An event-driven product data management method and related equipment

Abstract

This application belongs to the field of artificial intelligence technology and relates to an event-driven product data management method and related equipment. The method includes: real-time acquisition of end-to-end data of a target product; when the end-to-end data is detected to meet the triggering conditions of a risk event, constructing an event task corresponding to the target product based on the end-to-end data; calling a task orchestration model and decomposing the event task according to the task orchestration model to obtain N event sub-tasks, where N is a positive integer; generating multimedia materials and auxiliary description text based on the task description text of each event sub-task; constructing product display data corresponding to the event task based on the multimedia materials and auxiliary description text; and outputting the product display data to the user terminal corresponding to the target product. This application can be used in fintech business systems to generate relevant product content, improving the efficiency and accuracy of product data management.

Description

Technical Field

[0001] This application relates to the field of artificial intelligence technology, and in particular to an event-driven product data management method and related equipment. Background Technology

[0002] In today's digital age, end-to-end product data encompasses a wealth of information from product design, production, sales to after-sales service. This data is crucial for businesses to understand product status, optimize product strategies, and improve user satisfaction, especially for products in fintech, healthcare, and elderly care. However, current product data management methods have many shortcomings.

[0003] Existing product data management methods often focus on static data storage and simple queries, lacking the ability to perceive and respond to dynamic changes in data in real time. When product data becomes abnormal or meets certain conditions, the corresponding processing procedures cannot be triggered in a timely manner, making it difficult for enterprises to discover potential risks and take effective measures in a timely manner.

[0004] This shows that existing product data management methods suffer from low management efficiency. Summary of the Invention

[0005] The purpose of this application is to propose an event-driven product data management method and related equipment to solve the problem of low management efficiency in existing product data management methods.

[0006] To address the aforementioned technical problems, this application provides an event-driven product data management method, employing the following technical solution: Real-time collection of end-to-end data for target products; When the end-to-end data is detected to meet the risk event triggering conditions, an event task corresponding to the target product is constructed based on the end-to-end data. The task orchestration model is invoked, and the event task is decomposed according to the task orchestration model to obtain N event subtasks, where N is an integer greater than zero; Multimedia materials and auxiliary descriptive text are generated based on the task description text of each event subtask; Based on the multimedia materials and the auxiliary descriptive text, construct product display data corresponding to the event task; The product display data is output to the user terminal corresponding to the target product.

[0007] To address the aforementioned technical problems, this application also provides an event-driven product data management device, employing the following technical solution: The data acquisition module is used to collect real-time data from the entire supply chain of the target product. The event task construction module is used to construct an event task corresponding to the target product based on the end-to-end data when the end-to-end data is detected to meet the risk event triggering conditions. The task decomposition module is used to call the task orchestration model and decompose the event task according to the task orchestration model to obtain N event subtasks, where N is an integer greater than zero. The subtask execution module is used to generate multimedia materials and auxiliary description text based on the task description text of each event subtask; The product data construction module is used to construct product display data corresponding to the event task based on the multimedia materials and the auxiliary descriptive text. The product data output module is used to output the product display data to the user terminal corresponding to the target product.

[0008] To address the aforementioned technical problems, this application also provides a computer device that employs the following technical solution: It includes a memory and a processor, wherein the memory stores computer-readable instructions, and the processor executes the computer-readable instructions to implement the steps of the event-driven product data management method described above.

[0009] To address the aforementioned technical problems, this application also provides a computer-readable storage medium, employing the technical solution described below: The computer-readable storage medium stores computer-readable instructions, which, when executed by a processor, implement the steps of the event-driven product data management method described above.

[0010] This application provides an event-driven product data management method, comprising: real-time collection of end-to-end data of a target product; when the end-to-end data is detected to meet the triggering conditions of a risk event, constructing an event task corresponding to the target product based on the end-to-end data; invoking a task orchestration model and decomposing the event task according to the task orchestration model to obtain N event sub-tasks, where N is a positive integer; generating multimedia materials and auxiliary description text based on the task description text of each event sub-task; constructing product display data corresponding to the event task based on the multimedia materials and auxiliary description text; and outputting the product display data to the user terminal corresponding to the target product. Compared with the prior art, this application can collect end-to-end data of a target product in real time, automatically construct and decompose event tasks when the data is detected to meet the triggering conditions of a risk event, generate multimedia materials and auxiliary description text, and finally construct product display data and output it to the user terminal, thereby improving the efficiency and accuracy of product data management. Attached Figure Description

[0011] To more clearly illustrate the solutions in this application, the accompanying drawings used in the description of the embodiments of this application will be briefly introduced below. Obviously, the accompanying drawings described below are some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0012] Figure 1 This is an exemplary system architecture diagram to which this application can be applied; Figure 2 This is a flowchart illustrating the implementation of the event-driven product data management method provided in this application embodiment; Figure 3 This is a schematic diagram of the structure of the event-driven product data management device provided in the embodiments of this application; Figure 4 This is a schematic diagram of the structure of one embodiment of the computer device according to this application. Detailed Implementation

[0013] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application pertains; the terminology used herein in the specification of the application is for the purpose of describing particular embodiments only and is not intended to be limiting of the application; the terms "comprising" and "having," and any variations thereof, in the specification, claims, and foregoing drawings of this application, are intended to cover non-exclusive inclusion. The terms "first," "second," etc., in the specification, claims, or foregoing drawings of this application are used to distinguish different objects, not to describe a particular order.

[0014] In this document, the term "embodiment" means that a particular feature, structure, or characteristic described in connection with an embodiment may be included in at least one embodiment of this application. The appearance of this phrase in various places throughout the specification does not necessarily refer to the same embodiment, nor is it a separate or alternative embodiment mutually exclusive with other embodiments. It will be explicitly and implicitly understood by those skilled in the art that the embodiments described herein can be combined with other embodiments.

[0015] To enable those skilled in the art to better understand the present application, the technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the accompanying drawings.

[0016] like Figure 1 As shown, system architecture 100 may include terminal device 101, network 102, and server 103. Terminal device 101 may be a laptop 1011, tablet 1012, or mobile phone 1013. Network 102 is used as a medium to provide a communication link between terminal device 101 and server 103. Network 102 may include various connection types, such as wired, wireless communication links, or fiber optic cables, etc.

[0017] Users can use terminal device 101 to interact with server 103 via network 102 to receive or send messages, etc. Various communication client applications can be installed on terminal device 101, such as web browser applications, shopping applications, search applications, instant messaging tools, email clients, social media platform software, etc.

[0018] Terminal device 101 can be various electronic devices with a display screen and support web browsing. In addition to laptops 1011, tablets 1012, or mobile phones 1013, terminal device 101 can also be an e-book reader, an MP3 player (Moving Picture Experts Group Audio Layer III), an MP4 player (Moving Picture Experts Group Audio Layer IV), a laptop computer, and a desktop computer, etc.

[0019] Server 103 can be a server that provides various services, such as a backend server that provides support for the pages displayed on terminal device 101.

[0020] It should be noted that the event-driven product data management method provided in this application is generally executed by a server / terminal device, and correspondingly, the event-driven product data management device is generally located in the server / terminal device.

[0021] It should be understood that Figure 1 The number of terminal devices, networks, and servers shown is merely illustrative. Depending on implementation needs, any number of terminal devices, networks, and servers can be included.

[0022] Continue to refer to Figure 2 The diagram illustrates a flowchart of an embodiment of the event-driven product data management method according to this application. The event-driven product data management method described above includes steps S201, S202, S203, S204, S205, and S206.

[0023] In this application embodiment, an event-driven, four-agent high-performance enabling system is proposed. By constructing an agentic core, the system achieves independence and high performance among the agents. Specifically, the event-driven, four-agent high-performance enabling system includes: (1) Unified Knowledge and Data Layer: Includes real-time formatted data (CRM, performance) and supported unformatted data (product knowledge, dialogue scripts, marketing materials); (2) Real-time event wake-up: It adopts the publish-subscribe (Pub / Sub) mode and is the only participant in communication between various intelligent agents; (3) Task orchestrator: responsible for receiving high-level instructions, hierarchical tasks, and forwarding coordinated instructions issued by other intelligent agents; (4) Four major intelligent agents: software agents with independent model and tool calling capabilities, including: business analysis agent, task planning agent, content generation agent, and product display agent. Specifically: ① Business Analysis Intelligent Agent: Its core responsibility is to gain insights and analysis, monitor data across the entire value chain (performance, behavior, market) in real time, achieve accurate performance insights, and proactively issue event tasks that meet the triggering conditions of risk events; ②Task planning agent: Its core responsibility is to make decisions and schedule event tasks, automatically create, dynamically adjust and sort daily work tasks to form a dynamic task flow; ③ Content Generation Intelligent Agent: Its core responsibility is creation and empowerment. Based on hot topics and task instructions, it uses AIGC capabilities to quickly generate compliant copywriting, images, videos and other multimedia materials, as well as auxiliary descriptive text. ④ Product Display Intelligent Agent: Its core responsibility is to display product data to user terminals and monitor the interaction between specialists and customers (clicks, conversations) in real time, providing insightful customer perspectives, real-time script suggestions, and objection handling solutions.

[0024] In step S201, the entire chain data of the target product is collected in real time.

[0025] In the embodiments of this application, this application can collect the full-link data of the target product in real time through the above-mentioned business analysis intelligent agent and various data acquisition devices and interfaces. Among them, the data acquisition devices and interfaces include, but are not limited to, sensors, Internet of Things devices, enterprise information system interfaces, etc. It should be understood that the examples of the methods of obtaining full-link data are only for the convenience of understanding and are not intended to limit this application.

[0026] In the embodiments of this application, the target product refers to the specific object on which the data is focused, managed and displayed in this application. As an example, the target product may be a high-end medical product that the company is currently promoting, "High-End Medical". The target product may also be a financial insurance product that the company is currently promoting. It should be understood that the examples of target products here are only for the convenience of understanding and are not intended to limit this application.

[0027] In this application embodiment, full-link data refers to the data of the target product in all stages of its entire life cycle, including design, production, sales, and after-sales service. This full-link data includes, but is not limited to, information such as product performance parameters, production progress, sales volume, and user feedback.

[0028] In this embodiment of the application, an efficient data acquisition network is established to ensure the timeliness and accuracy of the data, providing a foundation for subsequent data analysis and processing.

[0029] In step S202, when it is detected that the end-to-end data meets the risk event triggering conditions, an event task corresponding to the target product is constructed based on the end-to-end data.

[0030] In the embodiments of this application, a series of risk event triggering conditions are pre-defined, and these conditions can be flexibly configured according to the characteristics of the product and business needs. As an example, for instance, when a certain performance parameter of the product exceeds the normal range, the production schedule is severely delayed, or the sales volume fluctuates abnormally, the corresponding event is triggered. It should be understood that the examples of risk event triggering conditions here are only for ease of understanding and are not intended to limit this application.

[0031] In this application embodiment, once the end-to-end data is detected to meet the triggering conditions, this application will construct an event task corresponding to the target product through the above-mentioned task planning agent and the specific content and business logic of the end-to-end data, and send the event task to the event entry system. The event task clarifies the problem and goal that needs to be dealt with, providing direction for subsequent task execution.

[0032] In practical application, taking the aforementioned high-end medical product "High-End Medical" as an example, suppose that during the monitoring process, "Event" A is discovered: The company's recently promoted high-end medical product, "High-End Medical," has significantly underperformed expectations in a certain regional market over the past two weeks, and its market share has also declined compared to similar products. Simultaneously, a highly interested customer is also detected. This customer previously showed strong interest in high-end medical products, repeatedly inquiring about product details, but has not made a purchase for seven consecutive days. Furthermore, this application pre-sets product performance assessment conditions and customer purchase conditions. Through testing, this application finds that the aforementioned "Event" A will trigger the risk event triggering conditions. "Event" A will be constructed as a "high-end medical product performance lag event" and a "highly interested customer failing to make a purchase for seven consecutive days event" corresponding to the high-end medical product "High-End Medical."

[0033] In step S203, the task orchestration model is invoked, and the event task is decomposed according to the task orchestration model to obtain N event subtasks, where N is an integer greater than zero.

[0034] In this embodiment, the task orchestration model is primarily used to rationally decompose complex event tasks. Based on the nature and requirements of the event task, it is broken down into multiple interrelated sub-tasks with a clear execution order. For example, an event task involving product quality issue handling can be decomposed into data verification sub-tasks, problem localization sub-tasks, and solution formulation sub-tasks, etc.

[0035] In this embodiment of the application, when the event task is published to the event entry system, the task planning agent is awakened by subscribing to the event task, decomposes the event task, transforms the complex task into multiple simple and easy-to-execute sub-tasks, assigns the sub-tasks, and publishes the sub-tasks to the event tracking system to improve the efficiency and accuracy of task processing.

[0036] In step S204, multimedia materials and auxiliary description text are generated based on the task description text of each event subtask.

[0037] In the embodiments of this application, multimedia materials are mainly used to intuitively display relevant information of the event sub-task, and may include various forms such as images, videos, and audio. As an example, for an event sub-task of product performance abnormality, the generated multimedia material may be a video of product performance testing. It should be understood that the examples of multimedia materials given here are for ease of understanding only and are not intended to limit this application.

[0038] In the embodiments of this application, the auxiliary description text is mainly used to provide detailed explanations and supplements to multimedia materials, helping users to better understand the task content. For example, for a subtask involving product performance anomalies, the generated auxiliary description text details the testing environment, testing methods, and the specific manifestations of the anomaly. It should be understood that the examples of auxiliary description text provided here are for ease of understanding only and are not intended to limit this application.

[0039] In this embodiment of the application, for each event subtask published in the event tracking system, the above-mentioned content generation agent subscribes to the assigned event and, based on the description text of each event subtask, uses natural language processing technology and multimedia generation algorithms to automatically generate corresponding multimedia materials and auxiliary description text, and stores the generated materials (including tracking tags) into the above-mentioned unified knowledge base.

[0040] In step S205, product display data corresponding to the event task is constructed based on multimedia materials and auxiliary descriptive text.

[0041] In this application embodiment, product display data refers to data that possesses a clear information hierarchy and a good user experience, capable of comprehensively and accurately conveying key product information and event handling status. For example, product display data can be displayed in the form of web pages, mobile application interfaces, etc., with reasonable layout and design allowing users to quickly obtain the information they need. It should be understood that the examples of product display data provided here are for ease of understanding only and are not intended to limit this application.

[0042] In this embodiment of the application, the generated multimedia materials and auxiliary descriptive text can be integrated and arranged, and according to a certain logical structure and display rules, product display data corresponding to the event task can be constructed, and the product display data can be published as a "content is ready" event.

[0043] In step S206, product display data is output to the user terminal corresponding to the target product.

[0044] In this embodiment, after a "content is ready" event is published, the product display agent subscribes to the event, determines the corresponding user terminal based on the target product's user group and usage scenario, and transmits the constructed product display data to the user terminal in real time via wired or wireless network, enabling users to promptly understand the latest product status and event handling. The user terminal can be a mobile terminal such as a mobile phone, smartphone, laptop, digital radio receiver, PDA (personal digital assistant), PAD (tablet computer), PMP (portable multimedia player), navigation device, etc., or a fixed terminal such as a digital TV, desktop computer, etc. It should be understood that the examples of user terminals given here are for convenience of understanding only and are not intended to limit this application.

[0045] In some optional implementations of the embodiments of this application, this application may also provide interactive functions, allowing users to provide feedback and query the displayed data, further enhancing the interaction between users and the product.

[0046] In some optional implementations of this application's embodiments, all execution results and transaction data, including customer clicks, triggering states of communication with customers, and final transaction information, are fed back to the unified knowledge and data layer in real time. The business analysis agent can periodically retrieve this data from the unified knowledge and data layer and re-examine whether the conditions for triggering risk events have been met, thereby continuously optimizing the business processes of the target product and improving business performance.

[0047] This application provides an event-driven product data management method, comprising: real-time collection of end-to-end data of a target product; when the end-to-end data is detected to meet the triggering conditions of a risk event, constructing an event task corresponding to the target product based on the end-to-end data; invoking a task orchestration model and decomposing the event task according to the task orchestration model to obtain N event sub-tasks, where N is a positive integer; generating multimedia materials and auxiliary description text based on the task description text of each event sub-task; constructing product display data corresponding to the event task based on the multimedia materials and auxiliary description text; and outputting the product display data to the user terminal corresponding to the target product. Compared with the prior art, this application can collect end-to-end data of a target product in real time, automatically construct and decompose event tasks when the data is detected to meet the triggering conditions of a risk event, generate multimedia materials and auxiliary description text, and finally construct product display data and output it to the user terminal, thereby improving the efficiency and accuracy of product data management.

[0048] In some optional implementations of the embodiments of this application, the step of constructing an event task corresponding to the target product based on end-to-end data specifically includes the following steps: Retrieve the relevant fields corresponding to the target product from the end-to-end data; Read the task template database and retrieve the target task template corresponding to the type of risk event triggering condition from the task template database; Fill the parameters in the associated fields into the target task template to obtain the event task.

[0049] In this application embodiment, the application can filter fields directly associated with the target product based on a predefined field mapping rule base (such as a JSON format rule file) and convert them into a unified data format (such as key-value pairs {"product_id":"D20251226","temperature":85}).

[0050] In this embodiment, the application reads the trigger condition type (such as "temperature exceeds threshold" or "operation interruption") from the risk event rule base, and the condition type is associated with the target product; then, in the task template database, the application obtains the target task template corresponding to the trigger condition type (such as the template content containing task steps, execution logic, and parameter placeholders) by searching for condition type keywords (such as the SQL query SELECT * FROM templates WHERE condition_type="temperature exceeds threshold") or classifying by machine learning models (such as using decision trees to match condition types with template IDs).

[0051] In this embodiment, the application extracts parameter values ​​(e.g., temperature=85) from the associated fields, verifies their data type (e.g., numeric, string) and range (e.g., temperature threshold 0-100℃); then, the verified parameters are filled into placeholders in the target task template (e.g., the template "If temperature > {{temp}}℃, perform cooling operation" is filled into "If temperature > 85℃, perform cooling operation"), generating an executable event task (e.g., JSON format task instruction {"task_id": "T20251226-001","action":"cooling","threshold":85}); finally, the generated event task is pushed to a task queue (e.g., a Kafka message queue) or stored in a task database (e.g., MySQL) for downstream systems to call and execute.

[0052] Compared with existing technologies, this application achieves automatic association of data fields, intelligent template matching, and dynamic parameter filling, thereby improving task generation efficiency and risk response speed.

[0053] In some optional implementations of the embodiments of this application, the step of decomposing the event task according to the task orchestration model to obtain N event subtasks specifically includes the following steps: Determine whether the event task is a standard event; If the event task is a standard task, then read the decomposition template database and retrieve the target decomposition template that matches the event task from the decomposition template database; The event tasks are decomposed into event subtasks based on the target decomposition template. If the event task is a non-standard task, then the event features are extracted based on the pre-trained model to obtain the event feature vector; The decomposition strategy is obtained based on the event feature vector, and the event task is decomposed according to the strategy to obtain the event subtask.

[0054] In this application embodiment, this application determines whether an event task is a standard event based on the task type field (such as task_type) or task characteristics (such as parameter structure, keywords). Standard events refer to task types that are pre-registered in the task type library (such as "temperature over-threshold alarm" and "order anomaly processing"), and their decomposition rules have been defined. Non-standard events refer to task types that are not registered in the task type library or that require dynamic analysis (such as "custom equipment maintenance" and "emergency fault troubleshooting").

[0055] In this embodiment, if the event task is a standard task, this application retrieves a target decomposition template matching the task type from the decomposition template database (e.g., an SQL query SELECT * FROM decomposition_templates WHERE task_type="over_temp_alert"). Then, based on the target decomposition template, the event task is decomposed into N subtasks (e.g., generating subtask 1 {"subtask_id": "ST001", "action": "check_cooling_system", "threshold": 88}, and subtask 2 {"subtask_id": "ST002", "action": "notify_operator", "message": "Equipment A001 temperature exceeds limit"}). The target decomposition template includes a list of subtasks, parameter mapping rules (e.g., mapping the parent task parameter temp to the threshold field of the subtask check_cooling_system), and execution order (e.g., parallel / serial).

[0056] In this embodiment, if the event task is a non-standard task, this application extracts event feature vectors using a pre-trained model (such as BERT or Transformer). For example, the task description text "Device A001 experiences abnormal temperature due to sensor malfunction; the sensor needs to be checked and the device restarted" is converted into a 128-dimensional vector. Then, the feature vectors are input into a decomposition strategy generation model (such as a decision tree or neural network), which outputs a decomposition strategy (such as "Perform sensor diagnosis first, then perform device restart"). Finally, event subtasks are generated according to the decomposition strategy (such as subtask 1 {"subtask_id": "ST001", "action": "diagnose_sensor", "device_id": "A001"}, and subtask 2 {"subtask_id": "ST002", "action": "restart_device", "device_id": "A001", "condition": "sensor_fault=True"}). The decomposition strategy includes subtask type, parameter dependencies (such as subtask 2 requiring the output of subtask 1), and execution conditions (such as restarting the device only when the sensor malfunctions).

[0057] Compared with existing technologies, this application combines standard template matching with non-standard feature learning to achieve automated and efficient decomposition of event tasks, thereby improving system adaptability and task execution efficiency.

[0058] In some optional implementations of the embodiments of this application, the step of obtaining the decomposition strategy based on the event feature vector specifically includes the following steps: The event feature vector is compared with fuzzy matching rules in the rule knowledge base, and a decomposition strategy is output based on the matched target rule; or The system queries historical similar paths that match the event feature vectors in the knowledge graph and then decomposes them based on the generation strategy of the matched historical similar paths.

[0059] In this embodiment of the application, a fuzzy matching rule base is predefined, and the rules include the following elements: (1) Rule conditions: Fuzzy constraints based on feature vectors (e.g., "temperature is abnormal and sensor failure probability > 0.7"); (2) Decomposition strategy template: The strategy output when the conditions are met (e.g., "execute sensor diagnosis first, then execute device restart"). Then, this application calculates the similarity between the event feature vector and the rule conditions (such as cosine similarity, Euclidean distance); if the similarity exceeds the threshold (such as 0.8), the match is successful and the corresponding decomposition strategy template is output; if multiple rules are successfully matched, the rule with the highest similarity is selected or the final strategy is determined by weighted voting.

[0060] In this embodiment, the application can also construct a knowledge graph using historical event tasks as nodes and decomposition strategies as edges. Nodes contain event feature vectors and task parameters, and edges are labeled with decomposition strategies (e.g., "Event X → Strategy Y → Subtask 1, Subtask 2"). Then, the application queries the knowledge graph for the historical event node most similar to the current event feature vector (e.g., searching for Top-K similar nodes by vector similarity). The decomposition strategy path corresponding to the similar node is extracted (e.g., the decomposition strategy for the historical event "Device B002 sensor failure" is "diagnosis → restart"). Finally, the application adjusts the historical strategy according to the current task parameters (e.g., device ID, failure type) (e.g., replacing device ID with A001) to generate an adapted decomposition strategy.

[0061] Compared with existing technologies, this application achieves rapid and accurate decomposition of non-standard event tasks through a dual-path mechanism of fuzzy rule matching and historical path reuse, thereby improving the system's adaptive capabilities and task processing efficiency.

[0062] In some optional implementations of the embodiments of this application, the step of querying historical similar paths that match event feature vectors in the knowledge graph specifically includes the following steps: Acquire historical task data and store the sub-task sequences and sub-task parameters decomposed from the historical task data as structured path objects; Extract key features from each historical path of the structured path object to obtain a path feature vector; The similarity between event feature vectors and path feature vectors is calculated based on cosine similarity or Euclidean distance. The path feature vectors are filtered according to a preset similarity threshold to obtain a candidate path set; Acquire multi-dimensional environmental data corresponding to the triggering conditions of risk events; Based on the dimensional environment data, historical path matching is performed on the candidate path set to obtain historically similar paths.

[0063] In this embodiment of the application, historical task data is collected from the task execution system (such as industrial control system, operation and maintenance log). The data includes task description, sub-task sequence, sub-task parameters (such as device ID, operation type, parameter threshold) and execution result (success / failure). Then, each historical task data is converted into a structured path object.

[0064] In this application embodiment, the following key features are extracted from the structured path object: (1) Task characteristics: subtask type (e.g., diagnosis, restart), parameter distribution (e.g., hash value of device ID, parameter threshold range); (2) Environmental characteristics: continuous or categorical values ​​such as temperature, humidity, time, network load (e.g., time period "day / night"); Then, One-Hot encoding is used for categorical features (such as subtask types); continuous features (such as temperature) are normalized to obtain encoded features; finally, the encoded features are concatenated into a fixed-dimensional path feature vector (such as a 256-dimensional vector [0.12, 0.45, ..., 0.89]).

[0065] In this embodiment of the application, the application generates an event feature vector for the current event task (such as "device A001 sensor failure causing abnormal temperature") using a pre-trained model (such as BERT); then, it calculates the similarity between the event feature vector and all path feature vectors using cosine similarity or Euclidean distance; finally, based on a preset similarity threshold (such as 0.8), it filters out path feature vectors with similarity higher than the threshold to obtain a candidate path set Set_{candidate}.

[0066] In this embodiment, the application can obtain multi-dimensional environmental data (such as real-time device temperature, humidity, and network status) related to the current event from sensors, logging systems, or external APIs; then, the environmental data is converted into constraints (such as "temperature > 40℃ and humidity < 70%), and expressed as a logical expression or numerical range; for each candidate path in the candidate path set Set_{candidate}, its historical environmental data Env_{hist} is extracted; it is determined whether the current environmental data Env_{current} satisfies the constraints of Env_{hist} (such as Env_{current}.temp ∈ [Env_{hist}.temp_min, Env_{hist}.temp_max]); the paths that satisfy the constraints are retained to obtain the historical similar path set Set_{similar}.

[0067] Compared with existing technologies, this application achieves efficient screening and adaptation of historical paths through structured path modeling, feature vector matching and dynamic environment verification, thereby improving the quality of strategy generation in complex scenarios.

[0068] In some optional implementations of the embodiments of this application, the steps of generating multimedia materials and auxiliary description text based on the task description text of each event subtask specifically include the following steps: Entity recognition is performed on the task description text based on the pre-trained model to obtain the task description entities; The task description text is classified according to the intent classifier to obtain the multimedia material type. The multimedia material type includes at least data collection, experimental verification, solution demonstration and complex structure. When the multimedia material type is data acquisition, a two-dimensional drawing tool is used to generate a dynamic trend chart corresponding to the task description entity, thus obtaining the multimedia material; and / or When the multimedia material type is experimental verification, the video editing tool is called to generate a simulated video corresponding to the task description entity, thus obtaining the multimedia material; and / or When the multimedia material type is a solution demonstration type, the 3D drawing tool is invoked to generate a 3D process model corresponding to the task description entity, thus obtaining the multimedia material; and / or When the multimedia material type is a complex structure, a customized image tool is called to generate a high-definition schematic diagram corresponding to the task description entity, thus obtaining the multimedia material; By inputting task description entities and multimedia material types into a deep learning model, intelligent summaries are generated to obtain auxiliary descriptive text.

[0069] In this embodiment, a pre-trained Named Entity Recognition (NER) model (such as BERT-BiLSTM-CRF) is used to extract entities from a standardized text sequence. Then, key entities in the task description are identified, including device names (such as "device A001"), parameter types (such as "temperature"), and operation objects (such as "data"), generating an entity list Entities = [e1, e2, ..., e...]. n ].

[0070] In this embodiment, a pre-trained text classification model (such as TextCNN or RoBERTa) is used to classify the intent of standardized text sequences; then, the following four types of intents and their corresponding material types are defined: (1) Data acquisition: Dynamic trend charts (such as temperature change curves over time) need to be generated. (2) Experimental verification type: simulation videos (such as animations of chemical experiment processes) need to be generated; (3) Solution demonstration type: A three-dimensional process model (such as a three-dimensional diagram of mechanical assembly process) needs to be generated; (4) Complex structures: High-definition schematic diagrams (such as exploded views of the internal structure of a circuit board) need to be generated. Finally, determine the multimedia material type Type ∈ {Data Acquisition Class, Experiment Verification Class, Solution Demonstration Class, Complex Structure Class} for the current task description.

[0071] In this embodiment of the application, the application calls the corresponding tool to generate materials according to the material type (Type): (1) Data Acquisition Category: Input: A list of entities (e.g., [Device A001, Temperature]); Tools: Two-dimensional plotting tools (such as Matplotlib, ECharts); Output: Generate dynamic trend charts (such as "Temperature Trend Chart of Equipment A001 (2025-01-01 to 2025-01-07)").

[0072] (2) Experimental verification type: Input: A list of entities (e.g., [chemical reaction A, reagent B]); Tools: Video editing tools (such as Blender, Adobe After Effects); Output: Generate a simulation video (e.g., "Simulation of the addition of reagent B in chemical reaction A (30 seconds)").

[0073] (3) Solution presentation type: Input: A list of entities (e.g., [robotic arm, assembly steps]); Tools: 3D drawing tools (such as SolidWorks, Unity 3D); Output: Generate a 3D process model (e.g., "3D demonstration of robotic arm assembly steps (including interactive rotation function)").

[0074] (4) Complex structure class: Input: A list of entities (e.g., [circuit board, chip]); Tools: Customized image tools (such as Photoshop scripts, CAD software); Output: Generate a high-definition schematic diagram (such as "High-definition exploded view of circuit board chip layout (resolution 4000×3000)").

[0075] In this embodiment of the application, a pre-trained text generation model (such as T5 or GPT-3.5) is used as an intelligent summary generator. Then, the entity list and the material type (such as "equipment A001 temperature data acquisition class") are concatenated as input data for the text generation model to generate semantically enhanced auxiliary descriptive text.

[0076] Compared with existing technologies, this application achieves dynamic adaptation of material types and semantic enhancement of auxiliary text through entity recognition, intent classification and toolchain integration, thereby improving the accuracy and comprehensibility of task demonstrations.

[0077] In some optional implementations of the embodiments of this application, after the steps of generating multimedia materials and auxiliary description text based on the task description text of each event subtask, the following steps are further included: Obtain user profiles of target users corresponding to the target product; User profile features are extracted from the user profile to obtain user profile features; The association rule mining method is used to obtain product preferences corresponding to user profile features, and a mapping relationship library between user profiles and displayed content elements is constructed based on product preferences. Based on the mapping relationship library, the multimedia materials are visually adapted and / or their content is reconstructed in a contextualized manner to obtain optimized multimedia materials. The optimized auxiliary description text is obtained by prioritizing the selling points of the auxiliary description text and / or adapting it with emotional language based on the mapping relationship library.

[0078] In this application embodiment, the application obtains target user profile data related to the target product from a user database or a real-time interactive interface; wherein, the profile data type includes, but is not limited to, basic attributes (age, gender, region), behavioral data (browsing history, purchase records), interest tags (technology enthusiasts, environmentalists), psychological characteristics (price-sensitive, quality-seeking), etc.

[0079] In this embodiment, feature encoding techniques (such as One-Hot encoding and word embedding) are used to vectorize user profile data. Then, through feature importance analysis (such as random forest feature weights) or business rule filtering, features strongly correlated with product preferences (such as "age → design style preference" and "purchase history → functional requirement priority") are retained. Finally, the user profile feature vector FeatureVector = [f1, f2, ..., f m ].

[0080] In this embodiment, the application uses the Apriori or FP-Growth algorithm to mine association rules between features and product preferences (such as "age∈[18,25] ∧ interest tag=games→preference for cool visual style"), and obtains the association rule set Rules = {Rule1, Rule2, ..., Rule...} k Then, based on association rules, this application establishes MappingDB, a mapping database between user profile features and displayed content elements, including: (1) Visual style mapping: such as "young users → high saturation colors, dynamic effects"; "professional users → simple lines, data visualization"; (2) Content scene mapping: such as "environmentalists → highlighting product sustainability scenarios"; "technology enthusiasts → showcasing animations of technical principles"; (3) Textual selling point mapping: such as "price-sensitive users → prioritize presenting cost-effective selling points"; "quality-seeking users → emphasize material and craftsmanship details"; (4) Emotional language mapping: such as "impulsive users → use motivational words (such as "buy now")"; "rational users → use data-supported statements (such as "95% user positive reviews")".

[0081] In this embodiment of the application, visual style adaptation includes: (1) Color adjustment: Adjust the color scheme according to the age characteristics of users (e.g., young users → macaron colors, middle-aged and elderly users → calm colors). (2) Enhanced dynamic effects: Add interactive elements based on user interest tags (e.g., add click-to-zoom function to the trend chart for game enthusiasts).

[0082] In this embodiment of the application, content contextual reconstruction refers to reconstructing the background of materials based on the user's psychological characteristics (e.g., environmentalist → inserting a forest scene into the video).

[0083] In this embodiment of the application, the priority ranking of selling points refers to adjusting the order of selling points according to the user's focus (e.g., for price-sensitive users, "limited-time discount" is placed at the beginning).

[0084] In this embodiment of the application, emotional language adaptation refers to replacing words according to the user's psychological characteristics (e.g., for quality-oriented users, change "durable" to "military-grade material, 10-year warranty").

[0085] Compared with existing technologies, this application adds a user profile-driven optimization step to achieve dynamic adaptation of the visual style of multimedia materials and personalized adjustment of the language style of auxiliary text, thereby improving the target users' acceptance of the content and their willingness to take action.

[0086] The embodiments of this application can acquire and process relevant data based on artificial intelligence technology. Artificial intelligence (AI) refers to the theories, methods, technologies, and application systems that use digital computers or machines controlled by digital computers to simulate, extend, and expand human intelligence, perceive the environment, acquire knowledge, and use that knowledge to obtain optimal results.

[0087] Foundational technologies for artificial intelligence generally include sensors, dedicated AI chips, cloud computing, distributed storage, big data processing, operating / interactive systems, and mechatronics. AI software technologies mainly encompass computer vision, robotics, biometrics, speech processing, natural language processing, and machine learning / deep learning.

[0088] Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by instructing related hardware through computer-readable instructions. These computer-readable instructions can be stored in a computer-readable storage medium. When the program is executed, it can include the processes of the embodiments of the above methods. The aforementioned storage medium can be a non-volatile storage medium such as a magnetic disk, optical disk, or read-only memory (ROM), or random access memory (RAM).

[0089] It should be understood that although the steps in the flowcharts of the accompanying figures are shown sequentially as indicated by the arrows, these steps are not necessarily executed in the order indicated by the arrows. Unless explicitly stated herein, there is no strict order restriction on the execution of these steps, and they can be executed in other orders. Moreover, at least some steps in the flowcharts of the accompanying figures may include multiple sub-steps or multiple stages. These sub-steps or stages are not necessarily completed at the same time, but can be executed at different times, and their execution order is not necessarily sequential, but can be performed alternately or in turn with other steps or at least some of the sub-steps or stages of other steps.

[0090] Further reference Figure 3 As a response to the above Figure 2 The implementation of the method shown in this application provides an embodiment of an event-driven product data management device, which is similar to... Figure 2 Corresponding to the method embodiments shown, this device can be specifically applied to various electronic devices.

[0091] like Figure 3 As shown, the event-driven product data management device 200 of this application embodiment includes: Data acquisition module 210 is used to collect real-time data of the target product across the entire supply chain. The event task construction module 220 is used to construct an event task corresponding to the target product based on the full-link data when the full-link data is detected to meet the risk event triggering conditions. The task decomposition module 230 is used to call the task orchestration model and decompose the event task according to the task orchestration model to obtain N event subtasks, where N is an integer greater than zero. Subtask execution module 240 is used to generate multimedia materials and auxiliary description text based on the task description text of each event subtask; Product data construction module 250 is used to construct product display data corresponding to event tasks based on multimedia materials and auxiliary descriptive text; The product data output module 260 is used to output product display data to the user terminal corresponding to the target product.

[0092] This application provides an event-driven product data management device 200, comprising: a data acquisition module 210 for real-time acquisition of end-to-end data of a target product; an event task construction module 220 for constructing an event task corresponding to the target product based on the end-to-end data when the end-to-end data is detected to meet the triggering conditions of a risk event; a task decomposition module 230 for calling a task orchestration model and decomposing the event task into N event sub-tasks, where N is a positive integer; a sub-task execution module 240 for generating multimedia materials and auxiliary description text based on the task description text of each event sub-task; a product data construction module 250 for constructing product display data corresponding to the event task based on the multimedia materials and auxiliary description text; and a product data output module 260 for outputting product display data to a user terminal corresponding to the target product. Compared with the prior art, this application can acquire end-to-end data of a target product in real time, automatically construct event tasks and decompose tasks when the data is detected to meet the triggering conditions of a risk event, generate multimedia materials and auxiliary description text, and finally construct product display data and output it to the user terminal, thereby improving the efficiency and accuracy of product data management.

[0093] In some optional implementations of the embodiments of this application, the above-mentioned event task construction module includes: The associated field retrieval submodule is used to retrieve the associated fields corresponding to the target product from the end-to-end data. The task template acquisition submodule is used to read the task template database and retrieve the target task template corresponding to the type of risk event triggering condition from the task template database. The parameter population submodule is used to populate the parameters in the associated fields into the target task template to obtain the event task.

[0094] In some optional implementations of the embodiments of this application, the above-mentioned task decomposition module includes: The event task judgment submodule is used to determine whether an event task is a standard event; The decomposition template acquisition submodule is used to read the decomposition template database and retrieve the target decomposition template that matches the event task if the event task is a standard task. The template decomposition submodule is used to decompose event tasks based on the target decomposition template to obtain event subtasks. The event feature extraction submodule is used to extract event features from the event task based on the pre-trained model if the event task is a non-standard task, and obtain the event feature vector. The strategy decomposition submodule is used to obtain the decomposition strategy based on the event feature vector, and to decompose the event task into event subtasks based on the decomposition strategy.

[0095] In some optional implementations of the embodiments of this application, the above-mentioned strategy decomposition submodule includes: The rule matching unit compares the event feature vector with fuzzy matching rules in the rule knowledge base and outputs a decomposition strategy based on the matched target rule; or The historical path matching unit is used to query historical similar paths in the knowledge graph that match the event feature vector, and to generate and decompose the matched historical similar paths according to the generation strategy.

[0096] In some optional implementations of the embodiments of this application, the aforementioned historical path matching unit includes: The historical data acquisition subunit is used to acquire historical task data and store the sub-task sequence and sub-task parameters after the historical task data is decomposed into a structured path object. The key feature extraction subunit is used to extract key features from each historical path of the structured path object to obtain a path feature vector; The similarity calculation subunit is used to calculate the similarity between event feature vectors and path feature vectors based on cosine similarity or Euclidean distance. The path feature filtering subunit is used to filter path feature vectors according to a preset similarity threshold to obtain a candidate path set; The multi-dimensional environmental data acquisition subunit is used to acquire multi-dimensional environmental data corresponding to the triggering conditions of risk events; The historical path matching subunit is used to perform historical path matching on the candidate path set based on dimensional environment data to obtain historically similar paths.

[0097] In some optional implementations of the embodiments of this application, the above-mentioned subtask execution module includes: The entity recognition submodule is used to perform entity recognition on the task description text based on the pre-trained model to obtain the task description entities. The intent classification submodule is used to classify the task description text according to the intent classifier to obtain the multimedia material type. The multimedia material type includes at least data collection type, experimental verification type, solution demonstration type and complex structure type. The dynamic trend chart generation submodule is used to generate a dynamic trend chart corresponding to the task description entity when the multimedia material type is data acquisition, by calling a two-dimensional drawing tool to obtain the multimedia material; and / or The simulated video generation submodule is used to generate a simulated video corresponding to the task description entity when the multimedia material type is experimental verification, thereby obtaining the multimedia material; and / or The 3D process model generation submodule is used to generate a 3D process model corresponding to the task description entity when the multimedia material type is a solution presentation type, thereby obtaining the multimedia material; and / or The high-definition schematic diagram generation submodule is used to generate a high-definition schematic diagram corresponding to the task description entity when the multimedia material type is a complex structure class, thereby obtaining the multimedia material; The intelligent summary generation submodule is used to generate intelligent summaries from the input values ​​of task description entities and multimedia material types using a deep learning model, thus obtaining auxiliary descriptive text.

[0098] In some optional implementations of the embodiments of this application, the event-driven product data management device 200 described above further includes: The user profile acquisition module is used to acquire user profiles of target users corresponding to the target product. The user profile feature extraction module is used to extract user profile features to obtain user profile features. The mapping relationship library construction module is used to obtain product preferences corresponding to user profile features based on association rule mining, and to construct a mapping relationship library between user profiles and displayed content elements based on product preferences; The multimedia material optimization module is used to perform visual style adaptation and / or content scene reconstruction of multimedia materials based on the mapping relationship library to obtain optimized multimedia materials. The auxiliary description text optimization module is used to prioritize the selling points of the auxiliary description text and / or adapt it with emotional language based on the mapping relationship library to obtain the optimized auxiliary description text.

[0099] To address the aforementioned technical problems, embodiments of this application also provide a computer device. Please refer to [link / reference needed]. Figure 4 , Figure 4 This is a basic structural block diagram of a computer device according to an embodiment of this application.

[0100] Computer device 300 includes a memory 310, a processor 320, and a network interface 330 that are interconnected via a system bus. It should be noted that only computer device 300 with components 310-330 is shown in the figure; however, it should be understood that it is not required to implement all the shown components, and more or fewer components can be implemented alternatively. Those skilled in the art will understand that the computer device described here is a device capable of automatically performing numerical calculations and / or information processing according to pre-set or stored instructions, and its hardware includes, but is not limited to, microprocessors, application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), digital signal processors (DSPs), embedded devices, etc.

[0101] Computer devices can include desktop computers, laptops, handheld computers, and cloud servers. These devices allow for human-computer interaction with users through keyboards, mice, remote controls, touchpads, or voice-activated devices.

[0102] The memory 310 includes at least one type of readable storage medium, including flash memory, hard disk, multimedia card, card-type memory (e.g., SD or DX memory), random access memory (RAM), static random access memory (SRAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), programmable read-only memory (PROM), magnetic memory, magnetic disk, optical disk, etc. In some embodiments, the memory 310 may be an internal storage unit of the computer device 300, such as the hard disk or memory of the computer device 300. In other embodiments, the memory 310 may also be an external storage device of the computer device 300, such as a plug-in hard disk, smart media card (SMC), secure digital (SD) card, flash card, etc., equipped on the computer device 300. Of course, the memory 310 may include both internal storage units and external storage devices of the computer device 300. In the embodiments of this application, the memory 310 is typically used to store the operating system and various application software installed on the computer device 300, such as computer-readable instructions based on event-driven product data management methods. In addition, the memory 310 can also be used to temporarily store various types of data that have been output or will be output.

[0103] In some embodiments, processor 320 may be a central processing unit (CPU), controller, microcontroller, microprocessor, or other data processing chip. Processor 320 is typically used to control the overall operation of computer device 300. In embodiments of this application, processor 320 is used to execute computer-readable instructions stored in memory 310 or process data, such as executing computer-readable instructions for an event-driven product data management method.

[0104] The network interface 330 may include a wireless network interface or a wired network interface, which is typically used to establish a communication connection between the computer device 300 and other electronic devices.

[0105] The computer equipment provided in this application can collect the entire chain data of the target product in real time. When the data is detected to meet the triggering conditions of a risk event, it automatically constructs an event task and decomposes the task, generates multimedia materials and auxiliary descriptive text, and finally constructs product display data and outputs it to the user terminal, thereby improving the management efficiency and accuracy of product data.

[0106] This application also provides another embodiment, namely, providing a computer-readable storage medium storing computer-readable instructions that can be executed by at least one processor to cause the at least one processor to perform the steps of the event-driven product data management method described above.

[0107] The computer-readable storage medium provided in this application can collect end-to-end data of the target product in real time. When the data is detected to meet the triggering conditions of a risk event, it automatically constructs an event task and decomposes the task, generates multimedia materials and auxiliary descriptive text, and finally constructs product display data and outputs it to the user terminal, thereby improving the management efficiency and accuracy of product data.

[0108] Through the above description of the embodiments, those skilled in the art can clearly understand that the methods of the above embodiments can be implemented by means of software plus necessary general-purpose hardware platforms. Of course, they can also be implemented by hardware, but in many cases the former is a better implementation method. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product is stored in a storage medium (such as ROM / RAM, magnetic disk, optical disk) and includes several instructions to cause a terminal device (which may be a mobile phone, computer, server, air conditioner, or network device, etc.) to execute the methods of the various embodiments of this application.

[0109] Obviously, the embodiments described above are only some embodiments of this application, not all embodiments. The accompanying drawings show preferred embodiments of this application, but do not limit the patent scope of this application. This application can be implemented in many different forms; rather, the purpose of providing these embodiments is to provide a more thorough and comprehensive understanding of the disclosure of this application. Although this application has been described in detail with reference to the foregoing embodiments, those skilled in the art can still modify the technical solutions described in the foregoing specific embodiments, or make equivalent substitutions for some of the technical features. Any equivalent structures made using the content of this application's specification and drawings, directly or indirectly applied to other related technical fields, are similarly within the scope of patent protection of this application.

Claims

1. An event-driven product data management method, characterized in that, Includes the following steps: Real-time collection of end-to-end data for target products; When the end-to-end data is detected to meet the risk event triggering conditions, an event task corresponding to the target product is constructed based on the end-to-end data. The task orchestration model is invoked, and the event task is decomposed according to the task orchestration model to obtain N event subtasks, where N is an integer greater than zero; Multimedia materials and auxiliary descriptive text are generated based on the task description text of each event subtask; Based on the multimedia materials and the auxiliary descriptive text, construct product display data corresponding to the event task; The product display data is output to the user terminal corresponding to the target product.

2. The event-driven product data management method according to claim 1, characterized in that, The step of constructing an event task corresponding to the target product based on the end-to-end data specifically includes the following steps: Obtain the associated fields corresponding to the target product from the end-to-end data; Read the task template database and retrieve the target task template corresponding to the type of the risk event triggering condition from the task template database; The parameters in the associated fields are filled into the target task template to obtain the event task.

3. The event-driven product data management method according to claim 1, characterized in that, The step of decomposing the event task according to the task orchestration model to obtain N event sub-tasks specifically includes the following steps: Determine whether the event task is a standard event; If the event task is a standard task, then the decomposition template database is read, and a target decomposition template matching the event task is obtained from the decomposition template database; The event task is decomposed according to the target decomposition template to obtain the event subtask; If the event task is a non-standard task, then the event feature vector is obtained by extracting event features from the pre-trained model. The decomposition strategy is obtained based on the event feature vector, and the event task is decomposed according to the decomposition strategy to obtain the event subtask.

4. The event-driven product data management method according to claim 3, characterized in that, The step of obtaining the decomposition strategy based on the event feature vector specifically includes the following steps: The event feature vector is compared with fuzzy matching rules in the rule knowledge base, and the decomposition strategy is output based on the matched target rule; or The decomposition strategy is generated based on the historical similar paths that match the event feature vector in the knowledge graph.

5. The event-driven product data management method according to claim 4, characterized in that, The step of querying historical similar paths in the knowledge graph that match the event feature vector specifically includes the following steps: Acquire historical task data, and store the sub-task sequence and sub-task parameters decomposed from the historical task data as structured path objects; Key features are extracted from each historical path of the structured path object to obtain a path feature vector; The similarity between the event feature vector and the path feature vector is calculated based on cosine similarity or Euclidean distance. The path feature vectors are filtered according to a preset similarity threshold to obtain a candidate path set; Obtain multi-dimensional environmental data corresponding to the risk event triggering conditions; Based on the dimensional environment data, the candidate path set is matched with historical paths to obtain the historical similar paths.

6. The event-driven product data management method according to claim 1, characterized in that, The step of generating multimedia materials and auxiliary descriptive text based on the task description text of each event subtask specifically includes the following steps: The task description text is subjected to entity recognition based on the pre-trained model to obtain the task description entities; The task description text is classified according to the intent classifier to obtain the multimedia material type, wherein the multimedia material type includes at least data collection type, experimental verification type, solution demonstration type and complex structure type; When the multimedia material type is data acquisition, a two-dimensional drawing tool is invoked to generate a dynamic trend chart corresponding to the task description entity, thereby obtaining the multimedia material; and / or When the multimedia material type is experimental verification type, a video editing tool is invoked to generate a simulated video corresponding to the task description entity, thereby obtaining the multimedia material; and / or When the multimedia material type is a solution demonstration type, a 3D drawing tool is invoked to generate a 3D process model corresponding to the task description entity, thereby obtaining the multimedia material; and / or When the multimedia material type is a complex structure class, a customized image tool is invoked to generate a high-definition schematic diagram corresponding to the task description entity, thereby obtaining the multimedia material; The task description entity and multimedia material type are input values ​​into a deep learning model to generate an intelligent summary, thus obtaining the auxiliary description text.

7. The event-driven product data management method according to claim 1, characterized in that, After the step of generating multimedia materials and auxiliary description text based on the task description text of each of the event subtasks, the following steps are also included: Obtain user profiles of target users corresponding to the target product; The user profile is subjected to profile feature extraction to obtain user profile features; The product preferences corresponding to the user profile features are obtained using association rule mining, and a mapping relationship library between user profiles and displayed content elements is constructed based on the product preferences. Based on the mapping relationship library, the multimedia materials are visually adapted and / or their content is reconstructed in a contextualized manner to obtain optimized multimedia materials. Based on the mapping relationship library, the selling points of the auxiliary description text are prioritized and / or adapted with emotional language to obtain the optimized auxiliary description text.

8. An event-driven product data management device, characterized in that, include: The data acquisition module is used to collect real-time data from the entire supply chain of the target product. The event task construction module is used to construct an event task corresponding to the target product based on the end-to-end data when the end-to-end data is detected to meet the risk event triggering conditions. The task decomposition module is used to call the task orchestration model and decompose the event task according to the task orchestration model to obtain N event subtasks, where N is an integer greater than zero. The subtask execution module is used to generate multimedia materials and auxiliary description text based on the task description text of each event subtask; The product data construction module is used to construct product display data corresponding to the event task based on the multimedia materials and the auxiliary descriptive text. The product data output module is used to output the product display data to the user terminal corresponding to the target product.

9. A computer device, comprising a memory and a processor, characterized in that, The memory stores computer-readable instructions, and when the processor executes the computer-readable instructions, it implements the steps of the event-driven product data management method as described in any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores computer-readable instructions, which, when executed by a processor, implement the steps of the event-driven product data management method as described in any one of claims 1 to 7.

Images