Digital marketing method and system, storage medium and computer

Abstract

The invention provides a digital marketing method and system, a storage medium and a computer, and the method comprises the steps: obtaining multi-dimensional data of a customer group, encrypting the data, and integrating the encrypted data; carrying out feature extraction on the integrated data based on a distributed big data calculation framework; performing label setting on customers in the customer group based on the extracted features to form a to-be-screened customer group, and performing intelligent product matching according to the customers in the to-be-screened customer group to form a matching white list of the products and the customers; and a marketing model is constructed, the matched white list is evaluated and analyzed according to the marketing model, so that the purchase intention of the customer is quantified, and the intention customer is screened out for product marketing recommendation. According to the digital marketing method provided by the invention, the target intention customer can be accurately obtained, and the marketing conversion rate is improved.

Description

Technical Field

[0001] The present invention relates to the field of digital marketing technology, and particularly relates to a digital marketing method, system, storage medium and computer. Background Art

[0002] Marketing is a process of creating, delivering and exchanging products or values to achieve organizational goals. Its core lies in effectively communicating and exchanging to transfer products or services from producers to consumers, while establishing long-term customer relationships.

[0003] Currently, traditional marketing includes static marketing and dynamic marketing; static marketing mainly includes passive marketing methods such as advertising placement, which is difficult to respond to market changes in real time, has low advertising placement efficiency and over-relies on manual work in marketing, with low marketing efficiency and high costs, and is unable to accurately obtain target potential customers, resulting in low conversion rates; dynamic marketing, by obtaining single-platform data such as customers' e-commerce sales records, cannot comprehensively capture customers' cross-platform behavior characteristics, is difficult to analyze multi-industry customer behavior information, and is difficult to accurately obtain target potential customers. Summary of the Invention

[0004] Based on this, the purpose of the present invention is to provide a digital marketing method, system, and storage medium to solve the technical problems existing in the prior art.

[0005] The present invention proposes a digital marketing method, including: Obtaining multi-dimensional data of a customer group based on a multi-source data engine, and performing dynamic desensitization encryption processing on the multi-dimensional data; Integrating the encrypted multi-dimensional data, and performing feature extraction on the integrated data based on a distributed big data computing framework to capture key information and dynamic features in the multi-dimensional data; Setting labels for customers in the customer group based on the key information and dynamic features in the multi-dimensional data to form a customer group to be screened, and performing intelligent product matching according to the customers in the customer group to be screened to form a matching white list of products and customers; Constructing a marketing model, and evaluating and analyzing the matching white list according to the marketing model to quantify the purchase intention of customers, and screening out potential customers for product marketing recommendations.

[0006] Preferably, the step of obtaining multi-dimensional data of a customer group based on a multi-source data engine and performing dynamic desensitization encryption processing on the multi-dimensional data includes: Performing incremental extraction and full extraction in the data source based on a multi-source data engine to obtain multi-dimensional data of a customer group, where the multi-dimensional data at least includes regional data, time data, brand category data, online sales data, and offline sales data; Build a proxy model, where the proxy model includes at least a proxy side and a server side; Send a request message to the proxy side based on multi-dimensional data. After receiving the request message, the proxy side forwards it to the server side; Query sensitive information in the server side, return the result set message containing the sensitive information to the proxy side. The proxy side queries the record closest to the current time from the session time table, and obtains the corresponding customer name according to the session ID in the record; The proxy side obtains the desensitization scheme corresponding to the current customer name from the desensitization configuration table according to the corresponding customer name, performs data desensitization, assembles the desensitized data, forms a result set message, and completes the dynamic desensitization and encryption of the data.

[0007] Preferably, the steps of integrating the encrypted multi-dimensional data, extracting features from the integrated data based on a distributed big data computing framework, and capturing key information and dynamic features in the multi-dimensional data include: Deduplicate, fill, and perform outlier detection on the encrypted multi-dimensional data to complete data cleaning, and perform standardization, discretization, and stratification on the cleaned data to complete data transformation; Based on the ETL tool, batch load the transformed data into the distributed storage area, extract and integrate the features corresponding to the multi-dimensional data respectively, and perform visualization processing on the integrated multi-dimensional data to obtain a visualization image corresponding to the key information and dynamic features in the multi-dimensional data.

[0008] Preferably, the steps of setting labels for customers in the customer group based on the key information and dynamic features in the multi-dimensional data, forming a customer group to be screened, and performing intelligent product matching according to the customers in the customer group to be screened include: Input the visualization image into the IKG module. Based on the image encoder in the IKG module, receive the visualization image and perform feature extraction to generate the corresponding serialized image features; The image decoder makes predictions starting from the start sequence in the serialized image features, gradually predicts the text description sequence corresponding to the visualization image based on the greedy algorithm, obtains the token sequence text, and cleans the token sequence text to obtain candidate words; Encode all the token information in the candidate words based on the text encoder in the IKG module, calculate the graphic-text similarity between the encoded token information and the serialized image features, screen out several relevant keywords and their weights, and use the several keywords as the labels corresponding to the customers in the customer group to form a customer group to be screened; Obtain the tag library of the product, calculate the similarity between all keywords corresponding to the customers in the customer group to be screened and the tag library to obtain a similarity matrix, weight the weights corresponding to the keywords into the similarity matrix to obtain the probability distribution between the product and the customers, and determine the intelligent matching result according to the probability distribution.

[0009] Preferably, the expression of the serialized image feature is:

[0010] In the formula, respectively represent the global feature and the local feature of the visualized image, represents the image encoder, represents the input visualized image; The expression for the step-by-step prediction by the greedy algorithm is:

[0011]

[0012]

[0013] In the formula, represents the output after passing through the linear layer and normalization during the prediction of the th image decoder, represents the result after the self-attention layer and the residual connection during the prediction of the th image decoder, represents the intermediate result of fusing the global feature of the visualized image and the cross-attention layer during the prediction of the th image decoder; represents the multi-head self-attention layer, represents the multi-head cross-attention layer, represents the multi-layer perceptron, represents the normalization layer.

[0014] The present invention also proposes a digital marketing system, including; An encryption module for obtaining multi-dimensional data of the customer group based on a multi-source data engine and performing dynamic desensitization encryption processing on the multi-dimensional data; A capture module for integrating the encrypted multi-dimensional data, extracting features from the integrated data based on a distributed big data computing framework, and capturing key information and dynamic features in the multi-dimensional data; A matching module for setting tags for the customers in the customer group based on the key information and dynamic features in the multi-dimensional data to form a customer group to be screened, and performing intelligent matching of products according to the customers in the customer group to be screened to form a matching white list of product-customer adaptation; A recommendation module, which is used to build a marketing model, evaluate and analyze a matching whitelist according to the marketing model, quantify the purchase intention of customers, and screen out potential customers for product marketing recommendations.

[0015] Preferably, the encryption module includes: An extraction unit, which is used to perform incremental extraction and full extraction in a data source based on a multi-source data engine to obtain multi-dimensional data of a customer group, and the multi-dimensional data at least includes regional data, time data, brand category data, online sales data, and offline sales data; A construction unit, which is used to build an agent model, and the agent model at least includes an agent side and a server side; A sending unit, which is used to send a request message to the agent side based on the multi-dimensional data. After receiving the request message, the agent side forwards it to the server side; A query unit, which is used to query sensitive information in the server side, return a result set message containing sensitive information to the agent side, the agent side queries the record closest to the current time from the session time table, and obtains the corresponding customer name according to the session ID in the record; A desensitization unit, which is used to perform data desensitization based on the agent side obtaining a desensitization scheme corresponding to the current customer name from a desensitization configuration table according to the corresponding customer name, assemble the desensitized data to form a result set message, and complete the dynamic desensitization and encryption of the data.

[0016] Preferably, the capture module includes: A processing unit, which is used to perform deduplication, filling, and outlier checking on the encrypted multi-dimensional data to complete data cleaning, and perform standardization, discretization, and stratification processing on the cleaned data to complete data conversion; A visualization unit, which is used to batch load the converted data into a distributed storage area based on an ETL tool, extract and integrate the features corresponding to the multi-dimensional data respectively, and perform visualization processing on the integrated multi-dimensional data to obtain a visualization image of the key information and dynamic features in the multi-dimensional data.

[0017] The present invention also provides a storage medium, on which a computer program is stored, and when the program is executed by a processor, the above digital marketing method is implemented.

[0018] The present invention also provides a computer, which includes a memory, a processor, and a computer program stored on the memory and executable on the processor. When the processor executes the computer program, the above digital marketing method is implemented.

[0019] The beneficial effects of the present invention compared with the prior art are as follows: The present application provides a dynamic active marketing method. First, a multi-source data engine is used to obtain multi-dimensional data of the customer group, and the multi-dimensional data is desensitized and encrypted to build a security defense line from the data storage source, providing full-life-cycle security protection for the data assets of the business system, ensuring the safe circulation and application of data within the compliance framework; the encrypted multi-dimensional data is integrated, and the multi-dimensional data can comprehensively capture the cross-platform behavior characteristics of customers, improving the accuracy of target customer selection; the customers in the customer group are tagged through the integrated data, and then the products are intelligently matched with the tagged customers to form a matching white list of product-customer adaptation; the potential customers are accurately located through the matching white list, greatly reducing the customer volume of subsequent marketing, which is conducive to improving the subsequent marketing reach rate and conversion rate, building a marketing model, and evaluating and analyzing the matching white list according to the marketing model to quantify the purchase intention of customers and screen out potential customers for product marketing recommendations; the digital marketing method provided by the present application has accurate product-customer matching, high marketing reach rate and conversion rate, and is suitable for large-scale promotion.

[0020] Additional aspects and advantages of the present invention will be given in part in the following description, become apparent in part from the following description, or be learned through the practice of the present invention. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] Figure 1 is a flowchart of the digital marketing method in Embodiment 1 of the present invention; Figure 2 is a structural block diagram of a computer in Embodiment 4 of the present invention.

[0022] The following specific embodiments will further illustrate the present invention in conjunction with the above-mentioned drawings. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0023] To facilitate the understanding of the present invention, the present invention will be described more comprehensively with reference to the relevant drawings. Several embodiments of the present invention are shown in the drawings. However, the present invention can be implemented in many different forms and is not limited to the embodiments described herein. On the contrary, these embodiments are provided to make the disclosure of the present invention more thorough and comprehensive.

[0024] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by those of ordinary skill in the technical field to which the present invention belongs. The terms used in the description of the present invention herein are only for the purpose of describing specific embodiments and are not intended to limit the present invention. The term "and / or" used herein includes any and all combinations of one or more of the related listed items.

[0025] Embodiment 1 Please refer toFigure 1 , the digital marketing method in the first embodiment of the present invention is shown. Specifically, the digital marketing method specifically includes steps S10 to S40: S10, obtaining multi-dimensional data of the customer group based on a multi-source data engine, and performing dynamic desensitization and encryption processing on the multi-dimensional data; Optionally, the step of obtaining multi-dimensional data of the customer group based on a multi-source data engine and performing dynamic desensitization and encryption processing on the multi-dimensional data includes: Performing incremental extraction and full extraction in the data source based on a multi-source data engine to obtain multi-dimensional data of the customer group, where the multi-dimensional data at least includes regional data, time data, brand data, online sales data, and offline sales data; Constructing an agent model, where the agent model at least includes an agent side and a server side; Sending a request message to the agent side based on the multi-dimensional data. After receiving the request message, the agent side forwards it to the server side; Querying for sensitive information in the server side, returning the result set message containing sensitive information to the agent side. The agent side queries the record closest to the current time from the session time table and obtains the corresponding customer name according to the session ID in the record; The agent side obtains the desensitization scheme corresponding to the current customer name from the desensitization configuration table according to the corresponding customer name for data desensitization, assembles the desensitized data, forms a result set message, and completes the dynamic desensitization and encryption of the data.

[0026] Optionally, using a self-developed multi-source data fusion engine to extract data in different dimensions such as regions, e-commerce, and fast-moving consumer goods. The agent model can be a MySQL agent model, which is used to monitor, analyze, or transmit the communication between data, and perform data communication and desensitization through the agent side and the server side. Specifically, the agent side queries the record closest to the current time from the session time table to obtain the session ID, and queries the corresponding customer name in the session time table according to the session ID, and determines whether the current customer needs to be desensitized. If so, obtains the desensitization scheme corresponding to the customer in the desensitization configuration table for data desensitization, and replaces the desensitized data with the original data for storage; Data dynamic desensitization and encryption can build a solid security defense from the data storage source, provide full-life-cycle security protection for the data assets of the business system, and ensure the safe circulation and application of data within the compliance framework.

[0027] S20, integrating the encrypted multi-dimensional data, and performing feature extraction on the integrated data based on a distributed big data computing framework to capture key information and dynamic features in the multi-dimensional data; The step of integrating the encrypted multi-dimensional data, extracting features from the integrated data based on a distributed big data computing framework, and capturing key information and dynamic features in the multi-dimensional data includes: Deduplicate, fill, and perform outlier detection on the encrypted multi-dimensional data to complete data cleaning, and perform standardization, discretization, and stratification on the cleaned data to complete data transformation; Based on an ETL tool, batch load the transformed data into a distributed storage area, extract and integrate the features corresponding to the multi-dimensional data respectively, and perform visualization processing on the integrated multi-dimensional data to obtain a visualization image corresponding to the key information and dynamic features in the multi-dimensional data.

[0028] Since the data source types of data in different dimensions are intricate and complex, before data integration, data processing is required to extract the data required for precision marketing from the complex original data format. Data cleaning is to solve quality problems such as data integrity, legality, and consistency, making the data more suitable for mining; data transformation is to make the data more consistent and easier to be processed by the model through standardization, discretization, and stratification, etc.; based on the ETL (Extract-Transform-Load) tool, batch load the transformed data into a distributed storage area; store the data in partitions and segments to improve the analysis ability and response ability; for example, store structured, semi-structured, and unstructured data in partitions and segments. Data visualization is to use computer graphics and image processing technologies to convert data into a visualization image and display it on the screen. The main characteristics of data visualization are: interactivity, multi-dimensionality, and visualization; performing visualization can help quickly analyze the dynamic changes of customers; in this embodiment, by obtaining data in different dimensions, the cross-platform behavior characteristics of customers can be comprehensively captured, and the accuracy of target customer selection can be improved. The integrated data is analyzed through a distributed big data computing framework, which can realize real-time feature extraction, accurately capture key information and dynamic trends in data in various fields, and provide solid data support and powerful technical impetus for subsequent data mining, analysis, and business decision-making.

[0029] S30, set labels for the customers in the customer group based on the key information and dynamic features in the multi-dimensional data to form a customer group to be screened, and perform intelligent product matching according to the customers in the customer group to be screened to form a matching white list of product-customer adaptation; The step of setting labels for the customers in the customer group based on the key information and dynamic features in the multi-dimensional data to form a customer group to be screened, and performing intelligent product matching according to the customers in the customer group to be screened includes: Input the visualization image into the IKG module, and based on the image encoder in the IKG module, receive the visualization image and perform feature extraction to generate corresponding serialized image features; The image decoder makes predictions starting from the start sequence in the serialized image features, gradually predicts the text description sequence corresponding to the visualized image based on the greedy algorithm, obtains the token sequence text, and cleans the token sequence text to obtain candidate words; Based on the text encoder in the IKG module, encode all the token information in the candidate words, calculate the text-image similarity between the encoded token information and the serialized image features, screen out several relevant keywords and their weights, and use the several keywords as the labels corresponding to the customers in the customer group to form the customer group to be screened; Obtain the tag library of the product, calculate the similarity between all the keywords corresponding to the customers in the customer group to be screened and the tag library, obtain the similarity matrix, weight the weights corresponding to the keywords into the similarity matrix, obtain the probability distribution between the product and the customers, and determine the intelligent matching result according to the probability distribution.

[0030] The expression of the serialized image features is:

[0031] In the formula, respectively represent the global feature and the local feature of the visualized image, represents the image encoder, represents the input visualized image; The expression for the step-by-step prediction by the greedy algorithm is:

[0032]

[0033]

[0034] In the formula, represents the output after passing through the linear layer and normalization during the prediction of the th image decoder, represents the result after the self-attention layer and residual connection during the prediction of the th image decoder, represents the intermediate result of fusing the global feature of the visualized image and the cross-attention layer during the prediction of the th image decoder; represents the multi-head self-attention layer, represents the multi-head cross-attention layer, represents the multi-layer perceptron, represents the normalization layer.

[0035] Optionally, the customer's visual image is a dynamic customer portrait model. The visual image is input into the image keyword generation (IKG) module. In the IKG module, the contrastive language-image pre-training (CLIP) model is used as an image encoder for image encoding to generate corresponding serialized image features. A Transformer-based bidirectional encoder is used as an image decoder to gradually predict the text description sequence corresponding to the visual image through a greedy algorithm. Then, stop words, invalid words, and redundant words in the token sequence text are removed to obtain candidate words. All token information is encoded through the text encoder in the CLIP model. In the CLIP model, the visual and text encoders are fully aligned, so text-image similarity calculation can be performed to screen out k most relevant keywords and their weights. Customer tagging is set through the keywords and weights to generate a multi-dimensional and accurate customer tag system. The tag library of the product is obtained, and the accuracy of target customer identification is improved through a similarity calculation method to achieve intelligent matching of "product - customer", and a customer white list is obtained.

[0036] S40. Build a marketing model, and evaluate and analyze the matching white list according to the marketing model to quantify the purchase intention of customers, and screen out potential customers for product marketing recommendations.

[0037] After obtaining the customer white list, the intelligent AI outbound call system can be used to make customer calls according to preset scripts and strategies to communicate and interact with customers efficiently. The intention model is used to comprehensively classify the customer intention level. Optionally, the intention labels can be divided into the following categories: A level (definite intention), B level (possible intention), C level (definite rejection), D level (customer is busy), E level (call failed), F level (invalid customer), etc., to scientifically quantify and analyze the purchase intention of customers and accurately screen out high-intention customers. The AI outbound call system has significantly increased the outbound call volume and communication efficiency compared with the traditional manual outbound call. The intention model evaluation helps the bank focus on high-value customer resources and optimize the allocation of marketing resources. An AI customer service robot is introduced, which can provide consulting services to customers 24 hours a day, answer common questions, and collect customer feedback. Optionally, the marketing model can also include marketing tools such as SMS, carrier-grade phone lines, and business process outsourcing (BPO). The carrier-grade phone lines ensure a 99% connection rate, and the SMS channel direct reach rate leads the industry, enabling efficient reach to target customers.

[0038] In summary, the present application provides a dynamic active marketing method. First, a multi-source data engine is used to obtain multi-dimensional data of the customer group, and the multi-dimensional data is desensitized and encrypted to build a security defense line from the data storage source, providing full-life cycle security protection for the data assets of the business system, ensuring the safe circulation and application of data within the compliance framework; the encrypted multi-dimensional data is integrated, and the multi-dimensional data can comprehensively capture the cross-platform behavior characteristics of customers, improving the accuracy of target customer selection; the customers in the customer group are labeled based on the integrated data, and then the customers with set labels are intelligently matched with products to form a matching whitelist of product-customer adaptation; the intention customers are accurately located through the matching whitelist, greatly reducing the customer volume of subsequent marketing, facilitating the improvement of subsequent marketing reach rate and conversion rate, building a marketing model, and evaluating and analyzing the matching whitelist according to the marketing model to quantify the purchase intention of customers, screening out intention customers for product marketing recommendation; the digital marketing method provided by the present application has accurate product-customer matching, high marketing reach rate and conversion rate; it is suitable for large-scale promotion.

[0039] Embodiment 2 This embodiment provides a digital marketing system, including: An encryption module, configured to obtain multi-dimensional data of the customer group based on a multi-source data engine, and perform dynamic desensitization and encryption processing on the multi-dimensional data; A capture module, configured to integrate the encrypted multi-dimensional data, and perform feature extraction on the integrated data based on a distributed big data computing framework to capture key information and dynamic features in the multi-dimensional data; A matching module, configured to set labels for customers in the customer group based on the key information and dynamic features in the multi-dimensional data to form a customer group to be screened, and perform intelligent product matching on the customers in the customer group to be screened to form a matching whitelist of product-customer adaptation; A recommendation module, configured to build a marketing model, and evaluate and analyze the matching whitelist according to the marketing model to quantify the purchase intention of customers, and screen out intention customers for product marketing recommendation.

[0040] Preferably, the step of obtaining multi-dimensional data of the customer group based on a multi-source data engine and performing dynamic desensitization and encryption processing on the multi-dimensional data includes: Performing incremental extraction and full extraction in the data source based on a multi-source data engine to obtain multi-dimensional data of the customer group, where the multi-dimensional data at least includes regional data, time data, brand data, online sales data, and offline sales data; Building an agent model, where the agent model at least includes an agent side and a server side; Sending a request message to the agent side based on the multi-dimensional data, and after receiving the request message, the agent side forwards it to the server side; Query sensitive information on the server side, return the result set message containing sensitive information to the proxy side. The proxy side queries the record closest to the current time from the session time table, and obtains the corresponding customer name according to the session ID in the record; The proxy side obtains the desensitization scheme corresponding to the current customer name from the desensitization configuration table according to the corresponding customer name, performs data desensitization, assembles the desensitized data, forms a result set message, and completes the dynamic desensitization encryption of the data.

[0041] Preferably, the step of integrating the encrypted multi-dimensional data, extracting features from the integrated data based on the distributed big data computing framework, and capturing key information and dynamic features in the multi-dimensional data includes: Deduplicate, fill, and perform outlier detection on the encrypted multi-dimensional data to complete data cleaning, and perform standardization, discretization, and stratification on the cleaned data to complete data transformation; Based on the ETL tool, batch load the transformed data into the distributed storage area, extract and integrate the features corresponding to the multi-dimensional data respectively, and perform visualization processing on the integrated multi-dimensional data to obtain the visualization image corresponding to the key information and dynamic features in the multi-dimensional data.

[0042] Preferably, the step of setting labels for customers in the customer group based on the key information and dynamic features in the multi-dimensional data, forming a customer group to be screened, and performing intelligent product matching according to the customers in the customer group to be screened includes: Input the visualization image into the IKG module. Based on the image encoder in the IKG module, receive the visualization image and perform feature extraction to generate the corresponding serialized image features; The image decoder makes predictions starting from the start sequence in the serialized image features, gradually predicts the text description sequence corresponding to the visualization image based on the greedy algorithm, obtains the token sequence text, and cleans the token sequence text to obtain candidate words; Based on the text encoder in the IKG module, encode all the token information in the candidate words, calculate the text-image similarity between the encoded token information and the serialized image features, screen out several relevant keywords and their weights, and use the several keywords as the labels corresponding to the customers in the customer group to form a customer group to be screened; Obtain the label library of the product, calculate the similarity between all the keywords corresponding to the customers in the customer group to be screened and the label library, obtain the similarity matrix, weight the weights corresponding to the keywords into the similarity matrix, obtain the probability distribution between the product and the customer, and determine the intelligent matching result according to the probability distribution.

[0043] Preferably, the expression of the serialized image features is:

[0044] In the formula, respectively represent the global feature and the local feature of the visualization image, represents the image encoder, represents the input visualization image; The expression for the step-by-step prediction by the greedy algorithm is:

[0045]

[0046]

[0047] In the formula, represents the output after passing through the linear layer and normalization during the prediction of the th image decoder, represents the result after the self-attention layer and the residual connection during the prediction of the th image decoder, represents the intermediate result of fusing the global feature of the visualization image and the cross-attention layer during the prediction of the th image decoder; represents the multi-head self-attention layer, represents the multi-head cross-attention layer, represents the multi-layer perceptron, represents the normalization layer.

[0048] Embodiment III This embodiment proposes a storage medium, on which a computer program is stored, and when the program is executed by a processor, it implements the digital marketing method as described above.

[0049] Embodiment IV The present invention also proposes a computer. Please refer to Figure 2 , which shows the computer in the embodiment of the present invention, including a memory 10, a processor 20, and a computer program 30 stored on the memory 10 and executable on the processor 20. When the processor 20 executes the computer program 30, it implements the digital marketing method as described above.

[0050] Among them, the memory 10 includes at least one type of storage medium, and the storage medium includes flash memory, hard disk, multimedia card, card-type memory (such as SD or DX memory, etc.), magnetic memory, magnetic disk, optical disc, etc. The memory 10 can be an internal storage unit of a computer in some embodiments, such as the hard disk of the computer. The memory 10 can also be an external storage device in other embodiments, such as a plug-in hard disk, Smart Media Card (SMC), Secure Digital (SD) card, Flash Card, etc. Further, the memory 10 can also include both an internal storage unit of the computer and an external storage device. The memory 10 can be used not only to store application software installed on the computer and various types of data, but also to temporarily store data that has been output or will be output.

[0051] Among them, the processor 20 can be an Electronic Control Unit (ECU, also known as a vehicle computer), a Central Processing Unit (CPU), a controller, a microcontroller, a microprocessor, or other data processing chips in some embodiments, and is used to run program codes stored in the memory 10 or process data, such as executing an access restriction program, etc.

[0052] It should be noted that Figure 2 the structure shown does not constitute a limitation on the computer. In other embodiments, the computer may include fewer or more components than shown in the figure, or combine certain components, or have different component arrangements.

[0053] Those skilled in the art can understand that the logic and / or steps represented in the flowchart or described in other ways herein, for example, can be considered as a definite sequence list of executable instructions for implementing logical functions, and can be specifically implemented in any computer-readable medium for use by an instruction execution system, apparatus, or device (such as a computer-based system, a system including a processor, or other systems that can fetch and execute instructions from the instruction execution system, apparatus, or device), or in combination with these instruction execution systems, apparatuses, or devices. For the purposes of this specification, a "computer-readable medium" can be any device that can contain, store, communicate, propagate, or transmit a program for use by or in combination with an instruction execution system, apparatus, or device.

[0054] More specific examples (a non-exhaustive list) of computer-readable media include the following: electrical connections (electronic devices) having one or more wirings, portable computer disk cartridges (magnetic devices), random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fiber devices, and portable compact disc read-only memory (CDROM). Additionally, the computer-readable media can even be paper or other suitable media on which the program can be printed, since the program can be obtained electronically, for example, by optically scanning the paper or other media, followed by editing, interpretation, or, if necessary, other suitable processing, and then storing it in a computer memory.

[0055] It should be understood that various parts of the present invention can be implemented by hardware, software, firmware, or a combination thereof. In the above embodiments, multiple steps or methods can be implemented by software or firmware stored in a memory and executed by a suitable instruction execution system. For example, if implemented by hardware, as in another embodiment, any one or a combination of the following techniques well known in the art can be used: discrete logic circuits having logic gate circuits for implementing logical functions on data signals, application specific integrated circuits having appropriate combinational logic gate circuits, programmable gate arrays (PGA), field programmable gate arrays (FPGA), etc.

[0056] The technical features of the above-described embodiments can be combined arbitrarily. For the sake of brevity of description, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, it should be considered as falling within the scope described in this specification.

[0057] The above-described embodiments merely represent several implementation manners of the present application. Their descriptions are relatively specific and detailed, but should not be construed as limiting the scope of the invention patent. It should be noted that for those of ordinary skill in the art, without departing from the concept of the present application, several modifications and improvements can be made, and these all belong to the protection scope of the present application. Therefore, the protection scope of the patent of the present application shall be subject to the appended claims.

Claims

1. A digital marketing method, characterized in that, Including: Obtain multi-dimensional data of the customer group based on a multi-source data engine, and perform dynamic desensitization encryption processing on the multi-dimensional data; Integrate the encrypted multi-dimensional data, and perform feature extraction on the integrated data based on a distributed big data computing framework to capture key information and dynamic features in the multi-dimensional data; Set labels for customers in the customer group based on the key information and dynamic features in the multi-dimensional data to form a customer group to be screened, and perform intelligent product matching according to the customers in the customer group to be screened to form a matching white list of product-customer adaptation; Construct a marketing model, and perform evaluation and analysis on the matching white list according to the marketing model to quantify the purchase intention of customers, and screen out potential customers for product marketing recommendations.

2. The digital marketing method according to claim 1, characterized in that, The steps of obtaining multi-dimensional data of the customer group based on a multi-source data engine and performing dynamic desensitization encryption processing on the multi-dimensional data include: Perform incremental extraction and full extraction in the data source based on a multi-source data engine to obtain multi-dimensional data of the customer group, and the multi-dimensional data at least includes regional data, time data, brand data, online sales data, and offline sales data; Construct an agent model, and the agent model at least includes an agent side and a server side; Send a request message to the agent side based on the multi-dimensional data, and the agent side forwards the request message to the server side after receiving it; Query sensitive information in the server side, return the result set message containing sensitive information to the agent side, the agent side queries the record closest to the current time from the session time table, and obtains the corresponding customer name according to the session ID in the record; The agent side obtains the desensitization scheme corresponding to the current customer name from the desensitization configuration table according to the corresponding customer name for data desensitization, assembles the desensitized data to form a result set message, and completes the dynamic desensitization encryption of the data.

3. The digital marketing method according to claim 2, characterized in that, The steps of integrating the encrypted multi-dimensional data, performing feature extraction on the integrated data based on a distributed big data computing framework, and capturing key information and dynamic features in the multi-dimensional data include: Perform deduplication, filling, and outlier detection on the encrypted multi-dimensional data to complete data cleaning, and perform standardization, discretization, and stratification processing on the cleaned data to complete data transformation; Based on an ETL tool, batch load the transformed data into a distributed storage area, extract and integrate the features corresponding to the multi-dimensional data respectively, and perform visualization processing on the integrated multi-dimensional data to obtain a visualization image corresponding to the key information and dynamic features in the multi-dimensional data.

4. The digital marketing method according to claim 3, characterized in that, The steps of setting labels for customers in the customer group based on the key information and dynamic features in the multi-dimensional data to form a customer group to be screened, and performing intelligent product matching according to the customers in the customer group to be screened include: Input the visualization image into the IKG module, and the image encoder in the IKG module receives the visualization image and performs feature extraction to generate corresponding serialized image features; The image decoder makes predictions starting from the start sequence in the serialized image features, gradually predicts the text description sequence corresponding to the visualized image based on the greedy algorithm, obtains the text of the token sequence, and cleans the text of the token sequence to obtain candidate words; Based on the text encoder in the IKG module, encode all the token information in the candidate words, calculate the text-image similarity between the encoded token information and the serialized image features, screen out several relevant keywords and their weights, and use the several keywords as the labels corresponding to the customers in the customer group to form a customer group to be screened; Obtain the label library of the product, calculate the similarity between all the keywords corresponding to the customers in the customer group to be screened and the label library to obtain a similarity matrix, weight the weights corresponding to the keywords into the similarity matrix to obtain the probability distribution between the product and the customers, and determine the intelligent matching result according to the probability distribution.

5. The digital marketing method according to claim 4, wherein The expression of the serialized image features is: In the formula, respectively represent the global feature and the local feature of the visualization image, represents the image encoder, represents the input visualization image; The expression for the gradual prediction by the greedy algorithm is: Wherein, represents the output after passing through the linear layer and normalization during the prediction of the th image decoder, represents the result after the self-attention layer and residual connection during the prediction of the th image decoder, represents the intermediate result of fusing the global features of the visualization image and the cross-attention layer during the prediction of the th image decoder; represents the multi-head self-attention layer, represents the multi-head cross-attention layer, represents the multi-layer perceptron, represents the normalization layer.

6. A digital marketing system, characterized in that, including; An encryption module for obtaining multi-dimensional data of the customer group based on a multi-source data engine and performing dynamic desensitization encryption processing on the multi-dimensional data; A capture module for integrating the encrypted multi-dimensional data, extracting features from the integrated data based on a distributed big data computing framework, and capturing key information and dynamic features in the multi-dimensional data; A matching module for setting labels for the customers in the customer group based on the key information and dynamic features in the multi-dimensional data to form a customer group to be screened, and performing intelligent product matching on the customers in the customer group to be screened to form a matching white list of product-customer adaptation; A recommendation module for building a marketing model, evaluating and analyzing the matching white list according to the marketing model to quantify the purchase intention of the customers, and screening out potential customers for product marketing recommendations.

7. The digital marketing system according to claim 6, characterized in that, The encryption module includes: An extraction unit for performing incremental extraction and full extraction in the data source based on a multi-source data engine to obtain multi-dimensional data of the customer group, where the multi-dimensional data at least includes regional data, time data, brand data, online sales data, and offline sales data; A construction unit for constructing an agent model, where the agent model at least includes an agent side and a server side; A sending unit for sending a request message to the agent side based on the multi-dimensional data, and the agent side forwards the request message to the server side after receiving it; A query unit for querying sensitive information in the server side, returning the result set message containing sensitive information to the agent side, and the agent side queries the record closest to the current time from the session time table and obtains the corresponding customer name according to the session ID in the record; A desensitization unit for performing data desensitization based on the agent side obtaining the desensitization scheme corresponding to the current customer name from the desensitization configuration table according to the corresponding customer name, assembling the desensitized data to form a result set message, and completing the dynamic desensitization encryption of the data.

8. The digital marketing system according to claim 7, wherein The capture module includes: A processing unit for performing deduplication, filling, and outlier checking on encrypted multi-dimensional data to complete data cleaning, and performing standardization, discretization, and stratification processing on the cleaned data to complete data transformation; A visualization unit for batch loading the transformed data into a distributed storage area based on an ETL tool, separately extracting and integrating the features corresponding to the multi-dimensional data, and performing visualization processing on the integrated multi-dimensional data to obtain the key information in the multi-dimensional data and the visualization images corresponding to the dynamic features.

9. A storage medium, on which a computer program is stored, characterized in that, When the program is executed by a processor, it implements the digital marketing method according to any one of claims 1 to 5.

10. A computer, comprising a memory, a processor, and a computer program stored on the memory and executable on the processor, characterized in that, When the processor executes the computer program, it implements the digital marketing method according to any one of claims 1 to 5.

Images