Electronic commerce data processing system and processing method thereof

Abstract

The invention discloses an electronic commerce data processing system and a processing method thereof, and relates to the technical field of electronic commerce data processing. The objective of the invention is to solve the problems that certain details of an image may be lost due to equal-proportion scaling, the accuracy and integrity of image information are affected especially when a complex or high-resolution image is processed, and non-image information cannot be effectively processed due to the fact that data of an electronic commerce platform further comprises a large amount of non-image information. The electronic commerce data processing system comprises a data acquisition unit, a data processing unit, a data security protection unit and a data visualization unit, data are collected from multiple sources, a high-quality data set is generated through preprocessing and integration, personalized demand data and purchase ability data are generated according to user information and a classification result of the data set, the security and privacy of the data are ensured, a user is helped to visually understand and analyze the data, comprehensive management and utilization of electronic commerce data are achieved, and the user experience is improved. And the service efficiency and the user experience are improved.

Description

Technical Field

[0001] The present invention relates to the technical field of e-commerce data processing, and in particular to an e-commerce data processing system and a processing method thereof. Background Art

[0002] With the rapid development of e-commerce, a large amount of transaction data is generated. How to efficiently process this data and extract valuable information is crucial to the operational decisions and user experience of the e-commerce platform. Regarding e-commerce data processing, the Chinese patent with publication number: CN115564643A discloses a data processing system and application of an e-commerce platform, including an information acquisition module, an information extraction module, an information processing module, an information storage module, and a data transmission module; the image information is obtained through the information acquisition module, and the image information is transmitted to the information extraction module through the data transmission module to extract the image information, and the extracted image information is transmitted to the information processing module through the data transmission module to adjust the resolution, and the adjusted image information is transmitted to the information storage module through the data transmission module. This patent realizes the simultaneous processing of multiple image information and improves the image processing rate by setting multiple image data processing units.

[0003] Although the above patent improves the efficiency of conveying correct image information by performing proportional scaling, it still has the following problems: Proportional scaling may cause loss of some image details, especially when processing complex or high-resolution images, affecting the accuracy and completeness of image information. In addition, the data of e-commerce platforms also include a large amount of non-image information, which cannot be effectively processed. Summary of the invention

[0004] The purpose of the present invention is to provide an e-commerce data processing system and a processing method thereof, to provide more valuable information for business decisions and operations by classifying and processing the generated e-commerce data, to generate e-commerce data with business value, to provide important decision support for the operation of the e-commerce platform, to help enterprises better understand user needs and market trends, to improve business efficiency and user experience, so as to solve the problems raised in the above-mentioned background technology.

[0005] To achieve the above object, the present invention provides the following technical solutions: E-commerce data processing system, including: A data acquisition unit is used to collect data from multiple data sources such as e-commerce platforms, user devices, and payment systems, pre-process the collected data, and integrate the data from different sources after pre-processing to generate a data set; A data processing unit, used to obtain the data set, generate personalized demand data and purchasing power data information corresponding to the user based on the user login information of the mobile terminal information, and at the same time, process and analyze the classification results of the data set, and generate e-commerce data based on the analysis results of the image data and text data; Data security protection unit, used to protect e-commerce data, encrypt data storage and transmission; The data visualization unit is used to visualize the processed data and generate push results based on the search information input by the user and the personalized demand data and purchasing power data information of the data processing unit.

[0006] Furthermore, the data acquisition unit includes: The data collection module is used to regularly collect and extract user behavior data, transaction data, and product information data from different data sources, and integrate the collected data based on different data sources to generate a data set; The data preprocessing module is used to preprocess the raw data in the collected data set, remove invalid data and duplicate data, convert the data format of the collected data according to the acquired data receiving format and record it, and generate the target data transmission file for each data set; The data transmission module is used to transmit the target data transmission file based on the communication link, obtain the data transmission feedback of the data set in real time, read the feedback data, and determine whether there is any omission in the data in the target data transmission file.

[0007] Furthermore, the data processing unit comprises: A data classification module, used to obtain a data set based on the communication link, and classify the data in the data set according to data type; The data analysis module is used to perform association mining and cluster analysis on the classified data based on the data classification results, determine the rules and trends of the data, and generate e-commerce data based on the analysis results; The classification processing module is used to match corresponding processing strategies according to different categories of data, including building user profiles for user behavior data, and performing order analysis and product recommendation operations on transaction data.

[0008] Furthermore, the data analysis module is specifically: User behavior analysis, obtaining multiple behavior patterns of users on the platform and the behavior characteristics under each behavior pattern, and performing correlation analysis based on the behavior patterns and their behavior characteristics to obtain analysis results; determining the user's interest preferences, demand characteristics and purchasing needs based on the analysis results; Sales trend forecasting: obtaining historical sales data and market environment data of each product on the platform, and analyzing platform products based on historical sales data and market environment, determining the sales patterns and trends of the platform, and conducting inventory management, promotional decisions and market positioning based on the analysis results; Supply chain optimization: monitor and predict each link in the supply chain, analyze supplier delivery time, product quality and sales data, screen and identify suppliers, and optimize procurement strategies for the platform; Risk control: discover and predict potential risk points by analyzing users' credit ratings, transaction records and complaint data, and take corresponding risk control measures for different risk points.

[0009] Furthermore, the classification processing module is specifically: Obtaining classification results of the data classification module, extracting picture data based on the classification results, performing target image capture on the picture data based on product features and user gesture features, and extracting product features and key information of user gestures from the image data; Extract text data based on the classification results, classify a known text in the text based on the constructed sentiment dictionary, and perform sentiment analysis on the expressions in the text; Performing word segmentation on the type sentences in the text according to the part-of-speech statistical features, extracting a number of words, and performing de-duplication processing on the extracted words to obtain keywords of the text; Through cluster analysis, user data is grouped according to similarities, the characteristics and needs of user groups are determined, user data is labeled according to groups, and classified based on the labels.

[0010] Furthermore, the data security protection unit includes: A data integrity verification module, for reading adjacent data of the missing data when the data transmission module determines that the transmission data is missing, and supplementing the missing data according to a preset similarity algorithm, and generating a complete target data transmission file based on the supplementation result; An image compression module is used to construct an image compression framework, insert an enhanced multi-scale residual block and a simplified attention module into the framework, and adopt a two-stage training method to perform stable training, extract image class data from the classification results and input the compressed image generated by the image compression framework; The data quality management module is used to monitor the data quality of the complete target data transmission file in real time, and issue an early warning when the data quality is lower than a preset quality standard.

[0011] Furthermore, the data visualization unit includes: The user management module is used to manage the permissions of users who use the data visualization function, record and manage users' personal information, and provide an interactive interface for users; The search module is used to provide multiple query methods, allowing users to filter data according to the required conditions and display the query results to users in a clear and intuitive way; The display module is used to display the processed data in the form of graphics, images, and charts in real time, and the user can operate the displayed data.

[0012] Furthermore, the data analysis module includes: Identification module for: Obtain classified data and generate several classification sets; Based on the data classification results, determine the entity dictionary for each category; Perform matching processing on several classification sets based on entity dictionaries of corresponding categories respectively, and determine several entity recognition results of each classification set; The first determination module is used to: Perform forward expansion and backward expansion on each entity recognition result to generate a logical sequence of thinking; Make logical judgments on each thinking logic sequence, connect them based on the judgment results, and obtain a thinking logic relationship diagram; Determine the central node based on the thinking logic relationship diagram, and determine the central theme of the corresponding classification set based on the central node; Generate a module to define the underlying structure of the sample center theme, define the relationship between each sample center theme, and abstract the analysis operation to form the basic architecture model of the system; set up an analysis engine to generate analysis tasks based on the sample center theme and support the analysis requirements of different analysis business application scenarios; set up the core layer to develop extension plug-ins based on the distributed architecture, and encapsulate the changes in business function extensions in the plug-ins; according to the relationship between each sample center theme, set it as a business layer in the form of a plug-in; define the data interface, and according to the analysis capabilities provided by the core layer, open each business function in the business layer in the form of a standard interface; set up business calculation logic to develop analysis rules, and verify and manage them through the core layer. The core layer drives the plug-in to run in the form of events to generate an association mining model; The second determination module is used to input the central theme of each classification set into the association mining model to determine the association relationship between the classified data.

[0013] Furthermore, the data processing unit includes: The third determination module is used to analyze the user login information of the mobile terminal information, determine the indicators involved in the user quantification framework and the purchase tendency value triggered by the indicators, and determine the user's purchase tendency;

[0014] in, The user's purchasing tendency; The purchase tendency value triggered by the indicator corresponding to the i1th indicator involved in the user quantitative framework; i1 is the value in the q set; q is the indicator of preference for purchase in the user quantitative framework; e is a natural constant; is the value of the i1th indicator of the commodity pool; The purchase tendency value triggered by the indicator corresponding to the i2th indicator involved in the user quantitative framework; The value of i2 is the value in the W set; W is the indicator that tends to not buy in the user quantitative framework; is the value of the i2th indicator of the commodity pool; The fourth determination module is used to make a purchase prediction for the user according to the user's purchase tendency, the prediction time is a preset time, determine the values ​​corresponding to the M predicted product viewing indicators, and determine the user's purchasing power data information according to the values ​​corresponding to the M predicted product viewing indicators:

[0015] in, The purchasing power data information of users; is the value corresponding to the s1th viewed product indicator; is the average value of the s1th viewed commodity index in the commodity pool within the preset time; M is the number of viewed commodity indexes.

[0016] The present invention provides another technical solution, an e-commerce data processing method, comprising the following steps: Step 1: Data acquisition, collecting data from multiple data sources, such as e-commerce platforms, user devices, and payment systems, including user behavior data, transaction data, and product information data; Step 2: Feature extraction: Process and analyze the acquired data set. Use corresponding algorithms and tools to extract features according to different data types. For image data, use image processing technology to extract key information such as product features and user gestures. For text data, perform text analysis, including sentiment analysis and keyword extraction, to explore users’ real needs and intentions. Step 3: Personalized data processing, based on the user login information of the mobile terminal information, generating the user's corresponding personalized demand data and purchasing power data information; Step 4: Data security protection: Protect e-commerce data to ensure that the data is not accessed, used or leaked without authorization; Step 5: Data visualization and push, visualize the processed data, and generate push results based on the search information, personalized demand data, and purchasing power data entered by the user.

[0017] Compared with the prior art, the present invention has the following beneficial effects: 1. The e-commerce data generated through classification processing provides more valuable information for business decision-making and operation. Through data analysis, we can better understand information such as user behavior, market demand and product information, generate e-commerce data with business value, provide important decision-making support for the operation of e-commerce platforms, help enterprises better understand user needs and market trends, and improve business efficiency and user experience; 2. Through the component image compression framework, a two-stage multi-scale residual block is adopted to enhance the reconstruction effect of the generated image to improve the quality of the generated image and avoid the problem that proportional scaling may cause the loss of some details of the image. Through the collaborative work of the data integrity verification module, the image compression module and the data quality management module, comprehensive security protection of e-commerce data is achieved to ensure the integrity, accuracy and high quality of the data and avoid the negative impact of low-quality data on business decisions. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] Figure 1 It is a module diagram of the e-commerce data processing system of the present invention; Figure 2 This is a schematic diagram of a data security protection unit module of the present invention; Figure 3 The figure is a flow chart of the e-commerce data processing method of the present invention. DETAILED DESCRIPTION

[0019] 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.

[0020] In order to solve the technical problem that in the prior art, proportional scaling may cause the loss of some details of the image, especially when processing complex or high-resolution images, affecting the accuracy and completeness of the image information, and the data of the e-commerce platform also includes a large amount of non-image information, and the non-image information cannot be effectively processed, please refer to Figure 1-2 , this embodiment provides the following technical solutions: E-commerce data processing system, including: A data acquisition unit is used to collect data from multiple data sources such as e-commerce platforms, user devices, and payment systems, pre-process the collected data, and integrate the data from different sources after pre-processing to generate a data set; The data acquisition unit comprises: The data collection module is used to regularly collect and extract user behavior data, transaction data, and product information data from different data sources, and integrate the collected data based on different data sources to generate a data set; The data preprocessing module is used to preprocess the raw data in the collected data set, remove invalid data and duplicate data, convert the data format of the collected data according to the acquired data receiving format and record it, and generate the target data transmission file for each data set; A data transmission module is used to transmit the target data transmission file based on a communication link, obtain data transmission feedback of the data set in real time, read the feedback data, and determine whether there is any omission in the data in the target data transmission file; A data processing unit is used to obtain the data set, generate personalized demand data and purchasing power data information corresponding to the user based on the user login information of the mobile terminal information, and at the same time, process and analyze the classification results of the data set, generate e-commerce data based on the analysis results of the image data and text data, process and extract image data, extract key information such as product features and user gestures, perform text analysis on text data, perform sentiment analysis, keyword extraction and other operations on user comments and feedback, so as to explore the real needs and intentions of users, and use distributed computing and memory computing technologies to efficiently calculate and store large-scale data; Data security protection unit is used to protect e-commerce data, encrypt data for storage and transmission, ensure that data is not accessed, used or leaked without authorization, and use encryption technology, access control mechanism and other security measures to protect data security. At the same time, data desensitization technology is used to process sensitive data to protect user privacy and business secrets. The data visualization unit is used to visualize the processed data. At the same time, it generates push results based on the search information input by the user and the personalized demand data and purchasing power data information of the data processing unit, so that users can understand the data and analysis results more intuitively. It provides various charts, reports and other visualization tools to help users better understand the data and gain insights into business trends. In this embodiment, the data visualization unit includes: The user management module is used to manage the permissions of users who use the data visualization function. Users with different permissions can access and operate different data scopes and functions. It records and manages users' personal information, such as login accounts, roles, preferences, etc., in order to provide users with personalized data display and functional experience, and provides users with an interactive interface so that users can easily query, operate and manage data; The search module is used to provide multiple query methods, such as keyword query, condition query, etc., so that users can quickly locate the required data. Users can filter the data according to the required conditions to obtain the data subset that meets the requirements, and the query results are displayed to users in a clear and intuitive way; The display module is used to display the processed data in the form of graphics, images, and charts in real time. Users can operate the displayed data to synchronize the displayed content with the data source. It provides interactive functions so that users can operate the displayed data, such as zooming, rotating, filtering, etc., providing users with an intuitive and easy-to-use interface to help users better understand and analyze e-commerce data and make more effective business decisions.

[0021] In this embodiment, the data acquisition unit collects data from multiple sources, and generates a high-quality data set after preprocessing and integration. The data processing unit generates personalized demand data and purchasing power data based on the classification results of the user information and the data set, and performs efficient calculation and storage at the same time. The data security protection unit ensures the security and privacy of the data. The data visualization unit provides a user-friendly interface to help users intuitively understand and analyze data. Through the collaborative work of multiple modules, comprehensive management and utilization of e-commerce data is achieved, thereby improving business efficiency and user experience.

[0022] In this embodiment, the data processing unit includes: A data classification module, used to obtain a data set based on the communication link, and classify the data in the data set according to data type; The data analysis module is used to perform association mining and cluster analysis on the classified data based on the data classification results, determine the rules and trends of the data, and generate e-commerce data based on the analysis results; In this embodiment, the data analysis module is specifically: User behavior analysis, obtaining multiple behavior patterns of users on the platform and the behavior characteristics under each behavior pattern, and performing correlation analysis based on the behavior patterns and their behavior characteristics to obtain analysis results; determining the user's interest preferences, demand characteristics and purchasing needs based on the analysis results; Sales trend forecasting: obtaining historical sales data and market environment data of each product on the platform, and analyzing platform products based on historical sales data and market environment, determining the sales patterns and trends of the platform, and conducting inventory management, promotional decisions and market positioning based on the analysis results; Supply chain optimization: monitor and predict each link in the supply chain, analyze supplier delivery time, product quality and sales data, screen and identify suppliers, and optimize procurement strategies for the platform. Through supply chain optimization, the operational efficiency and customer satisfaction of the e-commerce platform can be improved, and operating costs and risks can be reduced; Risk control: by analyzing the user's credit rating, transaction records and complaint data, we can discover and predict potential risk points, such as fraud, malicious attacks and credit risks, and take corresponding risk control measures for different risk points, such as establishing a risk early warning mechanism, implementing security measures and formulating a credit evaluation system. Through risk control, we can protect the interests of the e-commerce platform and users and improve the reliability and security of the platform. The classification processing module is used to match the corresponding processing strategies according to different categories of data, including building user profiles for user behavior data, and performing order analysis and product recommendation operations on transaction data; In this embodiment, the classification processing module is specifically: Obtain the classification results of the data classification module, extract image data based on the classification results, perform target image capture on the image data based on product features and user gesture features, and extract product features and key information of user gestures from the image data; extract product features such as shape, texture, color, and other features, as well as information such as the action, direction, and speed of the user gesture from the image; Extract text data based on the classification results, classify a known text in the text based on the constructed sentiment dictionary, and perform sentiment analysis on the expressions in the text; Performing word segmentation operations on type sentences in the text according to part-of-speech statistical features, extracting a number of words, and performing de-duplication processing on the extracted words to obtain keywords of the text; extracting meaningful information such as keywords, phrases, sentences, etc. from the text, and analyzing its sentiment tendency (positive, negative or neutral) and importance; Through cluster analysis, user data is grouped according to similarities to determine the characteristics and needs of user groups. User data is labeled according to groups and classified based on the labels to predict user purchasing behavior, preferences and other information.

[0023] In this embodiment, through classification processing, more accurate e-commerce data can be generated to provide more valuable information for business decisions and operations; statistics, association rule mining, cluster analysis and other operations are performed on various types of data to discover hidden patterns and trends in the data. Through data analysis, information on user behavior, market demand, and product information can be better understood, and e-commerce data with business value can be generated, providing important decision-making support for the operation of e-commerce platforms, helping companies to better understand user needs and market trends, and improve business efficiency and user experience.

[0024] In this embodiment, the data security protection unit includes: A data integrity verification module, for reading adjacent data of the missing data when the data transmission module determines that the transmission data is missing, and supplementing the missing data according to a preset similarity algorithm, and generating a complete target data transmission file based on the supplementation result; An image compression module is used to construct an image compression framework, insert an enhanced multi-scale residual block and a simplified attention module into the framework, and adopt a two-stage training method to perform stable training, extract image class data from the classification results and input the compressed image generated by the image compression framework; The data quality management module is used to monitor the data quality of the complete target data transmission file in real time, and issue an early warning when the data quality is lower than a preset quality standard.

[0025] In this embodiment, a two-stage multi-scale residual block is used to enhance the reconstruction effect of the generated image through the component image compression framework to improve the quality of the generated image. The loss can also be reduced under moderate complexity, avoiding the problem that proportional scaling may cause loss of certain details of the image. Through the collaborative work of the data integrity verification module, the image compression module and the data quality management module, comprehensive security protection of e-commerce data is achieved to ensure the integrity, accuracy and high quality of the data and avoid the negative impact of low-quality data on business decisions.

[0026] Furthermore, the data analysis module includes: Identification module for: Obtain classified data and generate several classification sets; Based on the data classification results, determine the entity dictionary for each category; Perform matching processing on several classification sets based on entity dictionaries of corresponding categories respectively, and determine several entity recognition results of each classification set; The first determination module is used to: Perform forward expansion and backward expansion on each entity recognition result to generate a logical sequence of thinking; Make logical judgments on each thinking logic sequence, connect them based on the judgment results, and obtain a thinking logic relationship diagram; Determine the central node based on the thinking logic relationship diagram, and determine the central theme of the corresponding classification set based on the central node; Generate a module to define the underlying structure of the sample center theme, define the relationship between each sample center theme, and abstract the analysis operation to form the basic architecture model of the system; set up an analysis engine to generate analysis tasks based on the sample center theme and support the analysis requirements of different analysis business application scenarios; set up the core layer to develop extension plug-ins based on the distributed architecture, and encapsulate the changes in business function extensions in the plug-ins; according to the relationship between each sample center theme, set it as a business layer in the form of a plug-in; define the data interface, and according to the analysis capabilities provided by the core layer, open each business function in the business layer in the form of a standard interface; set up business calculation logic to develop analysis rules, and verify and manage them through the core layer. The core layer drives the plug-in to run in the form of events to generate an association mining model; The second determination module is used to input the central theme of each classification set into the association mining model to determine the association relationship between the classified data.

[0027] The working principle of the above technical solution: In this embodiment, the entity dictionary includes: the basic dictionary is constructed based on the metadata information and data information of the data table to which the data belongs, and the proprietary dictionary includes dimension categories and dimension value categories; the dimension categories correspond to the field names in the data table, and the dimension value categories correspond to the specific values ​​of the fields in the data table; the overall dictionary includes time categories, unit categories, and intent categories; the intent category represents the central idea and semantic representation of the data.

[0028] In this embodiment, matching processing is performed on several classification sets based on entity dictionaries of corresponding categories, and data matching is performed based on entity dictionaries of corresponding categories, so as to improve the accuracy of data recognition and obtain accurate entity recognition results.

[0029] In this embodiment, forward expansion and backward expansion are performed on each entity recognition result to generate a thinking logic sequence, including forward expansion and backward expansion of each entity. If a complete sentence is obtained for semantic representation, a thinking logic sequence is generated.

[0030] In this embodiment, a logical judgment is performed on each thinking logic sequence, and two thinking logic sequences that are logically consistent are connected, and the connections are processed in sequence to obtain a thinking logic relationship diagram; the various logical relationships in the classified data are extracted and the corresponding schematic diagrams are generated.

[0031] In this embodiment, the central node is determined based on the thinking logic relationship diagram. The central node is a thinking logic relationship diagram generated around a certain node. The theme corresponding to the central node serves as the central theme of the classification set to achieve accurate data extraction and organization.

[0032] In this embodiment, the underlying structure of the sample center theme is defined, the relationship between each sample center theme is defined, and the analysis operation is abstracted to form the basic architecture model of the system, including: in the underlying structure of the sample center theme, define the sample: refers to a specific data point or instance, which can be text, image, audio or other forms of data. Define the feature: describe the attribute or characteristic of the sample, which can be each dimension or attribute of the sample. Define the label: used to identify the category or attribute of the sample. Define the data set: a collection of multiple samples, which can be a labeled data set (each sample has a corresponding label) or an unlabeled data set. Define the sample center theme: it is a summary of the clustering or classification results of the samples in the data set. It can be understood that each sample center theme represents a cluster or category, which has certain commonalities and similarities. In the relationship between the sample center themes, it includes clustering relationship: the sample center theme can cluster similar samples together through the clustering algorithm to form different clusters. There can be a relationship with low similarity between different clusters. Classification relationship: the sample center theme can classify the sample into different categories through the classification algorithm, and each sample center theme represents a category. There can be an inclusion relationship or an exclusion relationship between different categories. Similarity relationship: The similarity between different sample central themes can be measured by calculating similarity indicators. Similarity indicators can be based on the distance between samples in feature space, similarity measurement methods (such as cosine similarity) or other related indicators. For the abstraction of analysis operations and the basic architecture model of the system, it includes feature engineering: feature extraction, selection or construction of samples to better describe the properties and characteristics of samples. Clustering algorithm and classification algorithm: algorithms used to cluster and classify samples. Common clustering algorithms include K-means, hierarchical clustering, etc., and common classification algorithms include decision trees, support vector machines, etc. Similarity calculation: methods or algorithms used to calculate similarity indicators between samples or sample central themes. Model evaluation and optimization: performance evaluation and optimization of clustering models or classification models to improve the accuracy and generalization ability of the model. Based on the above elements and relationships, we can build a system infrastructure model for processing the analysis task of sample central themes.

[0033] In this embodiment, an analysis engine is set to facilitate task management: support the creation, scheduling and management of analysis tasks, including the creation, editing, deletion and scheduling functions of tasks. Data preprocessing: provide functions such as data cleaning, data conversion, missing value processing, standardization, etc. to ensure data quality and consistency. Feature engineering: provide various feature extraction, selection and construction functions to help users better describe the attributes and characteristics of samples. Clustering and classification: integrate common clustering algorithms and classification algorithms, including K-means, DBSCAN, decision trees, neural networks, etc., so as to cluster and classify samples according to user needs. Similarity calculation: provide various similarity calculation methods and algorithms, including distance measurement methods, cosine similarity, etc., to help users calculate the similarity between samples or sample central themes. Visual analysis: support visual analysis functions, including data visualization, clustering result visualization, classification result visualization, etc., so that users can intuitively understand and interpret the analysis results. Model evaluation and optimization: provide model evaluation and optimization functions, including model performance evaluation, parameter tuning, feature selection, etc., to improve the accuracy and generalization ability of the model. Scalability and flexibility: With good scalability and flexibility, new algorithms, functions and modules can be added according to user needs to adapt to different analytical business application scenarios. The analysis engine can meet the needs of different analytical business application scenarios and provide users with powerful analysis capabilities and support. Users can use the analysis engine to perform data processing, feature engineering, clustering and classification and other analysis operations according to specific analysis tasks and needs, and obtain accurate analysis results through visual analysis and model evaluation optimization.

[0034] In this embodiment, the core layer is set to develop extension plug-ins based on a distributed architecture, and the changes in the extension of business functions are encapsulated in the plug-ins; according to the relationship between the subject of each sample center, the plug-in is set as the business layer, including: Design of the core layer: The core layer can adopt a distributed architecture, use a group of distributed nodes to work together, and process large-scale data and complex computing tasks. Distributed computing frameworks such as Apache Hadoop, Apache Spark, etc. can be used to ensure the scalability and high availability of the system. Development of extension plug-ins: In order to support the expansion of business functions, a group of extension plug-ins can be designed and developed to encapsulate different business functions inside the plug-in. Each plug-in can provide specific functions, such as different clustering algorithms, classification algorithms, or similarity calculation methods. The plug-in can be loaded and unloaded in a pluggable manner to meet different business needs. Plug-in management: A plug-in management module can be designed to manage the loading, unloading, updating and configuration of plug-ins. The module can provide a user interface or a command line interface to facilitate the user to manage the installation and use of plug-ins. Design of the business layer: According to the relationship between different sample center themes, the plug-in can be used as a component of the business layer. In the business layer, appropriate plug-ins are used to implement specific analysis tasks, such as clustering tasks, classification tasks, etc. The business layer can be configured and expanded according to needs, and different plug-ins can be combined to meet the analysis needs of specific business scenarios. Through the above design, the distributed architecture of the core layer can be realized, and the changes in business functions can be encapsulated in the plug-ins by extending the plug-ins. Such a design can effectively improve the flexibility and scalability of the system, allowing specific functions to be customized and integrated according to business needs. At the same time, the configuration and combination of the business layer can be dynamically adjusted according to the relationship between the sample center themes to meet the analysis needs in different scenarios.

[0035] In this embodiment, a data interface is defined, and each business function in the business layer is opened in the form of a standard interface according to the analysis capability provided by the core layer; business calculation logic development and analysis rules are set, and verification and management are performed through the core layer. The core layer drives the plug-in to run in the form of events to generate an association mining model, including: definition of data interface: according to the needs of business functions, appropriate data interfaces can be defined for the transmission of input and output data. These interfaces may include data input interfaces, data output interfaces, parameter configuration interfaces, etc. The definition of interfaces needs to be designed according to specific business needs and data formats to ensure the correct transmission and processing of data. Business functions are opened in the form of interfaces: in the business layer, each business function is opened according to the defined interface. This means that the module that implements each business function needs to implement the corresponding interface to receive input data, perform calculations, generate output data, and return data to the core layer. Development and analysis rules of business calculation logic: in the business layer, business calculation logic and analysis rules are developed. These rules can be defined according to specific business needs and analysis goals to guide data processing, analysis and mining. Business calculation logic can include steps such as feature extraction, data conversion, model training, and related parameter configuration and tuning. Verification and management of the core layer: The core layer can verify and manage the business calculation logic and analysis rules. The core layer can verify the legality and correctness of the business rules and provide corresponding management functions, such as configuration management, model management, event-driven management, etc. The core layer can also optimize the performance and resource management of the business calculation logic to improve the efficiency and accuracy of the analysis. Event-driven plug-in operation: The core layer can execute plug-in operation in an event-driven manner. When the core layer receives a suitable event or request, it triggers the corresponding plug-in operation according to the configuration and definition. This method can realize the dynamic scheduling and management of plug-ins, generate association mining models, and update and maintain models as needed. Through the above settings, efficient interaction and management between the business layer and the core layer can be achieved. Business functions are opened according to the defined standard interface. The core layer can verify and manage the business calculation logic and drive the plug-in operation in an event-driven manner to realize the generation and update of association mining models. Such a design can improve the flexibility and maintainability of the system and support the customization and management of different business needs and analysis rules.

[0036] The beneficial effects of the above technical solution are: realizing entity recognition of classified data, obtaining entity recognition results, making logical judgments and associations based on the entity recognition results, and accurately determining the central theme of the classification set; inputting the central theme of each classification set into the association mining model to accurately determine the association relationship between the classified data.

[0037] Furthermore, the data processing unit includes: The third determination module is used to analyze the user login information of the mobile terminal information, determine the indicators involved in the user quantification framework and the purchase tendency value triggered by the indicators, and determine the user's purchase tendency;

[0038] in, The user's purchasing tendency; The purchase tendency value triggered by the indicator corresponding to the i1th indicator involved in the user quantitative framework; i1 is the value in the q set; q is the indicator of preference for purchase in the user quantitative framework; e is a natural constant; is the value of the i1th indicator of the commodity pool; The purchase tendency value triggered by the indicator corresponding to the i2th indicator involved in the user quantitative framework; The value of i2 is the value in the W set; W is the indicator that tends to not buy in the user quantitative framework; is the value of the i2th indicator of the commodity pool; The fourth determination module is used to make a purchase prediction for the user according to the user's purchase tendency, the prediction time is a preset time, determine the values ​​corresponding to the M predicted product viewing indicators, and determine the user's purchasing power data information according to the values ​​corresponding to the M predicted product viewing indicators:

[0039] in, The purchasing power data information of users; is the value corresponding to the s1th viewed product indicator; is the average value of the s1th viewed commodity index in the commodity pool within the preset time; M is the number of viewed commodity indexes.

[0040] The working principle and beneficial effects of the above technical solution: Analyze the user login information of the mobile terminal information to determine the indicators involved in the user quantification framework, including browsing time, browsing history, search history, etc. The q set is the indicator of interest. The W set is the indicator of no interest. Accurately determine the user's purchasing tendency; the fourth determination module is used to make purchase predictions for the user based on the user's purchasing tendency, the prediction time is a preset time, and determine the values ​​corresponding to the predicted M product viewing indicators. According to the values ​​corresponding to the predicted M product viewing indicators, the user's purchasing power data information is determined: the product viewing indicators include the viewing time, style, inquiry record, function, and usage method of the product. It is convenient to accurately determine the user's purchasing power data information, facilitate the subsequent corresponding processing and analysis, and generate e-commerce data.

[0041] See also Figure 3The present invention provides an e-commerce data processing method, comprising the following steps: Step 1: Data acquisition, collecting data from multiple data sources, such as e-commerce platforms, user devices, and payment systems, including user behavior data, transaction data, and product information data; Step 2: Feature extraction: Process and analyze the acquired data set. Use corresponding algorithms and tools to extract features according to different data types. For image data, use image processing technology to extract key information such as product features and user gestures. For text data, perform text analysis, including sentiment analysis and keyword extraction, to explore users’ real needs and intentions. Step 3: Personalized data processing: Based on the user login information of the mobile terminal information, the corresponding personalized demand data and purchasing power data information of the user are generated, which reflects the user's interests, preferences and consumption capacity, which is crucial for the accurate recommendation and personalized service of the e-commerce platform; Step 4: Data security protection: Protect e-commerce data to ensure that the data is not accessed, used or leaked without authorization; Step 5: Data visualization and push, visualize the processed data to help users better understand the data and analysis results. At the same time, generate push results based on the search information, personalized demand data and purchasing power data information entered by the user.

[0042] The above description is only a preferred specific implementation manner of the present invention, but the protection scope of the present invention is not limited thereto. Any technician familiar with the technical field can make equivalent replacements or changes according to the technical scheme and inventive concept of the present invention within the technical scope disclosed by the present invention, which should be covered by the protection scope of the present invention.

Claims

1. An electronic commerce data processing system, characterized in that: include: A data acquisition unit is used to collect data from multiple data sources such as e-commerce platforms, user devices, and payment systems, pre-process the collected data, and integrate the data from different sources after pre-processing to generate a data set; A data processing unit, used to obtain the data set, generate personalized demand data and purchasing power data information corresponding to the user based on the user login information of the mobile terminal information, and at the same time, process and analyze the classification results of the data set, and generate e-commerce data based on the analysis results of the image data and text data; Data security protection unit, used to protect e-commerce data, encrypt data storage and transmission; The data visualization unit is used to visualize the processed data and generate push results based on the search information input by the user and the personalized demand data and purchasing power data information of the data processing unit.

2. The electronic commerce data processing system according to claim 1, wherein: The data acquisition unit comprises: The data collection module is used to regularly collect and extract user behavior data, transaction data, and product information data from different data sources, and integrate the collected data based on different data sources to generate a data set; The data preprocessing module is used to preprocess the raw data in the collected data set, remove invalid data and duplicate data, convert the data format of the collected data according to the acquired data receiving format and record it, and generate the target data transmission file for each data set; The data transmission module is used to transmit the target data transmission file based on the communication link, obtain the data transmission feedback of the data set in real time, read the feedback data, and determine whether there is any omission in the data in the target data transmission file.

3. The electronic commerce data processing system according to claim 2, wherein: Data processing unit, including: A data classification module, used to obtain a data set based on the communication link, and classify the data in the data set according to data type; The data analysis module is used to perform association mining and cluster analysis on the classified data based on the data classification results, determine the rules and trends of the data, and generate e-commerce data based on the analysis results; The classification processing module is used to match corresponding processing strategies according to different categories of data, including building user profiles for user behavior data, and performing order analysis and product recommendation operations on transaction data.

4. The electronic commerce data processing system according to claim 3, wherein: Data analysis module, specifically: User behavior analysis, obtaining multiple behavior patterns of users on the platform and the behavior characteristics under each behavior pattern, and performing correlation analysis based on the behavior patterns and their behavior characteristics to obtain analysis results; determining the user's interest preferences, demand characteristics and purchasing needs based on the analysis results; Sales trend forecasting: obtaining historical sales data and market environment data of each product on the platform, and analyzing platform products based on historical sales data and market environment, determining the sales patterns and trends of the platform, and conducting inventory management, promotional decisions and market positioning based on the analysis results; Supply chain optimization: monitor and predict each link in the supply chain, analyze supplier delivery time, product quality and sales data, screen and identify suppliers, and optimize procurement strategies for the platform; Risk control: discover and predict potential risk points by analyzing users' credit ratings, transaction records and complaint data, and take corresponding risk control measures for different risk points.

5. The electronic commerce data processing system according to claim 4, characterized in that: Classification processing module, specifically: Obtaining classification results of the data classification module, extracting picture data based on the classification results, performing target image capture on the picture data based on product features and user gesture features, and extracting product features and key information of user gestures from the image data; Extract text data based on the classification results, classify a known text in the text based on the constructed sentiment dictionary, and perform sentiment analysis on the expressions in the text; Performing word segmentation on the type sentences in the text according to the part-of-speech statistical features, extracting a number of words, and performing de-duplication processing on the extracted words to obtain keywords of the text; Through cluster analysis, user data is grouped according to similarities, the characteristics and needs of user groups are determined, user data is labeled according to groups, and classified based on the labels.

6. The electronic commerce data processing system according to claim 5, characterized in that: Data security protection unit, including: A data integrity verification module, for reading adjacent data of the missing data when the data transmission module determines that the transmission data is missing, and supplementing the missing data according to a preset similarity algorithm, and generating a complete target data transmission file based on the supplementation result; An image compression module is used to construct an image compression framework, insert an enhanced multi-scale residual block and a simplified attention module into the framework, and adopt a two-stage training method to perform stable training, extract image class data from the classification results and input the compressed image generated by the image compression framework; The data quality management module is used to monitor the data quality of the complete target data transmission file in real time, and issue an early warning when the data quality is lower than a preset quality standard.

7. The electronic commerce data processing system according to claim 6, characterized in that: Data Visualization Unit, including: The user management module is used to manage the permissions of users who use the data visualization function, record and manage users' personal information, and provide an interactive interface for users; The search module is used to provide multiple query methods, allowing users to filter data according to the required conditions and display the query results to users in a clear and intuitive way; The display module is used to display the processed data in the form of graphics, images, and charts in real time, and the user can operate the displayed data.

8. The electronic commerce data processing system according to claim 3, wherein: The data analysis module comprises: Identification module for: Obtain classified data and generate several classification sets; Based on the data classification results, determine the entity dictionary for each category; Perform matching processing on several classification sets based on entity dictionaries of corresponding categories respectively, and determine several entity recognition results of each classification set; The first determination module is used to: Perform forward expansion and backward expansion on each entity recognition result to generate a logical sequence of thinking; Make logical judgments on each thinking logic sequence, connect them based on the judgment results, and obtain a thinking logic relationship diagram; Determine the central node based on the thinking logic relationship diagram, and determine the central theme of the corresponding classification set based on the central node; Generate a module to define the underlying structure of the sample center theme, define the relationship between each sample center theme, and abstract the analysis operation to form the basic architecture model of the system; set up an analysis engine to generate analysis tasks based on the sample center theme and support the analysis requirements of different analysis business application scenarios; set up the core layer to develop extension plug-ins based on the distributed architecture, and encapsulate the changes in business function extensions in the plug-ins; according to the relationship between each sample center theme, set it as a business layer in the form of a plug-in; define the data interface, and according to the analysis capabilities provided by the core layer, open each business function in the business layer in the form of a standard interface; set up business calculation logic to develop analysis rules, and verify and manage them through the core layer. The core layer drives the plug-in to run in the form of events to generate an association mining model; The second determination module is used to input the central theme of each classification set into the association mining model to determine the association relationship between the classified data.

9. The electronic commerce data processing system according to claim 1, wherein: The data processing unit comprises: The third determination module is used to analyze the user login information of the mobile terminal information, determine the indicators involved in the user quantification framework and the purchase tendency value triggered by the indicators, and determine the user's purchase tendency; in, The user's purchasing tendency; The purchase tendency value triggered by the indicator corresponding to the i1th indicator involved in the user quantitative framework; i1 is the value in the q set; q is the indicator of preference for purchase in the user quantitative framework; e is a natural constant; is the value of the i1th indicator of the commodity pool; The purchase tendency value triggered by the indicator corresponding to the i2th indicator involved in the user quantitative framework; The value of i2 is the value in the W set; W is the indicator that tends to not buy in the user quantitative framework; is the value of the i2th indicator of the commodity pool; The fourth determination module is used to make a purchase prediction for the user according to the user's purchase tendency, the prediction time is a preset time, determine the values ​​corresponding to the M predicted product viewing indicators, and determine the user's purchasing power data information according to the values ​​corresponding to the M predicted product viewing indicators: in, The purchasing power data information of users; is the value corresponding to the s1th viewed product indicator; is the average value of the s1th viewed commodity index in the commodity pool within the preset time; M is the number of viewed commodity indexes.

10. An e-commerce data processing method, based on the implementation of the e-commerce data processing system according to claim 9, characterized in that: The following steps are involved: Step 1: Data acquisition, collecting data from multiple data sources, such as e-commerce platforms, user devices, and payment systems, including user behavior data, transaction data, and product information data; Step 2: Feature extraction: Process and analyze the acquired data set. Use corresponding algorithms and tools to extract features according to different data types. For image data, use image processing technology to extract key information such as product features and user gestures. For text data, perform text analysis, including sentiment analysis and keyword extraction, to explore users’ real needs and intentions. Step 3: Personalized data processing, based on the user login information of the mobile terminal information, generating the user's corresponding personalized demand data and purchasing power data information; Step 4: Data security protection: Protect e-commerce data to ensure that the data is not accessed, used or leaked without authorization; Step 5: Data visualization and push, visualize the processed data, and generate push results based on the search information, personalized demand data, and purchasing power data entered by the user.

Images