An online shopping mall platform intelligent management method and system

Abstract

The application relates to the information technology field and particularly discloses an online shopping mall platform intelligent management method and system, which comprises the following steps: S1, a data acquisition system is constructed, platform operation data, user behavior data and external market data are acquired in real time; S2, historical sales data are acquired, a Prophet time series model is used to analyze the historical sales data, future sales trends and inventory demands are predicted, and dynamic replenishment suggestions are generated; S3, inventory management is optimized through a dynamic programming algorithm, and the optimal inventory threshold is calculated in combination with the external market data, and automatic replenishment triggering is realized; S4, a user portrait is established according to the user behavior data, and individualized coupon strategies are generated based on a collaborative filtering algorithm; S5, work orders and customer service are dynamically matched according to a multi-attribute decision algorithm; and S6, warranty information is stored, and the non-tamperability and traceability of construction time, material batch and technician ID data are ensured, so that the operation efficiency and user experience of the shopping mall platform are improved.

Description

Technical Field

[0001] This invention relates to the field of information technology, specifically to an intelligent management method and system for an online shopping mall platform. Background Technology

[0002] With the rapid development of e-commerce, online shopping platforms are facing increasingly complex challenges in areas such as product sales, customer service, and supply chain management. As the automotive consumer market flourishes, car wrapping films, as an important product for personalized car exterior decoration, are also experiencing a booming online market.

[0003] However, current online car wrapping film platforms face numerous management challenges. In inventory management, they struggle to accurately assess complex factors such as market demand fluctuations and supply chain uncertainties. In user marketing, they typically employ generic coupon distribution strategies without deeply understanding user needs and preferences. In work order processing, existing allocation methods fail to dynamically and effectively match work orders with customer service, resulting in low efficiency. Regarding warranty information management, traditional storage methods cannot guarantee data tamper-proofing and traceability, hindering rapid and accurate tracing of issues in the event of quality disputes.

[0004] Therefore, there is an urgent need for an innovative management method that can effectively integrate multi-source data, utilize advanced algorithms and technologies, solve the pain points in the management of the aforementioned online car color change film e-commerce platform, and improve platform operational efficiency, user experience, and market competitiveness. Summary of the Invention

[0005] In order to overcome the above-mentioned technical problems in the prior art, the present invention provides an intelligent management method and system for online shopping mall platforms. By integrating multi-dimensional data analysis, dynamic algorithm optimization and reliable data traceability, the intelligent management method for online shopping malls improves platform operation efficiency and user experience.

[0006] To achieve the above objectives, this invention provides an intelligent management method for an online shopping mall platform, comprising the following steps: S1: Constructing a data acquisition system to acquire platform operation data, user behavior data, and external market data in real time, wherein the external market data includes delivery cycle and material delay probability; S2: Acquiring historical sales data, analyzing the historical sales data using the Prophet time series model, predicting future sales trends and inventory demand, and generating dynamic replenishment suggestions; S3: Optimizing inventory management through dynamic programming algorithms and the dynamic replenishment suggestions, and calculating the optimal inventory threshold by combining the delivery cycle and the material delay probability to achieve automatic replenishment triggering; S4: Establishing user profiles based on the user behavior data, generating personalized coupon strategies based on collaborative filtering algorithms and user profiles, and updating the user profiles in real time; S5: Acquiring data from the work order management module, and dynamically matching work orders with customer service based on a multi-attribute decision algorithm; S6: Storing quality assurance information to ensure the immutability and traceability of construction time, material batches, and technician ID data.

[0007] Preferably, in step S2, generating dynamic replenishment suggestions specifically includes: acquiring weather data and supply chain fluctuation information, integrating the weather data and the supply chain fluctuation information, and dynamically adjusting the replenishment priority according to the gradient boosting tree; embedding a visual early warning interface into the data overview module of the online mall platform, acquiring inventory data in real time and generating inventory health, acquiring replenishment logs in real time, and displaying the inventory health and replenishment logs in real time on the visual early warning interface.

[0008] Preferably, step S5 specifically includes: establishing a customer service information database, which records customer service professional skills, real-time load, and historical satisfaction data; dynamically optimizing the work order allocation strategy by combining multi-attribute decision-making algorithms and reinforcement learning models to balance skill matching, response time, and load balancing; acquiring work order text and performing semantic classification on the work order text using the BERT model to identify the urgency of the work order and customer emotions.

[0009] Preferably, the dynamic optimization of the work order allocation strategy by combining multi-attribute decision-making algorithms and reinforcement learning models further includes: the state space defined by the reinforcement learning model is the real-time customer service load, the backlog of work orders, and the distribution of customer sentiment, and the action space is the work order allocation decision; the short-term reward function is set as work order resolution time and customer satisfaction score, and the long-term reward function is the customer service skill improvement rate and the overall work order resolution rate; the allocation logic is interpreted using the SHAP value interpretation algorithm to generate an interpretable allocation report.

[0010] Preferably, in step S6, storing warranty information specifically includes: setting up a private blockchain node, associating IoT device data of the construction store, and automatically uploading warranty operation records; providing a client blockchain query interface to support QR code verification of the entire warranty information chain.

[0011] Preferably, the method further includes: integrating AR augmented reality technology to provide a virtual color try-on function, supporting users to upload vehicle photos and simulate the color change film effect in real time; the virtual color try-on function includes: building a 3D vehicle model library to support the simulation of color change film effects for mainstream models; collecting user color try-on interaction data and optimizing the recommendation priority of popular colors through clustering algorithms.

[0012] Preferably, the method further includes: in the sensitive message recording module of the online shopping mall platform, using a natural language processing model to monitor work orders and chat records in real time, and identify sensitive keywords and customer emotions; generating risk warnings based on the sensitive keywords and customer emotions using an anomaly detection algorithm; classifying work order processing priorities based on the risk warnings and triggering automatic handling processes.

[0013] Accordingly, the present invention also provides an intelligent management system for an online shopping mall platform, the system comprising the following modules: a data acquisition module: used to collect platform operation data, user behavior data, and external market data; a sales forecasting module: connected to the data acquisition module, used to receive historical sales data, analyze the historical sales data using the Prophet time series model, predict future sales trends and inventory demand, and generate dynamic replenishment suggestions; an inventory management module: connected to the sales forecasting module, used to optimize inventory management through dynamic programming algorithms, calculate the optimal inventory threshold by combining the dynamic replenishment suggestions, delivery cycle, and material delay probability, and trigger automatic replenishment instructions; a user profiling module: connected to the data acquisition module, used to establish user profiles based on user behavior data, generate personalized coupon strategies based on collaborative filtering algorithms, and update the user profiles in real time; a work order management module: used to dynamically match work orders and customer service based on a multi-attribute decision algorithm, the attributes including customer service professional skills, real-time load, and historical processing efficiency, track work order status, and feed back processing results to the user profiling module; and a quality assurance management module: used to store quality assurance information, construction time, material batches, and technician ID data, and ensure the immutability and traceability of data based on blockchain technology.

[0014] Preferably, the system further includes: a virtual color-matching module: integrating AR augmented reality technology, supporting users to upload vehicle photos and simulate the effect of color-changing film, while collecting color-matching interaction data, and optimizing the recommendation priority of popular colors through clustering algorithms; and a sensitive message monitoring module: using a natural language processing model to analyze work orders and chat records in real time, identify sensitive keywords and customer emotions, and trigger risk warnings through anomaly detection algorithms.

[0015] The present invention has at least the following technical effects through the technical solution provided by the present invention:

[0016] This invention provides an online shopping mall platform management method. In inventory management, it integrates external data such as delivery cycles, weather, and logistics delays, and uses a dynamic programming algorithm to dynamically adjust inventory thresholds. It also adds a visual early warning interface, allowing managers to intuitively grasp inventory status and providing decision support, reducing manual intervention costs and improving platform operational efficiency. In user marketing, it updates user profiles in real time, enabling precise marketing based on user consumption behavior and preferences, generating personalized coupon strategies, and increasing coupon usage and user repurchase rates. In the work order processing stage, it uses a multi-attribute decision algorithm to match work orders with customer service, utilizes a reinforcement learning model to optimize allocation strategies, and uses a BERT model to analyze work order text, improving work order processing efficiency and service quality from multiple dimensions, and increasing allocation accuracy, thereby enhancing user experience. In warranty information management, it uses blockchain technology to store warranty information, linking the blockchain with IoT devices, building a private blockchain node to automatically upload records to associated IoT devices, and providing a query interface to ensure the authenticity, reliability, and traceability of warranty information, enhancing user trust. Attached Figure Description

[0017] The accompanying drawings are provided to further illustrate embodiments of the present invention and form part of the specification. They are used together with the following detailed description to explain the embodiments of the present invention, but do not constitute a limitation thereof. In the drawings:

[0018] Figure 1 This is a flowchart of an intelligent management method for an online shopping mall platform provided by an embodiment of the present invention;

[0019] Figure 2 This is a structural diagram of an intelligent management system for an online shopping mall platform provided in an embodiment of the present invention. Detailed Implementation

[0020] The specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings. It should be understood that the specific embodiments described herein are for illustration and explanation only and are not intended to limit the scope of the present invention.

[0021] In this invention, the terms "system" and "network" are used interchangeably. "Multiple" refers to two or more; therefore, in this invention, "multiple" can also be understood as "at least two." "And / or" describes the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can represent: A alone, A and B simultaneously, or B alone. Additionally, the character " / ", unless otherwise specified, generally indicates that the preceding and following related objects have an "or" relationship. Furthermore, it should be understood that in the description of this invention, terms such as "first" and "second" are used only for descriptive purposes and should not be construed as indicating or implying relative importance or order.

[0022] Please see Figure 1 This invention provides an intelligent management method for an online shopping mall platform, comprising the following steps:

[0023] S1: Build a data acquisition system to acquire platform operation data, user behavior data and external market data in real time. The external market data includes delivery cycle and material delay probability.

[0024] S2: Obtain historical sales data, analyze the historical sales data using the Prophet time series model, predict future sales trends and inventory demand, and generate dynamic replenishment suggestions;

[0025] S3: Optimize inventory management through dynamic programming algorithm and the dynamic replenishment suggestion, and calculate the optimal inventory threshold by combining the delivery cycle and the material delay probability to achieve automatic replenishment triggering;

[0026] S4: Establish a user profile based on the user behavior data, generate a personalized coupon strategy based on the collaborative filtering algorithm and the user profile, and update the user profile in real time;

[0027] S5: Obtain data from the work order management module and dynamically match work orders with customer service representatives based on a multi-attribute decision algorithm;

[0028] S6: Store quality assurance information to ensure the immutability and traceability of construction time, material batches, and technician ID data.

[0029] In this embodiment of the invention, an intelligent management method for an online marketplace platform is provided, applicable to an online marketplace platform for car color-changing films. This online marketplace platform includes the following functional modules: data overview, work order management, store management, SMS group messaging, product management, out-of-stock records, order management, warranty management, coupon management, sensitive message records, financial control center, painting business center, market business center, and system operation and maintenance configuration. The data overview module compiles key operational data for the entire platform, helping managers gain a comprehensive understanding of the platform's sales performance. The work order management module efficiently handles customer service needs and issues, providing complete work order tracking and management functions. Through this module, users can view and manage... The system manages detailed information for all work orders, helping administrators monitor their progress in real time and ensure timely responses to customer needs. The out-of-stock recording module optimizes inventory turnover and product supply efficiency by monitoring product inventory status in real time, helping managers quickly identify out-of-stock items and ensuring a dynamic balance between inventory levels and sales demand. The warranty management module provides comprehensive product warranty service support, ensuring transparency and traceability of warranty information. The coupon management module offers comprehensive coupon issuance and usage tracking. The sensitive message recording module provides comprehensive customer communication monitoring and risk warning functions, ensuring timely identification and handling of sensitive information during customer service, optimizing service processes, and improving customer satisfaction and compliance. This online marketplace platform for car color-changing films has many problems during use; therefore, this invention provides an intelligent management method for the online marketplace platform to improve its operational efficiency, user experience, and market competitiveness.

[0030] In this embodiment of the invention, for step S1, a data acquisition system is first constructed. After construction, the data acquisition system acquires platform operation data, user behavior data, and external market data in real time through an application programming interface (API). The acquired platform operation data includes order volume, sales data, product listing and delisting status, inventory data, and daily user access information. User behavior data is collected through tracking points, including user browsing paths, click heatmaps, shopping cart additions, browsing history, purchase history, and search history. Search keyword information can be extracted from the search history. External market data such as delivery cycles are acquired by connecting with external suppliers and logistics companies, and the probability of material delays is calculated by combining historical logistics data. For example, an online car color change film store monitors through the data acquisition system that the probability of logistics delays in South China has increased to 35% due to typhoon weather. The system automatically marks this area as a high-risk supply chain area.

[0031] Furthermore, in step S2, historical sales data is acquired through a data acquisition system and input into the Prophet time series model. The seasonality period is set to 12 months, and promotional events include "Double Eleven" and "618". The Prophet time series model analyzes the historical sales data, outputting a sales forecast curve and inventory demand for the next three months, and generating replenishment suggestions. In another implementation, after the Prophet model analyzes the historical sales data, the gradient boosting tree, combined with high-temperature weather forecast data and supply chain fluctuations such as recent frequent delivery delays by suppliers, dynamically adjusts replenishment priorities. A visual early warning interface is embedded in the data overview module of the online marketplace platform. This interface obtains real-time inventory data and replenishment logs. Based on inventory data and preset inventory safety thresholds, an inventory health score is generated. Inventory health is categorized into green, yellow, and red alerts. Managers can view the inventory health score through this visual alert interface and also view detailed replenishment logs for timely inventory monitoring. Furthermore, for step S3, inventory management is optimized using dynamic programming algorithms and replenishment suggestions. The dynamic programming algorithm comprehensively considers predicted sales volume, delivery cycle, and material delay probability to calculate the optimal inventory threshold. Automatic replenishment is triggered based on this optimal threshold to avoid stockouts. For example, taking an online car wrapping film store as an example, before summer, the system acquires the following data through a data collection system: sales data for the past three summers, current inventory data, and delivery cycles of each supplier (e.g., supplier A). The system considers several factors: Supplier A's delivery cycle is 15 days, Supplier B's is 20 days; material delay probability (Supplier A has a 10% delay probability due to transportation issues, Supplier B has a 15% probability); and high-temperature weather forecasts for the next two months issued by the meteorological department. Then, the Prophet time series model analyzes historical sales data and predicts a 40% increase in sales of a certain light-colored color-changing film during the summer. Simultaneously, a gradient boosting tree, combined with high-temperature weather forecasts and supply chain fluctuations from Supplier A's recent frequent delivery delays, prioritizes this color-changing film for replenishment. A dynamic programming algorithm, considering predicted sales, delivery cycle, and material delay probability, calculates an optimal inventory threshold of 800 rolls. However, with only 300 rolls currently in stock, the system automatically issues a replenishment order to Supplier A, preventing stockouts. On the visual alert interface, managers can see the real-time inventory health status of this color-changing film under a yellow alert and view detailed replenishment logs for timely inventory monitoring.

[0032] In this embodiment of the invention, for step S4, based on the user profile collected by the data acquisition system, a detailed user profile is established, including information such as purchase frequency, price sensitivity, color preference, user browsing path, click heatmap, shopping cart addition behavior, browsing history, purchase history, and search history. A personalized coupon strategy is generated based on the collaborative filtering algorithm and the user profile, and the user profile is updated in real time according to the real-time changes in user behavior data. For example, a user A on the platform has browsed high-end color-changing films priced above 2000 yuan multiple times in the past six months, but has only purchased a mid-to-low-end product once. The system establishes a detailed user profile based on their browsing history, purchase history, and other user behavior data, and labels their profile accordingly. Users A showed strong interest in high-end color-changing films but were cautious in their purchasing decisions. Collaborative filtering analysis revealed that user A shared similar browsing preferences and brand interests with other users who had already purchased high-end color-changing films. Based on this, the system generated a personalized coupon strategy, sending user A a "300 RMB off for purchases over 2000 RMB" exclusive coupon via SMS and app notifications. After receiving the coupon, user A completed the purchase within a week, achieving precise marketing and improving conversion rates and average order value. Alternatively, another user, B, browsed electroplated blue color-changing films multiple times within three days but did not place an order. The system updated their profile in real-time as a "hesitant user" and sent them a "limited-time 20% off + free installation coupon," ultimately leading to conversion.

[0033] In this embodiment of the invention, for step S5, the work order data of the work order management module is obtained, and work orders and customer service are dynamically matched according to the multi-attribute decision algorithm. Specifically, a customer service information database is first established, which records customer service professional skills, real-time load, and historical satisfaction data. This database is then combined with the multi-attribute decision algorithm and reinforcement learning model to dynamically optimize the work order allocation strategy, balancing skill matching, response time, and load balancing, thereby improving work order processing efficiency and customer experience. When optimizing the work order allocation strategy, the work order text needs to be obtained and semantically classified using the BERT model. This semantic classification allows for the identification of work order urgency and customer sentiment, which are then considered in the work order allocation strategy. Regarding the combination of the multi-attribute decision algorithm and reinforcement learning model, the state space defined in the reinforcement learning model represents the real-time load of customer service, the backlog of work orders, and the distribution of customer sentiment; the action space is the work order allocation decision. The short-term reward function set in the reinforcement learning model is the work order resolution time and the customer satisfaction score. The long-term reward function is the customer service skill improvement rate and the overall work order resolution rate. Furthermore, in addition to multi-attribute decision-making algorithms and reinforcement learning models, SHAP values ​​can be used to explain the algorithm's allocation logic and generate an interpretable allocation report. That is, after the reinforcement learning model outputs work order allocation decisions, for each allocation decision, the SHAP value of each feature is calculated. For example, customer service A's skill matching contributes +0.4 (positive impact), customer service B's current workload contributes -0.3 (negative impact), and work order urgency contributes +0.2. The sum of the SHAP values ​​equals the difference in the model's final decision score (e.g., customer service A's score is 0.3 higher than customer service B's). Then, an interpretable report is generated, including feature importance ranking (listing the top 3 factors with the greatest impact on this allocation), contribution value visualization (displaying the SHAP value of each feature through a bar chart or waterfall chart), and a summary of the decision logic (e.g., this allocation is due to customer service A's skill matching (+0.4) being significantly higher than customer service B's (+0.1), and customer service B's current workload being higher (-0.3)). This aims to achieve algorithm transparency and intelligent management.

[0034] Furthermore, regarding the work order allocation optimization strategy provided in this embodiment, for example, the work order management module of the online mall platform receives three work orders simultaneously: Work order 1: The user reports that the edge of the color-changing film in the new post is slightly curled, and the text content is relatively calm; Work order 2: The user claims that the color-changing film has peeled off in a large area, and the tone is very agitated, demanding an immediate solution; Work order 3: The user inquires about the warranty period of a certain color-changing film, and the attitude is friendly. The BERT model first performs semantic classification on the work order text, identifying work order 2 as urgent and indicating an agitated customer, work order 1 as a general issue, and work order 3 as an inquiry. Next, the customer service information database shows that Customer Service Representative A is skilled at handling after-sales issues and currently has a low workload, Customer Service Representative B is skilled at answering inquiries and is not overworked, and Customer Service Representative C is handling other urgent work orders. Then, a multi-attribute decision algorithm combined with a reinforcement learning model assigns work order 2 to Customer Service Representative A, work order 3 to Customer Service Representative B, and work order 1 is temporarily queued. Upon receiving the work order, Customer Service Representative A quickly communicates with the user, arranges for a professional to inspect and reapply the screen protector, and achieves a user satisfaction score of 4.8 out of 5. Customer Service Representative B promptly answers the user's inquiry, and the user gives positive feedback. Finally, the system generates a work order assignment report using the SHAP value interpretation algorithm, clearly showing the reasons for assigning work order 2 to Customer Service Representative A, facilitating management's evaluation and optimization of the assignment strategy.

[0035] In this embodiment of the invention, for step S6, storing warranty information, specifically, a private chain is first built, that is, a consortium chain is built based on Hyperledger Fabric. The nodes of the private chain include the headquarters, construction stores, and logistics companies; the data of the Internet of Things (IoT) devices (hereinafter referred to as IoT devices) of the construction stores are associated, and warranty operation records are automatically uploaded. That is, after the construction is completed, the IoT devices of the stores automatically upload the construction time, material batch, and technician ID to the blockchain; a blockchain query interface for clients is provided, which supports scanning QR codes to verify the entire chain of warranty information. That is, customers can scan the QR code on the warranty card to query the entire chain of information (such as the source of material batch and the hash value of construction video), so as to ensure the immutability and traceability of construction time, material batch, and technician ID data. For example, after a user purchases a color-changing film and completes the installation, the IoT devices at the installation store (such as cameras and electronic tag readers) automatically record the installation time (e.g., 2024-06-01 14:30), the batch number of the materials used (e.g., identified as NO.20240601A via electronic tag), and the technician's ID (e.g., recorded as TECH-001 via employee ID card check-in). These warranty operation records are automatically uploaded to the blockchain system through private chain nodes, forming an immutable record. One month later, if the user discovers that the color-changing film has faded, they can scan the QR code on the product packaging and use the client's blockchain query interface to view the full-chain warranty information of the color-changing film, including video clips of the installation process (recorded and uploaded by the camera), detailed information on the material batch, and technician qualification certificates. Based on the blockchain records, the merchant can quickly trace the material supplier and the problem in the installation process, promptly resolving disputes for the user, protecting the user's rights, and maintaining their own reputation.

[0036] In another embodiment, the intelligent management method for the online shopping mall platform provided by the present invention further includes integrating AR augmented reality technology to provide a virtual color try-on function, supporting users to upload vehicle photos and simulate the effect of color change film in real time; wherein the virtual color try-on function includes: building a 3D vehicle model library to support the simulation of color change film effects for mainstream models; and can also collect user color try-on interaction data and optimize the recommendation priority of popular colors through clustering algorithms. For example, an online shopping platform can build a 3D vehicle model library. When users select color-changing films, they can use the AR virtual try-on function to upload photos of their vehicles. The system matches the vehicle models from the 3D model library and uses AR technology to simulate the effect of the color-changing film in real time, displaying the real-time simulation effect of different color-changing films. Users tried various colors such as red, blue, and gray, and spent a relatively long time trying on the blue color-changing film, repeatedly adjusting the viewing angle to check the effect. After collecting user color-trying interaction data, the system analyzed the data using clustering algorithms and found that the number of users browsing and trying on blue color-changing films recently had increased significantly, and this group had a high purchase conversion rate. Based on this data, the online shopping platform can embed a color-trying data heatmap in the painting business center module, and mark the blue color-changing film as a popular product in the color-trying data heatmap of the painting business center module, and adjust the inventory strategy to increase the stock of blue color-changing films. At the same time, the platform can increase the promotion of blue color-changing films on the homepage and product recommendation page, thereby increasing the platform's sales.

[0037] In another embodiment, the intelligent management method for online shopping mall platforms provided by this invention further includes firstly, using a natural language processing model (hereinafter referred to as NLP model) to monitor work orders and chat logs in real time within the sensitive message recording module of the online shopping mall platform, and identifying sensitive keywords and customer emotions. Specifically, the NLP model detects sensitive keywords such as "complaint" and "counterfeit goods" in the chat logs and analyzes customer emotions; then, based on the sensitive keywords and customer emotions, it generates risk warnings through anomaly detection algorithms; finally, based on the risk warnings, it classifies the work order processing priority and triggers an automatic handling process, that is, the system automatically marks it as a high-risk work order and notifies the supervisor to intervene. For example, if a message appears in the chat log between customer service and a user saying, "Your color-changing film is completely counterfeit, I'm going to file a complaint with the Consumer Association," the NLP model identifies sensitive keywords such as "counterfeit goods" and "complaint" in real time, judges the customer's emotion to be anger, then uses an anomaly detection algorithm to generate a high-risk warning, the system immediately raises the processing priority of the work order to the highest level, and triggers an automatic handling process. Meanwhile, the system automatically notified the customer service manager to intervene and immediately sent a reassuring text message to the user, promising a solution within 24 hours. After communicating and negotiating with the user, the customer service manager arranged for professional inspectors to test the color-changing film. It was ultimately confirmed that the problem was caused by improper installation, and the system was redone for the user free of charge, successfully resolving the user's problem.

[0038] This invention provides an intelligent management method for online shopping mall platforms. Through steps such as constructing a data collection system, employing advanced data analysis and prediction models, optimizing inventory management, generating personalized marketing strategies, dynamically matching work orders and customer service, and storing warranty information, it achieves intelligent management of the online shopping mall platform. This significantly improves inventory turnover, increases work order response speed and customer satisfaction, and greatly enhances the efficiency of warranty dispute resolution. Furthermore, automated decision-making reduces manual intervention, leading to more accurate resource allocation and improved efficiency. AR color try-on, personalized recommendations, and rapid work order response further enhance customer satisfaction. Blockchain-based warranty traceability and sensitive message monitoring ensure data transparency and compliance, increasing customer trust. Reinforced learning and dynamic optimization capabilities support continuous business iteration. Overall, this method improves the operational efficiency and user experience of the shopping mall platform.

[0039] Please see Figure 2 Based on the same inventive concept, embodiments of the present invention provide an intelligent management system for an online shopping mall platform, the system comprising the following modules:

[0040] Data acquisition module: used to collect platform operation data, user behavior data, and external market data;

[0041] Sales forecasting module: Connected to the data acquisition module, it receives historical sales data, analyzes the historical sales data using the Prophet time series model, predicts future sales trends and inventory demand, and generates dynamic replenishment suggestions;

[0042] Inventory Management Module: Connected to the sales forecasting module, it is used to optimize inventory management through dynamic programming algorithms, calculate the optimal inventory threshold by combining the dynamic replenishment suggestions, delivery cycle and material delay probability, and trigger automatic replenishment instructions;

[0043] User profiling module: connected to the data acquisition module, used to build user profiles based on user behavior data, generate personalized coupon strategies based on collaborative filtering algorithms, and update the user profiles in real time;

[0044] Work order management module: used to dynamically match work orders with customer service representatives based on a multi-attribute decision algorithm. The attributes include customer service representative professional skills, real-time load, and historical processing efficiency. It tracks the status of work orders and feeds back the processing results to the user profile module.

[0045] Quality Assurance Management Module: Used to store quality assurance information, construction time, material batches, and technician ID data. Based on blockchain technology, it ensures the immutability and traceability of the data.

[0046] In another embodiment, the online shopping platform intelligent management system provided by this invention further includes:

[0047] Virtual color try-on module: Integrates AR augmented reality technology, supports users to upload vehicle photos and simulate the effect of color change film, and collects color try-on interaction data, and optimizes the recommendation priority of popular colors through clustering algorithms;

[0048] Sensitive message monitoring module: It uses a natural language processing model to analyze work orders and chat records in real time, identify sensitive keywords and customer sentiments, and trigger risk warnings through anomaly detection algorithms.

[0049] The optional embodiments of the present invention have been described in detail above with reference to the accompanying drawings. However, the embodiments of the present invention are not limited to the specific details in the above embodiments. Within the scope of the technical concept of the embodiments of the present invention, various simple modifications can be made to the technical solutions of the embodiments of the present invention, and these simple modifications all fall within the protection scope of the embodiments of the present invention.

[0050] It should also be noted that the various specific technical features described in the above embodiments can be combined in any suitable manner without contradiction. To avoid unnecessary repetition, the embodiments of the present invention will not describe the various possible combinations separately.

[0051] Those skilled in the art will understand that all or part of the steps in the methods of the above embodiments can be implemented by a program instructing related hardware. This program is stored in a storage medium and includes several instructions to cause a microcontroller, chip, or processor to execute all or part of the steps of the methods described in the various embodiments of this application. The aforementioned storage medium includes various media capable of storing program code, such as a USB flash drive, a portable hard drive, a read-only memory (ROM), a random access memory (RAM), a magnetic disk, or an optical disk.

[0052] Furthermore, various different implementations of the present invention can be combined arbitrarily, as long as they do not violate the spirit of the present invention, they should also be regarded as the content disclosed in the present invention.

Claims

1. An online mall platform intelligent management method, characterized in that, The method comprises the following steps: S1: constructing a data acquisition system to obtain platform operation data, user behavior data, and external market data in real time, wherein the external market data comprises a delivery cycle and a material delay probability; S2: obtaining historical sales data, analyzing the historical sales data by using a Prophet time series model, predicting future sales trends and inventory demand, and generating a dynamic replenishment suggestion, wherein generating the dynamic replenishment suggestion specifically comprises: obtaining weather data and supply chain fluctuation information, integrating the weather data and the supply chain fluctuation information, and dynamically adjusting a replenishment priority according to a gradient boosting tree; embedding a visual early warning interface in a data overview module of an online mall platform, obtaining inventory data in real time and generating an inventory health degree, and obtaining a replenishment log in real time, wherein the visual early warning interface displays the inventory health degree and the replenishment log in real time; S3: optimizing inventory management by using a dynamic programming algorithm and the dynamic replenishment suggestion, calculating an optimal inventory threshold in combination with the delivery cycle and the material delay probability, and realizing automatic replenishment triggering; S4: establishing a user portrait according to the user behavior data, generating an individualized coupon strategy based on a collaborative filtering algorithm and the user portrait, and updating the user portrait in real time; S5: obtaining work order management module data, dynamically matching work orders and customer service according to a multi-attribute decision algorithm, specifically comprising: establishing a customer service information database, wherein the customer service information database records customer service professional skills, real-time load, and historical satisfaction data; combining a multi-attribute decision algorithm and a reinforcement learning model to dynamically optimize a work order distribution strategy, balance skill matching degree, response time, and load balancing, obtain work order text, and perform semantic classification on the work order text by using a BERT model to identify work order urgency and customer emotion; wherein a state space defined by the reinforcement learning model is customer real-time load, work order backlog, and customer emotion distribution, an action space is a work order distribution decision; a short-term reward function is set as work order solution time and customer satisfaction score, and a long-term reward function is customer service skill improvement rate and overall work order solution rate; a SHAP value explanation algorithm is used to distribute logic to generate an explainable distribution report; S6: storing warranty information to ensure that construction time, material batch, and technician ID data are tamper-proof and traceable; wherein the storage of warranty information specifically comprises: building a private chain node, associating Internet of Things device data of a construction store, and automatically uploading warranty operation records; providing a client-side blockchain query interface to support scanning and verifying warranty information in the whole link, and the whole link information comprises material batch source and construction video hash value.

2. The intelligent management method of an online mall platform according to claim 1, characterized in that, The method further comprises: integrating AR augmented reality technology to provide a virtual color testing function, supporting users to upload vehicle photos and simulate color film effects in real time; the virtual color testing function comprises: constructing a vehicle 3D model library to support color film effect simulation for mainstream vehicle models; collecting user color testing interaction data and optimizing popular color recommendation priority by using a clustering algorithm.

3. The intelligent management method of an online mall platform according to claim 1, characterized in that, The method further comprises: in a sensitive message record module of an online mall platform, using a natural language processing model to monitor work order and chat records in real time, and identifying sensitive keywords and customer emotion; generating a risk warning through an anomaly detection algorithm according to the sensitive keywords and the customer sentiment; classifying the work order processing priority according to the risk warning and triggering an automatic handling process.

4. An intelligent management system for an online mall platform, characterized in that, The system comprises the following modules based on the intelligent management method of an online mall platform according to any one of claims 1-3: a data collection module for collecting platform operation data, user behavior data, and external market data; a sales prediction module connected with the data collection module, for receiving historical sales data, analyzing the historical sales data using a Prophet time series model, predicting future sales trends and inventory demand, and generating dynamic restocking suggestions, wherein weather data and supply chain fluctuation information are obtained, the weather data and the supply chain fluctuation information are integrated, and the restocking priority is dynamically adjusted according to a gradient boosting tree; an inventory management module connected with the sales prediction module, for optimizing inventory management through a dynamic programming algorithm, combining the dynamic restocking suggestions, delivery cycle, and material delay probability to calculate the optimal inventory threshold, and triggering automatic restocking instructions; a user portrait module connected with the data collection module, for establishing a user portrait based on user behavior data and generating personalized coupon strategies based on a collaborative filtering algorithm, while updating the user portrait in real time; a work order management module for dynamically matching work orders and customer service according to a multi-attribute decision algorithm, the attributes including customer service professional skills, real-time load, and historical processing efficiency, tracking work order status and feeding back processing results to the user portrait module; specifically including: combining a multi-attribute decision algorithm and a reinforcement learning model to dynamically optimize work order allocation strategies, and obtaining work order text and classifying it through a BERT model to identify the urgency of the work order and the customer sentiment; wherein the state space defined by the reinforcement learning model is the real-time load of the customer service, the work order backlog, and the customer sentiment distribution, and the action space is the work order allocation decision; the short-term reward function is set as the work order solution time and the customer satisfaction score, and the long-term reward function is the customer service skill improvement rate and the overall work order solution rate; a quality assurance management module for storing quality assurance information, construction time, material batch, and technician ID data, and ensuring data non-tamperability and traceability based on blockchain technology.

5. The intelligent management system of an online mall platform according to claim 4, characterized in that, The system further comprises: a virtual color testing module integrated with AR augmented reality technology, supporting users to upload vehicle photos and simulate color film effects, while collecting color testing interaction data, and optimizing popular color recommendation priority through a clustering algorithm; a sensitive message monitoring module using a natural language processing model to analyze work orders and chat records in real time, identify sensitive keywords and customer sentiment, and trigger a risk warning through an anomaly detection algorithm.

Images