Method for generating precision marketing strategies based on big data

Abstract

The present invention discloses a method for generating precision marketing strategies based on big data, which belongs to the field of precision marketing technology, and includes the following steps: S1: collecting real-time data of target users, processing the real-time data of target users, and determining target user characteristic data; S2: constructing a marketing strategy precision generation model, testing and optimizing the marketing strategy precision generation model, and determining the best marketing strategy precision generation model; S3: analyzing and predicting the target user characteristic data based on the best marketing strategy precision generation model, and generating a precision marketing strategy. The present invention solves the existing problem that different marketing strategies and promotional materials cannot be deployed in various market environments and target user groups, cannot meet user needs, and has poor marketing effects. The present invention can promote the creation of visually attractive promotional content, can better meet user needs, and increase the market share of products or services. At the same time, it can also effectively avoid waste of advertising and improve marketing effects.

Description

Technical Field

[0001] The present invention relates to the field of precision marketing technology, and in particular to a method for generating precision marketing strategies based on big data. Background Art

[0002] As the size and variety of data sets used in traditional data analytics solutions continue to increase, these traditional approaches have become increasingly limited, including the lack of ability to determine which data sets are relevant to the solution of any given complex data problem, given the increasing size and variety of data sets. For example, different styles of advertising and promotions may need to be designed for online and offline markets.

[0003] In an online environment, businesses may make greater use of digital advertising and social media promotions to attract consumers. However, offline, businesses may rely on traditional brochures, posters, outdoor billboards or host live events to increase brand awareness and customer interaction.

[0004] The Chinese patent with publication number CN113159839A discloses a multi-contact product marketing system, which relates to the field of marketing systems. The system includes a cloud platform resource system, which includes an offline marketing module, an online marketing module, a production module and an information sorting module. The input end of the offline marketing module and the input end of the online marketing module are both connected to the output end of the information sorting module. The output end of the offline marketing module is connected to the input end of the consulting unit, the input end of the experience unit and the input end of the feedback unit. The output end of the online marketing module is connected to the input end of the after-sales unit, the input end of the VR experience unit and the input end of the browsing unit. The feedback unit transmits customer feedback information to the information sorting module through the offline marketing module, and makes comprehensive contact with customers through a wide range of channels, so that customers can better understand the product, thereby facilitating the improvement of product sales. However, the patent has the following defects:

[0005] The existing ones cannot deploy different marketing strategies and promotional materials in various market environments and target user groups, cannot meet user needs well, and result in poor marketing effects. Summary of the invention

[0006] The purpose of the present invention is to provide a method for generating precise marketing strategies based on big data, which can promote the creation of visually attractive promotional content, better meet user needs, and increase the market share of products or services. At the same time, it can also effectively avoid advertising waste, improve marketing effects, and solve the problems raised in the above-mentioned background technology.

[0007] To achieve the above object, the present invention provides the following technical solutions:

[0008] The method of generating precision marketing strategies based on big data includes the following steps:

[0009] S1: Collect and process the real-time data of target users to determine the characteristic data of target users;

[0010] S2: Build a marketing strategy precision generation model, test and optimize the marketing strategy precision generation model, and determine the best marketing strategy precision generation model;

[0011] S3: Analyze and predict the target user feature data based on the best marketing strategy precision generation model to generate a precision marketing strategy;

[0012] S4: Adjust and optimize precision marketing strategies for different regions and cultures to determine more precise and effective marketing strategies.

[0013] Preferably, in S1, real-time data of target users is collected, and the following operations are performed:

[0014] Collect user identities in real time and legally to determine user identity data;

[0015] Conduct real-time and legal collection of behavior patterns to determine behavior pattern data;

[0016] Conduct real-time and legal collection of social network relationships to determine social network data;

[0017] Among them, based on user identity data, behavior pattern data and social network data, real-time data of target users based on big data is determined.

[0018] Preferably, in S1, the real-time data of the target user is processed, and the following operations are performed:

[0019] Obtain real-time data of target users based on big data;

[0020] Based on the sequential retrieval method, the real-time data of target users based on big data is retrieved;

[0021] Check the consistency of target user real-time data based on big data;

[0022] Check whether the target user's real-time data meets the requirements based on the reasonable value range and mutual relationship of each variable in the target user's real-time data;

[0023] Remove inconsistent data that is beyond the normal range, logically unreasonable or contradictory in the real-time data of target users;

[0024] Determine the consistent data that is within the normal range and logically reasonable in the real-time data of the target user;

[0025] Process invalid and missing values ​​for consistent data that is within a normal range and logically reasonable in the real-time data of target users;

[0026] Remove invalid and missing data that are of no value in generating precise marketing strategies from the target user's real-time data that is within a normal range and logically reasonable and consistent;

[0027] Identify real-time target user data that is valuable for generating precise marketing strategies;

[0028] Extract features from real-time target user data that is valuable for generating precision marketing strategies;

[0029] Determine target user characteristic data based on big data.

[0030] Preferably, removing inconsistent data that is beyond a normal range, logically unreasonable or contradictory in the real-time data of the target user, further includes using the inconsistent data that is beyond a normal range, logically unreasonable or contradictory to determine the data quality of the real-time data of the target user, including:

[0031] Extract data corresponding to inconsistent data that is beyond a normal range, logically unreasonable, or contradictory from the removed target user's real-time data;

[0032] Segment the data corresponding to the inconsistent data that exceeds a normal range, is logically unreasonable, or contradicts each other in the removed real-time data of the target user according to the collection time and number of the real-time data of the target user, and obtain the data corresponding to the inconsistent data that exceeds a normal range, is logically unreasonable, or contradicts each other in the removed real-time data of the target user corresponding to each data collection;

[0033] The data corresponding to the inconsistent data that is beyond the normal range, logically unreasonable or contradictory in the removed real-time data of the target user corresponding to each data collection is used as the target abnormal data;

[0034] The target abnormal data is used to obtain an abnormal data evaluation parameter; wherein the abnormal data evaluation parameter is obtained by the following formula:

[0035] ;

[0036] when season Where Z represents the abnormal data evaluation parameter; Z0 represents the preset abnormal data evaluation parameter benchmark value; n represents the number of times the target user's real-time data is collected; C fi represents the amount of target abnormal data corresponding to the i-th target user real-time data collection; C zirepresents the total amount of target user real-time data corresponding to the i-th target user real-time data collection; C f0 represents the amount of target abnormal data corresponding to the first target user real-time data collection; C0 represents the total amount of target user real-time data corresponding to the first target user real-time data collection; z represents the adjustment coefficient, and the adjustment coefficient is obtained by the following formula:

[0037] ;

[0038] Where n represents the number of times the target user's real-time data is collected; C fi represents the amount of target abnormal data corresponding to the i-th target user real-time data collection; C fi-1 represents the amount of target abnormal data corresponding to the i-1th target user real-time data collection; C zi represents the total amount of target user real-time data corresponding to the i-th target user real-time data collection; C zi-1 represents the total amount of target user real-time data corresponding to the i-1th target user real-time data collection;

[0039] Comparing the abnormal data evaluation parameter with a preset evaluation parameter threshold;

[0040] When the abnormal data evaluation parameter exceeds the preset evaluation parameter threshold, the data quality of the current target user's real-time data is determined to be abnormal, and a data quality abnormality alarm is issued;

[0041] When the abnormal data evaluation parameter does not exceed the preset evaluation parameter threshold, the invalid data that is worthless in generating precise marketing strategies in the real-time data of the target user that is within the normal range and logically reasonable is used in combination with the abnormal data evaluation parameter to perform a secondary evaluation on the data quality of the real-time data of the current target user to obtain a secondary evaluation result.

[0042] Preferably, when the abnormal data evaluation parameter does not exceed the preset evaluation parameter threshold, the invalid data that is worthless for generating a precision marketing strategy in the target user's real-time data within a normal range and logically reasonable is combined with the abnormal data evaluation parameter to perform a secondary evaluation on the data quality of the current target user's real-time data, and obtain a secondary evaluation result, including:

[0043] When the abnormal data evaluation parameter does not exceed the preset evaluation parameter threshold, extract invalid data that is of no value in generating a precision marketing strategy from the logically reasonable and consistent data in the real-time data of the target user within a normal range and is combined with the abnormal data evaluation parameter;

[0044] Extract the data volume corresponding to the invalid data in each target user real-time data collection;

[0045] The abnormal data evaluation parameter is adjusted using the data amount corresponding to the invalid data in each target user real-time data collection to obtain the adjusted abnormal data evaluation parameter; wherein the adjusted abnormal data evaluation parameter is obtained by the following formula:

[0046] ;

[0047] Among them, Z t represents the adjusted abnormal data evaluation parameter; Z represents the abnormal data evaluation parameter; n represents the number of times the target user's real-time data is collected; C si represents the amount of normal data corresponding to the real-time data collection of the i-th target user; C wi represents the amount of invalid data contained in the normal data corresponding to the real-time data collection of the i-th target user; C zi represents the total amount of target user real-time data corresponding to the i-th target user real-time data collection; z represents the adjustment coefficient;

[0048] Comparing the adjusted abnormal data evaluation parameter with a preset evaluation parameter threshold;

[0049] When the adjusted abnormal data evaluation parameter exceeds the preset evaluation parameter threshold, the data quality of the current target user's real-time data is determined to be abnormal, and a data quality abnormality alarm is issued;

[0050] When the adjusted abnormal data evaluation parameter does not exceed the preset evaluation parameter threshold, it is determined that the data quality of the current target user's real-time data is normal.

[0051] Preferably, in S2, a marketing strategy accurate generation model is constructed, and the following operations are performed:

[0052] According to the demand for generating precision marketing strategies based on big data, collect historical data on user identity, historical data on behavior patterns, and historical data on social networks;

[0053] Determine the target user's historical data based on the user's identity historical data, behavior pattern historical data and social network historical data;

[0054] Divide the historical data of target users to determine the training set and test set;

[0055] According to the need to generate precise marketing strategies based on big data, select the appropriate diffusion model architecture based on the training set;

[0056] Input the training set into the appropriate diffusion model architecture of choice;

[0057] Training the selected appropriate diffusion model architecture based on the training set;

[0058] Determine an accurate generation model for marketing strategies based on big data.

[0059] Preferably, in S2, the marketing strategy precision generation model is tested and optimized, and the following operations are performed:

[0060] Obtain accurate generation models of marketing strategies based on big data;

[0061] Based on the test set, the performance of the marketing strategy precision generation model based on big data is tested;

[0062] Determine the performance test results of the accurate generation model based on the marketing strategy;

[0063] Conduct in-depth mining and relevant analysis on the performance test results based on the accurate generation model of marketing strategies;

[0064] Determine the optimization and adjustment plan based on the accurate generation model of marketing strategy;

[0065] Optimize and adjust the marketing strategy precision generation model based on the optimization adjustment plan;

[0066] Determine the best marketing strategy accurate generation model after optimization and adjustment.

[0067] Preferably, in S3, the target user feature data is analyzed and predicted, and the following operations are performed:

[0068] Obtain target user feature data based on big data;

[0069] Obtain the best marketing strategy accurate generation model after optimization and adjustment;

[0070] Input the target user feature data into the optimal marketing strategy precision generation model after optimization and adjustment;

[0071] Analyze and predict the target user feature data based on the best marketing strategy accurate generation model after optimization and adjustment;

[0072] Generate precise marketing strategies based on big data, and deploy different marketing strategies and promotional materials in various market environments and target user groups.

[0073] Preferably, in S4, the precision marketing strategy is adjusted and optimized for different regions and cultures, and the following operations are performed:

[0074] Obtain precise marketing strategies based on big data;

[0075] Obtain feedback and suggestions from target users on the precision marketing strategy in real time based on the precision marketing strategy based on big data, and determine the feedback and suggestions from target users;

[0076] Based on target users’ feedback and suggestions, adjust and optimize the precision marketing strategy based on big data for different regions and cultures;

[0077] Determine more accurate and effective marketing strategies.

[0078] Preferably, in S4, a more accurate and effective marketing strategy includes:

[0079] Quickly generate marketing product background based on more accurate and effective marketing strategies;

[0080] Automatically add lighting effects to marketing products so that they blend naturally with the product background;

[0081] Make the generated marketing images directly usable in product promotional posters.

[0082] Compared with the prior art, the present invention has the following beneficial effects:

[0083] The present invention collects real-time data of target users, processes the real-time data of target users, determines characteristic data of target users, constructs a marketing strategy precision generation model according to the demand for generating precise marketing strategies based on big data, tests and optimizes the marketing strategy precision generation model, determines the best marketing strategy precision generation model, analyzes and predicts the characteristic data of target users based on the best marketing strategy precision generation model, generates precise marketing strategies, adjusts and optimizes the precise marketing strategies for different regions and cultures, determines more precise and effective marketing strategies, quickly generates marketing product backgrounds based on the more precise and effective marketing strategies, automatically adds lighting effects to marketing products so that they are naturally integrated with the product backgrounds, and directly uses the generated marketing images for product promotion posters, thereby promoting the creation of visually attractive promotional content, better meeting user needs, and increasing the market share of products or services. At the same time, it can also effectively avoid waste of advertising and improve marketing effects. Different marketing strategies and promotional materials can be deployed in various market environments and target user groups, and the content and design style of the marketing materials can be adjusted accordingly according to different regions and cultures. BRIEF DESCRIPTION OF THE DRAWINGS

[0084] Figure 1 This is a flow chart of the invention for generating a precision marketing strategy based on big data. DETAILED DESCRIPTION

[0085] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only 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.

[0086] In order to solve the existing problem of not being able to deploy different marketing strategies and promotional materials in various market environments and target user groups, not being able to meet user needs well, and resulting in poor marketing results, please refer to Figure 1 , this embodiment provides the following technical solutions:

[0087] The method of generating precision marketing strategies based on big data includes the following steps:

[0088] S1: Collect and process the real-time data of target users to determine the characteristic data of target users;

[0089] In this embodiment, as a preferred technical solution of the present invention, real-time data of target users is collected, and the following operations are performed:

[0090] Collect user identities in real time and legally to determine user identity data;

[0091] Conduct real-time and legal collection of behavior patterns to determine behavior pattern data;

[0092] Conduct real-time and legal collection of social network relationships to determine social network data;

[0093] Among them, based on user identity data, behavior pattern data and social network data, real-time data of target users based on big data is determined.

[0094] It should be noted that by legally collecting user identity data, behavior pattern data and social network data, real-time data of target users based on big data can be determined, which facilitates the generation of precise marketing strategies based on real-time data of target users based on big data.

[0095] In this embodiment, as a preferred technical solution of the present invention, the real-time data of the target user is processed by performing the following operations:

[0096] Obtain real-time data of target users based on big data;

[0097] Based on the sequential retrieval method, the real-time data of target users based on big data is retrieved;

[0098] Check the consistency of target user real-time data based on big data;

[0099] Check whether the target user's real-time data meets the requirements based on the reasonable value range and mutual relationship of each variable in the target user's real-time data;

[0100] Remove inconsistent data that is beyond the normal range, logically unreasonable or contradictory in the real-time data of target users;

[0101] Determine the consistent data that is within the normal range and logically reasonable in the real-time data of the target user;

[0102] Process invalid and missing values ​​for consistent data that is within a normal range and logically reasonable in the real-time data of target users;

[0103] Remove invalid and missing data that are of no value in generating precise marketing strategies from the target user's real-time data that is within a normal range and logically reasonable and consistent;

[0104] Identify real-time target user data that is valuable for generating precise marketing strategies;

[0105] Extract features from real-time target user data that is valuable for generating precision marketing strategies;

[0106] Determine target user characteristic data based on big data.

[0107] It should be noted that by searching the real-time data of target users based on big data, it is possible to determine the real-time data of target users that is valuable for generating precise marketing strategies, and to improve the subsequent processing speed and accuracy of the real-time data of target users. By performing feature extraction on the real-time data of target users, it is possible to determine the characteristic data of target users based on big data, and to facilitate the subsequent analysis and prediction of the characteristic data of target users, and to facilitate the rapid generation of precise marketing strategies.

[0108] Specifically, removing inconsistent data that is beyond a normal range, logically unreasonable or contradictory in the real-time data of the target user, and also including using the inconsistent data that is beyond a normal range, logically unreasonable or contradictory to determine the data quality of the real-time data of the target user, including:

[0109] Extract data corresponding to inconsistent data that is beyond a normal range, logically unreasonable, or contradictory from the removed target user's real-time data;

[0110] Segment the data corresponding to the inconsistent data that exceeds a normal range, is logically unreasonable, or contradicts each other in the removed real-time data of the target user according to the collection time and number of the real-time data of the target user, and obtain the data corresponding to the inconsistent data that exceeds a normal range, is logically unreasonable, or contradicts each other in the removed real-time data of the target user corresponding to each data collection;

[0111] The data corresponding to the inconsistent data that is beyond the normal range, logically unreasonable or contradictory in the removed real-time data of the target user corresponding to each data collection is used as the target abnormal data;

[0112] The target abnormal data is used to obtain an abnormal data evaluation parameter; wherein the abnormal data evaluation parameter is obtained by the following formula:

[0113] ;

[0114] when season Where Z represents the abnormal data evaluation parameter; Z0 represents the preset abnormal data evaluation parameter benchmark value; n represents the number of times the target user's real-time data is collected; C fi represents the amount of target abnormal data corresponding to the i-th target user real-time data collection; C zi represents the total amount of target user real-time data corresponding to the i-th target user real-time data collection; C f0 represents the amount of target abnormal data corresponding to the first target user real-time data collection; C0 represents the total amount of target user real-time data corresponding to the first target user real-time data collection; z represents the adjustment coefficient, and the adjustment coefficient is obtained by the following formula:

[0115] ;

[0116] Where n represents the number of times the target user's real-time data is collected; C fi represents the amount of target abnormal data corresponding to the i-th target user real-time data collection; C fi-1 represents the amount of target abnormal data corresponding to the i-1th target user real-time data collection; C zi represents the total amount of target user real-time data corresponding to the i-th target user real-time data collection; C zi-1 represents the total amount of target user real-time data corresponding to the i-1th target user real-time data collection;

[0117] Comparing the abnormal data evaluation parameter with a preset evaluation parameter threshold;

[0118] When the abnormal data evaluation parameter exceeds the preset evaluation parameter threshold, the data quality of the current target user's real-time data is determined to be abnormal, and a data quality abnormality alarm is issued;

[0119] When the abnormal data evaluation parameter does not exceed the preset evaluation parameter threshold, the invalid data that is worthless in generating precise marketing strategies in the real-time data of the target user that is within the normal range and logically reasonable is used in combination with the abnormal data evaluation parameter to perform a secondary evaluation on the data quality of the real-time data of the current target user to obtain a secondary evaluation result.

[0120] The technical effect of the above technical solution is: by removing inconsistent data that is beyond the normal range, logically unreasonable or contradictory in the real-time data of the target user, the accuracy and reliability of the data are directly improved. This helps to reduce decision-making errors caused by data errors or anomalies and improve the accuracy of subsequent data analysis and application.

[0121] By introducing the abnormal data evaluation parameter (Z), the solution can quantitatively evaluate the overall quality of the data based on factors such as the number of data collections, the amount of abnormal data in each collection, and the total amount of data. This refined evaluation method not only takes into account the changes in data volume, but also the dynamic changes in data quality over time, making the evaluation results more comprehensive and accurate.

[0122] By comparing the abnormal data evaluation parameters with the preset evaluation parameter thresholds, the solution can determine in real time whether the data quality of the target user's real-time data is abnormal, and issue an alarm in a timely manner when an abnormality occurs. This helps to quickly respond to data quality issues, avoid further dissemination and use of problematic data, and reduce potential losses.

[0123] After confirming that the data quality is basically reliable, the solution further uses the consistent data within the normal range and logically reasonable to identify and process the invalid data that is of no value in generating precision marketing strategies. This step helps to further purify the data and improve the effective utilization of the data, so that the data can better serve the formulation and implementation of precision marketing strategies.

[0124] Through the above series of data cleaning, quality assessment and value extraction processes, the solution provides more reliable and valuable support for data-based decision-making. Whether it is formulating precision marketing strategies or other data-based decisions, it can be based on more accurate and high-quality data, thereby improving the accuracy and effectiveness of decisions.

[0125] In summary, this technical solution significantly improves the quality and value of target users' real-time data through a series of refined data processing and evaluation steps, providing a more solid and reliable foundation for data-based decision-making.

[0126] Specifically, when the abnormal data evaluation parameter does not exceed the preset evaluation parameter threshold, the invalid data that is worthless in generating a precision marketing strategy in the target user's real-time data within a normal range and logically reasonable is combined with the abnormal data evaluation parameter to perform a secondary evaluation on the data quality of the current target user's real-time data, and obtain a secondary evaluation result, including:

[0127] When the abnormal data evaluation parameter does not exceed the preset evaluation parameter threshold, extract invalid data that is of no value in generating a precision marketing strategy from the logically reasonable and consistent data in the real-time data of the target user within a normal range and is combined with the abnormal data evaluation parameter;

[0128] Extract the data volume corresponding to the invalid data in each target user real-time data collection;

[0129] The abnormal data evaluation parameter is adjusted using the data amount corresponding to the invalid data in each target user real-time data collection to obtain the adjusted abnormal data evaluation parameter; wherein the adjusted abnormal data evaluation parameter is obtained by the following formula:

[0130] ;

[0131] Among them, Z t represents the adjusted abnormal data evaluation parameter; Z represents the abnormal data evaluation parameter; n represents the number of times the target user's real-time data is collected; C si represents the amount of normal data corresponding to the real-time data collection of the i-th target user; C wi represents the amount of invalid data contained in the normal data corresponding to the real-time data collection of the i-th target user; C zi represents the total amount of target user real-time data corresponding to the i-th target user real-time data collection; z represents the adjustment coefficient;

[0132] Comparing the adjusted abnormal data evaluation parameter with a preset evaluation parameter threshold;

[0133] When the adjusted abnormal data evaluation parameter exceeds the preset evaluation parameter threshold, the data quality of the current target user's real-time data is determined to be abnormal, and a data quality abnormality alarm is issued;

[0134] When the adjusted abnormal data evaluation parameter does not exceed the preset evaluation parameter threshold, it is determined that the data quality of the current target user's real-time data is normal.

[0135] The technical effect of the above technical solution is: after initially judging that the data quality of the target user's real-time data is basically reliable (i.e. the evaluation parameters of abnormal data do not exceed the preset threshold), the technical solution goes further into the data to identify and quantify the invalid data in the normal data. This process not only considers the integrity and accuracy of the data, but also the validity and practicality of the data, thereby achieving a deeper level of data quality evaluation.

[0136] By introducing the amount of data corresponding to invalid data and dynamically adjusting the evaluation parameters of abnormal data, the evaluation criteria for data quality are made more comprehensive and flexible. This adjustment method can more accurately reflect the true situation of data quality and avoid misjudging data quality due to the existence of some invalid data.

[0137] After identifying and quantifying invalid data, the technical solution can more accurately evaluate the overall quality of target users’ real-time data based on the quantity and proportion of invalid data. This helps companies make better use of valid data, develop more accurate marketing strategies, and improve the value of data utilization.

[0138] Through the secondary evaluation process, this technical solution enhances the ability to monitor data quality. Even when the data quality is initially judged to be normal, it can promptly detect and process invalid parts of the data to ensure that the data quality is always maintained at a high level.

[0139] Ultimately, this technical solution provides more reliable support for corporate decision-making by providing more accurate and comprehensive data quality assessment results. Whether it is formulating precision marketing strategies or other data-based decisions, they can be based on better data, thereby improving the scientificity and effectiveness of decision-making.

[0140] To sum up, this technical solution has significantly improved the quality and value of target users' real-time data through in-depth data quality assessment, dynamic adjustment of evaluation standards, improvement of data utilization value, enhancement of data quality monitoring capabilities, and optimization of decision support, providing a more solid data foundation for the development of the enterprise.

[0141] S2: Build a marketing strategy precision generation model, test and optimize the marketing strategy precision generation model, and determine the best marketing strategy precision generation model;

[0142] In this embodiment, as a preferred technical solution of the present invention, a marketing strategy accurate generation model is constructed, and the following operations are performed:

[0143] According to the demand for generating precision marketing strategies based on big data, collect historical data on user identity, historical data on behavior patterns, and historical data on social networks;

[0144] Determine the target user's historical data based on the user's identity historical data, behavior pattern historical data and social network historical data;

[0145] Divide the historical data of target users to determine the training set and test set;

[0146] According to the need to generate precise marketing strategies based on big data, select the appropriate diffusion model architecture based on the training set;

[0147] Input the training set into the appropriate diffusion model architecture of choice;

[0148] Training the selected appropriate diffusion model architecture based on the training set;

[0149] Determine an accurate generation model for marketing strategies based on big data.

[0150] In this embodiment, as a preferred technical solution of the present invention, the marketing strategy accurate generation model is tested and optimized, and the following operations are performed:

[0151] Obtain accurate generation models of marketing strategies based on big data;

[0152] Based on the test set, the performance of the marketing strategy precision generation model based on big data is tested;

[0153] Determine the performance test results of the accurate generation model based on the marketing strategy;

[0154] Conduct in-depth mining and relevant analysis on the performance test results based on the accurate generation model of marketing strategies;

[0155] Determine the optimization and adjustment plan based on the accurate generation model of marketing strategy;

[0156] Optimize and adjust the marketing strategy precision generation model based on the optimization adjustment plan;

[0157] Determine the best marketing strategy accurate generation model after optimization and adjustment.

[0158] S3: Analyze and predict the target user feature data based on the best marketing strategy precision generation model to generate a precision marketing strategy;

[0159] In this embodiment, as a preferred technical solution of the present invention, the target user feature data is analyzed and predicted, and the following operations are performed:

[0160] Obtain target user feature data based on big data;

[0161] Obtain the best marketing strategy accurate generation model after optimization and adjustment;

[0162] Input the target user feature data into the optimal marketing strategy precision generation model after optimization and adjustment;

[0163] Analyze and predict the target user feature data based on the best marketing strategy accurate generation model after optimization and adjustment;

[0164] Generate precise marketing strategies based on big data, and deploy different marketing strategies and promotional materials in various market environments and target user groups.

[0165] It should be noted that, based on the optimal marketing strategy precision generation model, the target user characteristic data is analyzed and predicted to generate a precise marketing strategy, and the precise marketing strategy is adjusted and optimized for different regions and cultures to determine a more precise and effective marketing strategy. Based on the more precise and effective marketing strategy, the marketing product background is quickly generated, and lighting effects are automatically added to the marketing product so that it blends naturally with the product background. The generated marketing images can be directly used for product promotional posters, which can promote the creation of visually attractive promotional content, better meet user needs, and increase the market share of products or services. At the same time, it can also effectively avoid advertising waste and improve marketing effectiveness. Different marketing strategies and promotional materials can be deployed in various market environments and target user groups, and the content and design style of the marketing materials can be adjusted accordingly according to different regions and cultures.

[0166] S4: Adjust and optimize precision marketing strategies for different regions and cultures to determine more precise and effective marketing strategies.

[0167] In this embodiment, as a preferred technical solution of the present invention, the precision marketing strategy is adjusted and optimized for different regions and cultures, and the following operations are performed:

[0168] Obtain precise marketing strategies based on big data;

[0169] Obtain feedback and suggestions from target users on the precision marketing strategy in real time based on the precision marketing strategy based on big data, and determine the feedback and suggestions from target users;

[0170] Based on target users’ feedback and suggestions, adjust and optimize the precision marketing strategy based on big data for different regions and cultures;

[0171] Determine more accurate and effective marketing strategies.

[0172] Among them, the content and design style of marketing materials need to be adjusted accordingly according to different regions and cultures to ensure their effectiveness and appeal through rigorous A / B testing. This test helps to determine which prompts and images can generate better engagement and conversion rates. The insights from this analysis are used to continuously improve the prompts and improve their accuracy based on real-world performance data. This process creates a self-loop of generation, evaluation and optimization, emphasizing the importance of iterative improvement, identifying the most effective features of successful prompts, thereby optimizing the overall marketing strategy and ensuring that each iteration can use the latest insights to create more attractive and targeted advertising content. This approach not only maximizes marketing effectiveness, but also continuously adapts to user behavior and trends through data-driven insights.

[0173] In this embodiment, as a preferred technical solution of the present invention, a more accurate and effective marketing strategy includes:

[0174] Quickly generate marketing product background based on more accurate and effective marketing strategies;

[0175] Automatically add lighting effects to marketing products so that they blend naturally with the product background;

[0176] Make the generated marketing images directly usable in product promotional posters.

[0177] It should be noted that based on a more accurate and effective marketing strategy, product backgrounds can be quickly generated, and lighting effects can be automatically added to marketing products so that they blend naturally with the product background. The generated marketing images can be directly used for product promotional posters, ensuring effective generation results while promoting the creation of visually attractive promotional content.

[0178] Therefore, by collecting real-time data of target users, processing the real-time data of target users, determining the characteristic data of target users, and according to the demand for generating precise marketing strategies based on big data, a precise generation model of marketing strategies is constructed, the precise generation model of marketing strategies is tested and optimized, and the best precise generation model of marketing strategies is determined. Based on the best precise generation model of marketing strategies, the characteristic data of target users is analyzed and predicted to generate precise marketing strategies, and the precise marketing strategies are adjusted and optimized for different regions and cultures to determine more precise and effective marketing strategies. Based on more precise and effective marketing strategies, marketing product backgrounds are quickly generated, and lighting effects are automatically added to marketing products so that they are naturally integrated with the product background. The generated marketing images can be directly used for product promotional posters, which can promote the creation of visually attractive promotional content, better meet user needs, and increase the market share of products or services. At the same time, it can also effectively avoid advertising waste and improve marketing effectiveness. Different marketing strategies and promotional materials can be deployed in various market environments and target user groups, and the content and design style of the marketing materials can be adjusted accordingly according to different regions and cultures.

[0179] It should be noted that, in this article, relational terms such as first and second, etc. are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device.

[0180] Although embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions and variations may be made to the embodiments without departing from the principles and spirit of the present invention, and that the scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. A method for generating precision marketing strategies based on big data, characterized in that: The steps include: S1: Collect and process the real-time data of target users to determine the characteristic data of target users; S2: Build a marketing strategy precision generation model, test and optimize the marketing strategy precision generation model, and determine the best marketing strategy precision generation model; S3: Analyze and predict the target user feature data based on the best marketing strategy precision generation model to generate a precision marketing strategy; S4: Adjust and optimize precision marketing strategies for different regions and cultures to determine more precise and effective marketing strategies; In S1, the real-time data of the target user is processed, and the following operations are performed: Obtain real-time data of target users based on big data; Based on the sequential retrieval method, the real-time data of target users based on big data is retrieved; Check the consistency of target user real-time data based on big data; Check whether the target user's real-time data meets the requirements based on the reasonable value range and mutual relationship of each variable in the target user's real-time data; Remove inconsistent data that is beyond the normal range, logically unreasonable or contradictory in the real-time data of target users; Determine the consistent data that is within the normal range and logically reasonable in the real-time data of the target user; Process invalid and missing values ​​for consistent data that is within a normal range and logically reasonable in the real-time data of target users; Remove invalid and missing data that are of no value in generating precise marketing strategies from the target user's real-time data that is within a normal range and logically reasonable and consistent; Identify real-time target user data that is valuable for generating precise marketing strategies; Extract features from real-time target user data that is valuable for generating precision marketing strategies; Determine the target user characteristic data based on big data; Removing inconsistent data that is beyond the normal range, logically unreasonable or contradictory in the real-time data of the target user, and also including using the inconsistent data that is beyond the normal range, logically unreasonable or contradictory to determine the data quality of the real-time data of the target user, including: Extract data corresponding to inconsistent data that is beyond a normal range, logically unreasonable, or contradictory from the removed target user's real-time data; Segment the data corresponding to the inconsistent data that exceeds a normal range, is logically unreasonable, or contradicts each other in the removed real-time data of the target user according to the collection time and number of the real-time data of the target user, and obtain the data corresponding to the inconsistent data that exceeds a normal range, is logically unreasonable, or contradicts each other in the removed real-time data of the target user corresponding to each data collection; The data corresponding to the inconsistent data that is beyond the normal range, logically unreasonable or contradictory in the removed real-time data of the target user corresponding to each data collection is used as the target abnormal data; The target abnormal data is used to obtain an abnormal data evaluation parameter; wherein the abnormal data evaluation parameter is obtained by the following formula: ； when season ; Where Z represents the abnormal data evaluation parameter; Z0 represents the preset abnormal data evaluation parameter benchmark value; n represents the number of times the target user's real-time data is collected; C fi represents the amount of target abnormal data corresponding to the i-th target user real-time data collection; C zi represents the total amount of target user real-time data corresponding to the i-th target user real-time data collection; C f0 represents the amount of target abnormal data corresponding to the first target user real-time data collection; C0 represents the total amount of target user real-time data corresponding to the first target user real-time data collection; z represents the adjustment coefficient, and the adjustment coefficient is obtained by the following formula: ； Where n represents the number of times the target user's real-time data is collected; C fi represents the amount of target abnormal data corresponding to the i-th target user real-time data collection; C fi-1 represents the amount of target abnormal data corresponding to the i-1th target user real-time data collection; C zi represents the total amount of target user real-time data corresponding to the i-th target user real-time data collection; C zi-1 represents the total amount of target user real-time data corresponding to the i-1th target user real-time data collection; Comparing the abnormal data evaluation parameter with a preset evaluation parameter threshold; When the abnormal data evaluation parameter exceeds the preset evaluation parameter threshold, the data quality of the current target user's real-time data is determined to be abnormal, and a data quality abnormality alarm is issued; When the abnormal data evaluation parameter does not exceed the preset evaluation parameter threshold, the invalid data that is worthless in generating a precision marketing strategy is combined with the abnormal data evaluation parameter in the target user's real-time data within a normal range and logically reasonable consistent data to perform a secondary evaluation on the data quality of the current target user's real-time data to obtain a secondary evaluation result; When the abnormal data evaluation parameter does not exceed the preset evaluation parameter threshold, the invalid data that is worthless in generating a precision marketing strategy in the target user's real-time data within the normal range and logically reasonable consistent data is combined with the abnormal data evaluation parameter to perform a secondary evaluation on the data quality of the current target user's real-time data, and obtain a secondary evaluation result, including: When the abnormal data evaluation parameter does not exceed the preset evaluation parameter threshold, extract invalid data that is of no value in generating a precision marketing strategy from the logically reasonable and consistent data in the real-time data of the target user within a normal range and is combined with the abnormal data evaluation parameter; Extract the data volume corresponding to the invalid data in each target user real-time data collection; The abnormal data evaluation parameter is adjusted using the data amount corresponding to the invalid data in each target user real-time data collection to obtain the adjusted abnormal data evaluation parameter; wherein the adjusted abnormal data evaluation parameter is obtained by the following formula: ； Among them, Z t represents the adjusted abnormal data evaluation parameter; Z represents the abnormal data evaluation parameter; n represents the number of times the target user's real-time data is collected; C si represents the amount of normal data corresponding to the real-time data collection of the i-th target user; C wi represents the amount of invalid data contained in the normal data corresponding to the real-time data collection of the i-th target user; C zi represents the total amount of target user real-time data corresponding to the i-th target user real-time data collection; z represents the adjustment coefficient; Comparing the adjusted abnormal data evaluation parameter with a preset evaluation parameter threshold; When the adjusted abnormal data evaluation parameter exceeds the preset evaluation parameter threshold, the data quality of the current target user's real-time data is determined to be abnormal, and a data quality abnormality alarm is issued; When the adjusted abnormal data evaluation parameter does not exceed the preset evaluation parameter threshold, it is determined that the data quality of the current target user's real-time data is normal.

2. The method for generating precision marketing strategies based on big data according to claim 1, characterized in that: In S1, real-time data of target users is collected, and the following operations are performed: Collect user identities in real time and legally to determine user identity data; Conduct real-time and legal collection of behavior patterns to determine behavior pattern data; Conduct real-time and legal collection of social network relationships to determine social network data; Among them, based on user identity data, behavior pattern data and social network data, real-time data of target users based on big data is determined.

3. The method for generating precision marketing strategies based on big data according to claim 2, characterized in that: In S2, a marketing strategy accurate generation model is constructed, and the following operations are performed: According to the demand for generating precision marketing strategies based on big data, collect historical data on user identity, historical data on behavior patterns, and historical data on social networks; Determine the target user's historical data based on the user's identity historical data, behavior pattern historical data and social network historical data; Divide the historical data of target users to determine the training set and test set; According to the need to generate precise marketing strategies based on big data, select the appropriate diffusion model architecture based on the training set; Input the training set into the appropriate diffusion model architecture of choice; Training the selected appropriate diffusion model architecture based on the training set; Determine an accurate generation model for marketing strategies based on big data.

4. The method for generating precision marketing strategies based on big data according to claim 3, characterized in that: In S2, the marketing strategy precision generation model is tested and optimized, and the following operations are performed: Obtain accurate generation models of marketing strategies based on big data; Based on the test set, the performance of the marketing strategy precision generation model based on big data is tested; Determine the performance test results of the accurate generation model based on the marketing strategy; Conduct in-depth mining and relevant analysis on the performance test results based on the accurate generation model of marketing strategies; Determine the optimization and adjustment plan based on the accurate generation model of marketing strategy; Optimize and adjust the marketing strategy precision generation model based on the optimization adjustment plan; Determine the best marketing strategy accurate generation model after optimization and adjustment.

5. The method for generating precision marketing strategies based on big data according to claim 4, characterized in that: In S3, the target user feature data is analyzed and predicted, and the following operations are performed: Obtain target user feature data based on big data; Obtain the best marketing strategy accurate generation model after optimization and adjustment; Input the target user feature data into the optimal marketing strategy precision generation model after optimization and adjustment; Analyze and predict the target user feature data based on the best marketing strategy accurate generation model after optimization and adjustment; Generate precise marketing strategies based on big data, and deploy different marketing strategies and promotional materials in various market environments and target user groups.

6. The method for generating precision marketing strategies based on big data according to claim 5, characterized in that: In S4, the precision marketing strategy is adjusted and optimized for different regions and cultures, and the following operations are performed: Obtain precise marketing strategies based on big data; Obtain feedback and suggestions from target users on the precision marketing strategy in real time based on the precision marketing strategy based on big data, and determine the feedback and suggestions from target users; Based on target users’ feedback and suggestions, adjust and optimize the precision marketing strategy based on big data for different regions and cultures; Determine more accurate and effective marketing strategies.

7. The method for generating precision marketing strategies based on big data according to claim 6, characterized in that: In S4, more accurate and effective marketing strategies include: Quickly generate marketing product background based on more accurate and effective marketing strategies; Automatically add lighting effects to marketing products so that they blend naturally with the product background; Make the generated marketing images directly usable in product promotional posters.

Images