parameter initialization:

[0128] Weight The initial value can be set through statistical analysis or prior knowledge;

[0129] Use the training data to optimize the weights to improve the model's predictive accuracy.

[0130] Mapping function training:

[0131] Training with emotional response data , so that it can accurately describe the impact of each input data on emotional response.

[0132] Affective Response Prediction:

[0133] Use the model to predict new data and output ;

[0134] Classify affective responses according to their predicted value (e.g., no affective response, positive affective response, negative affective response).

[0135] Dynamic adjustment and optimization:

[0136] Adjust advertising strategies based on emotional response results;

[0137] If the emotional response is not ideal, it is necessary to optimize the input variables (such as image features or advertising design) and recalculate .

[0138] S4. Based on the results output from steps S2 and S3, a correlation model between image features and consumer behavior is constructed to further analyze the correlation between image features and consumer behavior.

[0139] For example, we can determine whether a particular ad color increases consumer purchase intent, whether a specific ad layout attracts more attention, or whether certain visual elements foster brand loyalty. By analyzing the relationship between image elements and consumer behavior patterns, we can develop personalized marketing plans for different consumer groups and improve marketing effectiveness. Machine learning technology can be used to deeply explore the potential impact of different image elements on the behavior of specific consumer groups.

[0140] In this embodiment, the following parameters are set in the image feature and consumer behavior correlation model:

[0141] Represents the correlation score between image features and consumer behavior, ranging from [0,1];

[0142] is the image feature vector, including characteristics (e.g., color, texture, shape);

[0143] is the consumer behavior feature vector, including characteristics (such as browsing time, click-through rate, number of purchases);

[0144] is the comprehensive score output by the dynamic analysis model;

[0145] is the weight of the image feature, which indicates the influence of each feature on consumer behavior;

[0146] is the weight of the consumer behavior, indicating the contribution of the behavior to the relevance;

[0147] The correlation model between image features and consumer behavior is as follows:

[0148] ;

[0149] in,

[0150] : Image features and behavioral characteristics The correlation function represents the statistical relationship between the two in historical data (such as correlation coefficient or mutual information).

[0151] When used, the model includes:

[0152] Data preparation:

[0153] enter and :

[0154] : Image feature vector, including color, texture and other features.

[0155] : Consumer behavior feature vector, including click-through rate, browsing time, etc.

[0156] : Image marketing effectiveness score, input by the results of the dynamic analysis model.

[0157] Correlation calculation:

[0158] Through historical data and statistical analysis, calculate :

[0159] Statistical methods (such as Pearson correlation coefficient) or information theory methods (such as mutual information) are used to calculate the correlation between image features and behavioral features.

[0160] Weight optimization:

[0161] Determine weights based on behavioral impact experiments or model training and :

[0162] For example, if a certain color is found to have a significant impact on purchasing behavior, then the corresponding ;

[0163] For behavioral characteristics Importance, based on marketing goals adjustment .

[0164] Relevance score calculation:

[0165] Substitute the weight and correlation into the formula to calculate :

[0166] high A value of indicates a strong correlation between image design and consumer behavior patterns.

[0167] Model optimization and output:

[0168] right Verify and optimize weights based on actual market feedback and and correlation function .

[0169] Output relevance scores to guide the next step of sentiment analysis and advertising optimization.

[0170] S5. Based on the correlation analysis results, evaluate the impact of different images on marketing effectiveness and adjust the image elements in marketing activities in real time.

[0171] In this embodiment, real-time data from marketing campaigns is continuously monitored to evaluate the effectiveness of different image designs. Consumer behavior is fed back into a dynamic analysis model, allowing for real-time adjustments to image elements within marketing campaigns. Image elements are continuously adjusted based on real-time feedback to optimize marketing effectiveness. This approach enables rapid response to market changes, improving the flexibility and timeliness of image marketing.

[0172] S6. Based on the correlation analysis results, a marketing image recommendation model is constructed to provide personalized marketing image recommendations for each consumer group.

[0173] In this embodiment, the marketing image recommendation model integrates image features , consumer behavior characteristics , emotional response , comprehensive score of consumer behavior status , the correlation score between image features and consumer behavior This step builds and recommends attractive and commercially valuable marketing images based on consumer preferences and contextual needs. For example, personalized advertising images can be recommended based on information such as the types of products a consumer has previously purchased and the types of ads they frequently view. This step optimizes images for specific contexts, ensuring that each consumer is exposed to the images most likely to influence their purchasing decision in a specific scenario.

[0174] The marketing image recommendation model sets the following parameters:

[0175] To consumers Image recommendation score;

[0176] Indicates the The feature vector of each candidate image;

[0177] Indicates consumers The weight vector is based on consumer behavior characteristics , emotional response and a comprehensive score of consumer behavior status ;

[0178] Indicates the The context-optimized weight of each image is used to dynamically adjust the recommendation priority.

[0179] Furthermore, the marketing image recommendation model is as follows:

[0180] ;

[0181] in,

[0182] : The weighted inner product of consumer weight and image features, quantifying the consumer's preference for a specific image.

[0183] When used, the model includes:

[0184] Data input and integration

[0185] Input image features , consumer behavior characteristics , emotional response , comprehensive score of consumer behavior status , the correlation score between image features and consumer behavior .

[0186] Features of candidate images Obtained directly through data acquisition.

[0187] Consumer weight calculation

[0188] Using consumer behavior and emotional responses Establishing the consumer's weight vector , reflecting its preferences through weighted synthesis.

[0189] Calculation combination and sentiment analysis output data.

[0190] Scenario optimization weight calculation

[0191] Dynamically adjust the weight of each candidate image based on the current scenario (such as holiday promotions, shopping seasons) .

[0192] It can be obtained by training historical marketing data of specific scenarios.

[0193] Recommendation score calculation

[0194] The image features , consumer weight , correlation score and contextual optimization weights Substitute the formula to calculate the recommendation score .

[0195] The higher the score of an image, the higher the recommendation priority.

[0196] Image sorting and output

[0197] According to the recommendation score Sort the candidate images and output the best image as the recommendation result.

[0198] S7. Evaluate the effectiveness of image marketing on different platforms through cross-platform analysis and provide optimization solutions for multi-platform marketing strategies.

[0199] In this embodiment, a comparative analysis is conducted on the marketing effects of images published on different platforms, including social media, e-commerce platforms, and search engine advertising;

[0200] We use A / B testing and cross-platform experiments to quantify the effectiveness of image marketing strategies across different platforms and assess differences between them. Through comparative analysis, we can identify which image designs perform better on specific platforms and which visual elements resonate more effectively with users on different platforms.

[0201] Through multi-platform data integration and comparative analysis, we provide enterprises with directions for cross-platform marketing optimization and ensure the consistency and effectiveness of marketing strategies in multiple channels. Example 2

[0202] This embodiment also provides a computer device, which is suitable for a marketing-based image big data analysis method, including a memory and a processor; the memory is used to store computer-executable instructions, and the processor is used to execute computer-executable instructions to implement a marketing-based image big data analysis method proposed in the above embodiment.

[0203] This embodiment also provides a storage medium having a computer program stored thereon. When the program is executed by a processor, the image big data analysis method based on marketing proposed in the above embodiment is implemented.

[0204] The computer device may be a terminal, comprising a processor, memory, a communication interface, a display, and an input device connected via a system bus. The processor provides computing and control capabilities. The memory includes non-volatile storage media and internal memory. The non-volatile storage media stores an operating system and computer programs. The internal memory provides an environment for the operating system and computer programs stored in the non-volatile storage media. The communication interface of the computer device is used to communicate with external terminals via wired or wireless communication. Wireless communication may be achieved via Wi-Fi, a carrier network, NFC (near-field communication), or other technologies. The display of the computer device may be a liquid crystal display or an electronic ink display. The input device may be a touchscreen overlay on the display, buttons, a trackball, or a touchpad on the computer device housing, or an external keyboard, touchpad, or mouse.

[0205] If a function is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, or the portion that contributes to the prior art, or a portion of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions for causing a computer device (which can be a personal computer, server, or network device, etc.) to execute all or part of the steps of the various embodiments of the method of the present invention. The aforementioned storage medium includes various media that can store program code, such as USB flash drives, mobile hard drives, read-only memories (ROMs), random access memories (RAMs), magnetic disks, or optical disks.

[0206] The logic and / or steps represented in the flowcharts or otherwise described herein, for example, can be considered as a sequenced list of executable instructions for implementing the logical functions, and can be embodied in any computer-readable medium for use by, or in conjunction with, an instruction execution system, apparatus, or device (e.g., a computer-based system, a system including a processor, or other system that can fetch and execute instructions from an instruction execution system, apparatus, or device). For purposes of this specification, a "computer-readable medium" can be any device that can contain, store, communicate, propagate, or transport a program for use by, or in conjunction with, an instruction execution system, apparatus, or device.

[0207] More specific examples (a non-exhaustive list) of computer-readable media include the following: an electrical connection with one or more wires (electronic devices), a portable computer disk cartridge (magnetic devices), a random access memory (RAM), a read-only memory (ROM), an erasable and programmable read-only memory (EPROM or flash memory), a fiber optic device, and a portable compact disc read-only memory (CDROM). In addition, the computer-readable medium may even be paper or other suitable medium on which the program is printed, since the program may be obtained electronically, for example, by optically scanning the paper or other medium and then editing, interpreting, or processing it in another suitable manner as necessary, and then storing it in a computer memory.

[0208] It should be understood that various parts of the present invention can be implemented using hardware, software, firmware, or a combination thereof. In the above-described embodiments, multiple steps or methods can be implemented using software or firmware stored in a memory and executed by a suitable instruction execution system. For example, if implemented using hardware, as in another embodiment, any one of the following technologies known in the art or a combination thereof can be used: a discrete logic circuit having logic gate circuits for implementing logic functions on data signals, an application-specific integrated circuit having suitable combinational logic gate circuits, a programmable gate array (PGA), a field-programmable gate array (FPGA), etc.

[0209] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit the present invention. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical solutions of the present invention, which should all be included in the scope of the claims of the present invention.

Claims

1. A marketing-based image big data analysis method, characterized in that: The following steps are involved: S1. Extract image features based on marketing images; S2. Based on the extracted image features, a dynamic analysis model is constructed to perform marketing scenario analysis. In step S2, the dynamic analysis model combines the changes in time series, the dynamic evolution of image features, and the feedback mechanism of consumer behavior. The dynamic analysis model sets the following parameters: A comprehensive score representing the consumer behavior status at time t, including indicators such as click-through rate, conversion rate, and purchase intention; It is the image features extracted from the image at time t, including color distribution, texture features, shape information, brand logo, and advertising elements; is the change in image features, that is, the difference in image features between time t and time t-1; It is the influencing factor of the marketing environment, indicating the interference or support of the external market environment on marketing activities at a certain moment, including holiday effects and competitor advertising; It is a hyperparameter of the model, which represents the weight of image features, environmental factors, and historical behaviors in dynamic adjustment; The dynamic analysis model is as follows: ； in, : Indicates the impact of current image features on consumer behavior at time t, It is the mapping function of image features to consumer behavior; : Indicates the impact of changes in image features on consumer behavior, Reflects the contribution of feature changes to behavioral changes; : Indicates the impact of external market environment factors on consumer behavior; : Indicates the influence of consumer behavior at the previous moment, that is, the continuous influence of consumer historical behavior on current decision-making; S3. After conducting marketing scenario analysis, construct a sentiment analysis model to evaluate consumers’ emotional responses to different types of marketing images; S4. Based on the results outputted from steps S2 and S3, a correlation model between image features and consumer behavior is constructed to further analyze the correlation between image features and consumer behavior; S5. Based on the correlation analysis results, evaluate the impact of different images on marketing effectiveness and adjust the image elements in marketing activities in real time; S6. Based on the correlation analysis results, a marketing image recommendation model is constructed to provide personalized marketing image recommendations for each consumer group; S7. Evaluate the effectiveness of image marketing on different platforms through cross-platform analysis and provide optimization solutions for multi-platform marketing strategies.

2. The image big data analysis method based on marketing according to claim 1, characterized in that: In step S1, image feature extraction includes: Extract visual features from a large number of marketing-related images, including color distribution, texture features, shape information, brand logos, advertising elements, and consumer behavior; Through deep convolutional neural network (CNN), images are converted into structured data features.

3. The image big data analysis method based on marketing according to claim 1, characterized in that: In step S3, the sentiment analysis model is as follows: ； in, represents the consumer's emotional response to the image at time t, and the emotional categories include like, dislike, and excitement; is the consumer behavior characteristic at time t; is the consumer's emotional preference at time t; are the weight parameters of this model, corresponding to image features, consumer behavior, emotional preferences, and comprehensive scores. and affective memory effects; : Mapping function of the impact of consumer behavior data on emotional response; : Mapping function of the impact of consumer emotional preferences on emotional responses; : Comprehensive score Affect mapping function for emotional responses; : Emotional memory effect, the continuity of consumers' previous emotional reactions on current evaluations.

4. The image big data analysis method based on marketing according to claim 1, characterized in that: In step S4, the following parameters are set in the image feature and consumer behavior correlation model: Represents the correlation score between image features and consumer behavior, ranging from [0,1]; is the image feature vector, including characteristics; is the consumer behavior feature vector, including characteristics; is the weight of the image feature, which indicates the influence of each feature on consumer behavior; is the weight of the consumer behavior, indicating the contribution of the behavior to the relevance; The correlation model between image features and consumer behavior is as follows: ； Among them, i represents the i-th image feature, j represents the j-th consumer behavior feature, : Image features and behavioral characteristics The correlation function represents the statistical relationship between the two in historical data.

5. The image big data analysis method based on marketing according to claim 1, characterized in that: In step S5, the effectiveness of different image designs is evaluated by continuously monitoring real-time data of marketing activities; Feed consumer behavior status into dynamic analysis models to adjust image elements in marketing activities in real time.

6. The image big data analysis method based on marketing according to claim 1, characterized in that: In step S6, the marketing image recommendation model integrates image features , consumer behavior characteristics , emotional response , comprehensive score of consumer behavior status , the correlation score between image features and consumer behavior Build and recommend attractive and commercially valuable marketing images based on consumer preferences and situational needs; The marketing image recommendation model sets the following parameters: To consumers Image recommendation score; Indicates the The feature vector of each candidate image; Indicates consumers The weight vector is based on consumer behavior characteristics , emotional response and a comprehensive score of consumer behavior status ; Indicates the The context-optimized weight of each image is used to dynamically adjust the recommendation priority.

7. The marketing-based image big data analysis method according to claim 6, characterized in that: The marketing image recommendation model is as follows: ， in, : The weighted inner product of consumer weight and image features quantifies consumer preference for images.

8. The marketing-based image big data analysis method according to claim 1, characterized in that: In step S7, a comparative analysis is performed on the marketing effects of images published on different platforms, including social media, e-commerce platforms, and search engine advertisements; Use A / B testing and cross-platform experiments to quantify the effectiveness of image marketing strategies on different platforms and evaluate the differences between platforms.

Images