A cloud computing platform-based supply chain big data management method and system

Abstract

This application provides a supply chain big data management method and system based on a cloud computing platform, relating to the field of supply chain big data management technology. It receives and standardizes raw data streams from all parties in the supply chain in real time, performs automatic semantic recognition and standardized mapping on the standardized supply chain data, and integrates it according to preset business logic to construct a supply chain information view. Based on this information view, it performs multi-dimensional intelligent analysis and calculations to generate supply chain demand forecasting results, dynamic scheduling schemes, and supply chain risk warning information. These intelligent decision-making results are transmitted in real time to the execution systems or operators at each stage of the supply chain, driving the execution actions at each stage and improving collaboration efficiency. This overcomes the shortcomings of existing technologies, such as delayed supply chain decision-making and mismatch between transportation capacity and warehouse resources, significantly improving the overall operational efficiency and risk resistance of the supply chain.

Description

Technical Field

[0001] This application relates to the field of supply chain big data management technology, and more specifically, to a supply chain big data management method and system based on a cloud computing platform. Background Technology

[0002] In today's globalized world, supply chain operations are increasingly complex, encompassing numerous elements such as orders, inventory, transportation capacity, goods in transit, and the external environment, forming a vast and interconnected network. To achieve efficient supply chain operation, enterprises urgently need to shift from optimizing single links to coordinating and managing the entire supply chain. The core of this lies in the real-time collection and integration of key information such as "people, vehicles, goods, and locations," achieving millisecond-level response and building a digital twin of the logistics system. This requires aggregating data from various channels, including e-commerce platforms, Enterprise Resource Planning (ERP) systems, Warehouse Management Systems (WMS), Order Management Systems (OMS), IoT sensors (such as GPS / BeiDou temperature and humidity positioning devices), and third-party logistics companies. The goal is to more accurately predict market demand, optimize cargo loading, plan the best transportation routes, thereby reducing transportation costs, increasing vehicle occupancy rates, achieving full-process cargo visibility, and quickly identifying and handling anomalies.

[0003] However, realizing this vision faces numerous challenges in cross-regional, multi-enterprise collaborative supply chain scenarios. For example, data from various participants, such as cargo owners, carriers, and logistics service providers, is typically stored independently in their respective systems, forming "data silos." This prevents the real-time sharing of critical information such as inventory levels, changes in market demand, and the status of goods in transit. This data fragmentation makes traditional data management platforms insufficiently scalable and lacking in real-time response capabilities when handling massive amounts of data, and it is difficult to effectively integrate data from different sources and in different formats. As a result, existing technologies struggle to provide comprehensive inventory forecasting for the entire supply chain, respond quickly to market demands, and dynamically adjust logistics arrangements, leading to delayed supply chain decisions, mismatches between transportation capacity and warehouse resources, and severely impacting collaboration efficiency. Summary of the Invention

[0004] This application provides a supply chain big data management method and system based on a cloud computing platform, which aims to solve the problems in cross-regional, multi-enterprise collaborative supply chain scenarios, such as data silos leading to the inability to share key information in real time, insufficient scalability and real-time response capabilities of traditional data management platforms, and difficulty in effectively integrating data from different sources and in different formats, which result in delayed supply chain decision-making, mismatch between transportation capacity and warehouse resources, and seriously affect collaboration efficiency.

[0005] On the one hand, this application provides a supply chain big data management method based on a cloud computing platform, including:

[0006] The system receives raw data streams from all parties in the supply chain in real time. These raw data streams include real-time sensing data, data from internal business systems, and data from external platforms. The system performs format checks and type identification on the raw data streams to obtain standardized supply chain data.

[0007] The supply chain data is automatically semantically identified and standardized and mapped, and then associated and integrated according to preset business logic to construct a globally updated supply chain information view in real time.

[0008] Based on the aforementioned supply chain information view, multi-dimensional intelligent analysis and calculation are performed to generate supply chain demand forecast results, dynamic scheduling schemes adapted to real-time operating conditions, and supply chain risk early warning information.

[0009] The supply chain demand forecast results, dynamic scheduling schemes, and supply chain risk warning information are transmitted in real time to the execution systems or operators at each link of the supply chain to drive the execution actions at each link of the supply chain, and to establish a mechanism to ensure information security and privacy in order to support cross-enterprise information sharing within the authorized scope.

[0010] On the other hand, this application provides a supply chain big data management system based on a cloud computing platform, the system comprising:

[0011] The data receiving and normalization module is used to receive raw data streams from all parties in the supply chain in real time. The raw data streams include real-time sensing data, data from internal business systems of the enterprise, and data from external platforms. The module performs format checks and type identification on the raw data streams to obtain standardized supply chain data.

[0012] The information view construction module is used to automatically identify and standardize the semantics of the supply chain data, and to associate and integrate it according to preset business logic to build a globally updated supply chain information view in real time.

[0013] The intelligent decision analysis module is used to perform multi-dimensional intelligent analysis and calculation based on the supply chain information view, generate supply chain demand forecast results, adapt dynamic scheduling schemes to real-time working conditions, and supply chain risk early warning information.

[0014] The decision execution driving module is used to transmit the supply chain demand forecast results, dynamic scheduling schemes, and supply chain risk warning information to the execution systems or operators at each link of the supply chain in real time, so as to drive the execution actions at each link of the supply chain and establish a mechanism to ensure information security and privacy, so as to support cross-enterprise information sharing within the authorized scope.

[0015] This application discloses a supply chain big data management method and system based on a cloud computing platform. By receiving and standardizing raw data streams from all parties in the supply chain in real time, it effectively solves the data integration problem caused by diverse data sources and inconsistent formats in traditional supply chains. Through automatic semantic recognition and standardized mapping of standardized supply chain data, and by integrating it according to preset business logic, a globally updated supply chain information view is constructed, breaking down the limitations of "data silos" and achieving real-time sharing and visualization of key supply chain information. Based on this, the method can perform multi-dimensional intelligent analysis and calculations based on the supply chain information view, generating accurate supply chain demand forecasts, dynamic scheduling schemes adapted to real-time operating conditions, and comprehensive supply chain risk warning information. This effectively solves the problems of existing technologies that struggle to perform global inventory forecasting for the entire supply chain, quickly respond to market demands, and dynamically adjust logistics arrangements. Ultimately, these intelligent decision-making results are transmitted in real time to the execution systems or operators at each stage of the supply chain, and information security and privacy protection mechanisms are established. This not only drives the execution actions at each stage of the supply chain and improves collaboration efficiency, but also supports cross-enterprise information sharing within the authorized scope. This overcomes the shortcomings of existing technologies, such as delayed supply chain decision-making and mismatch between transportation capacity and warehouse resources, and significantly improves the overall operational efficiency and risk resistance of the supply chain. Attached Figure Description

[0016] To illustrate this application more clearly, the accompanying drawings used in the embodiments will be briefly described below. Obviously, those skilled in the art can obtain other drawings based on these drawings without any creative effort.

[0017] Figure 1 The diagram illustrates a process flow diagram of a supply chain big data management method based on a cloud computing platform.

[0018] Figure 2 The diagram illustrates the structure of a supply chain big data management system based on a cloud computing platform.

[0019] Figure reference numerals: 100, Supply Chain Big Data Management System Based on Cloud Computing Platform; 10, Data Receiving and Standardization Module; 20, Information View Construction Module; 30, Intelligent Decision Analysis Module; 40, Decision Execution Driven Module. Detailed Implementation

[0020] The technical solutions of this application will now be clearly and completely described with reference to the accompanying drawings. Obviously, the described embodiments are merely some embodiments of this application, and not all embodiments. The components of this application described and shown in the accompanying drawings can generally be arranged and designed in various different configurations. Therefore, the following detailed description of the embodiments of this application provided in the accompanying drawings is not intended to limit the scope of the claimed application, but merely to illustrate selected embodiments of this application. All other embodiments obtained by those skilled in the art based on the embodiments of this application without inventive effort are within the scope of protection of this application.

[0021] It should be noted that similar reference numerals and letters in the following figures indicate similar items; therefore, once an item is defined in one figure, it does not need to be further defined and explained in subsequent figures. Furthermore, in the description of this application, terms such as "first," "second," etc., are used only to distinguish descriptions and should not be construed as indicating or implying relative importance.

[0022] To achieve efficient supply chain operation, enterprises urgently need to shift from optimizing single links to coordinating and managing the entire supply chain. The core of this lies in the real-time collection and integration of key information such as "people, vehicles, goods, and locations," achieving millisecond-level response and building a digital twin of the logistics system. This requires aggregating data from various channels, including e-commerce platforms, Enterprise Resource Planning (ERP) systems, Warehouse Management Systems (WMS), Order Management Systems (OMS), IoT sensors (such as GPS / BeiDou temperature and humidity positioning devices), and third-party logistics companies. The goal is to more accurately predict market demand, optimize cargo loading, and plan the best transportation routes, thereby reducing transportation costs, increasing vehicle occupancy rates, achieving full cargo visibility, and quickly identifying and handling anomalies. However, in cross-regional, multi-enterprise collaborative supply chain scenarios, realizing this vision faces numerous challenges. For example, data from cargo owners, carriers, and logistics service providers are typically stored independently in their respective systems, forming "data silos," preventing the real-time sharing of key information such as inventory levels, changes in market demand, and the status of goods in transit. This fragmented data structure makes traditional data management platforms insufficient in scalability and real-time response when processing massive amounts of data, and makes it difficult to effectively integrate data from different sources and in different formats.

[0023] like Figure 1 The diagram illustrates a flowchart of a supply chain big data management method based on a cloud computing platform. This application proposes a supply chain big data management method based on a cloud computing platform, comprising:

[0024] S10, receive raw data streams from all parties in the supply chain in real time. The raw data streams include real-time sensing data, data from the enterprise's internal business system, and data from external platforms. Perform format checks and type identification on the raw data streams to obtain standardized supply chain data.

[0025] Raw data streams refer to unprocessed initial data collected in real time from various stages of the supply chain. This data may include real-time sensing data from IoT sensors (such as GPS and temperature / humidity sensors), data from internal business systems such as ERP, WMS, and OMS, and data from external platforms such as e-commerce platforms and third-party logistics platforms.

[0026] S20, perform semantic automatic recognition and standardized mapping on the supply chain data, and associate and integrate it according to preset business logic to construct a globally updated supply chain information view in real time;

[0027] Among them, automatic semantic recognition and standardized mapping refers to automatically understanding the business meaning of supply chain data through technical means and converting it into a unified and standardized format so as to facilitate the integration and analysis of data from different sources.

[0028] Pre-defined business logic refers to a set of rules and processes pre-defined according to supply chain management needs, which are used to guide the association, integration and analysis of data.

[0029] A supply chain information view is a comprehensive, real-time updated supply chain data model that integrates all standardized supply chain data and presents it in a structured manner, providing a unified data foundation for subsequent intelligent analysis.

[0030] S30, based on the supply chain information view, perform multi-dimensional intelligent analysis and calculation to generate supply chain demand forecast results, dynamic scheduling schemes adapted to real-time working conditions, and supply chain risk warning information.

[0031] Among them, multi-dimensional intelligent analysis and computation refers to using advanced data analysis technologies (such as machine learning and artificial intelligence algorithms) to deeply mine the data in the supply chain information view and discover potential patterns and trends from multiple perspectives (such as time, region, product, supplier, etc.).

[0032] Supply chain demand forecasting results refer to the prediction of the demand for various products in the supply chain over a future period based on historical data and real-time information.

[0033] Dynamic scheduling schemes refer to plans that flexibly adjust and optimize resources such as logistics, production, and inventory in the supply chain based on real-time operating conditions and forecast results.

[0034] Supply chain risk warning information refers to identifying and informing customers in advance of potential risks in the supply chain (such as supplier disruptions, transportation delays, and sudden drops in demand) and their potential impact.

[0035] S40, The supply chain demand forecast results, dynamic scheduling plan and supply chain risk early warning information are transmitted in real time to the execution system or operators of each link in the supply chain to drive the execution actions of each link in the supply chain, and to establish a mechanism to protect information security and privacy to support cross-enterprise information sharing within the authorized scope.

[0036] Among them, the execution system or operator refers to the automated system (such as automated warehouse system, transportation management system) or human operator responsible for specific operations in each link of the supply chain.

[0037] Mechanisms for ensuring information security and privacy refer to the technical and management measures taken to protect supply chain data from unauthorized access, use, disclosure, damage, or tampering.

[0038] Cross-enterprise information sharing within the scope of authorization refers to allowing different enterprises in the supply chain to share necessary information according to preset permission rules, while ensuring data security and privacy, in order to promote collaborative efficiency.

[0039] The core of the supply chain big data management method based on cloud computing platform proposed in this application lies in building an efficient and intelligent supply chain data processing and decision support system.

[0040] First, in the process of receiving raw data streams from various parties in the supply chain in real time and performing format checks and type identification on these raw data streams to obtain standardized supply chain data, the methods for receiving the raw data streams can be diverse. For example, message queue services (such as Kafka and RabbitMQ) can be used to collect data from IoT sensors, internal enterprise systems, and external platforms in real time. This data may exist in different formats, such as JSON, XML, CSV, or database records. After receiving the data, it is necessary to perform format checks to ensure that it conforms to a preset data model. For example, a set of data validation rules can be defined to check the data type, length, and value range of each field. At the same time, type identification is also required to determine which business type the data belongs to, such as order data, inventory data, or transportation data. This can be achieved through data source identification, data content feature matching, or metadata analysis. For example, data from a WMS system is usually identified as inventory data, while data from an e-commerce platform may be identified as order data. Through these processes, raw, heterogeneous data can be transformed into standardized supply chain data with a unified format and clear type.

[0041] Secondly, in the steps of automatically recognizing and standardizing the semantics of the supply chain data, and then associating and integrating it according to preset business logic to construct a globally updated supply chain information view, automatic semantic recognition is crucial. For example, for product names in "formatted supply chain data," different companies may have different naming methods (e.g., "screwdriver" and "Phillips head screwdriver" may refer to the same product). In this case, Natural Language Processing (NLP) technology, combined with preset industry dictionaries and ontology libraries, can be used to automatically identify the same business semantics under these different expressions. Standardization mapping maps the identified business semantics to unified standardized codes or identifiers. For example, all expressions referring to "screwdriver" are mapped to a unified product ID. After completing the standardization mapping, these standardized data need to be associated and integrated according to preset business logic. For example, order data can be associated with corresponding inventory data and transportation data to form a complete order fulfillment chain. These association and integration operations can be based on relational database join operations or on graph databases to build complex entity relationship networks. Through continuous association and integration, a "globally updated supply chain information view" that can reflect the overall status of the supply chain in real time is ultimately formed.

[0042] Furthermore, multi-dimensional intelligent analysis is the core of the steps involved in performing multi-dimensional intelligent analysis based on the aforementioned supply chain information view to generate supply chain demand forecast results, dynamic scheduling schemes adapted to real-time operating conditions, and supply chain risk early warning information. For example, machine learning models (such as time series forecasting models ARIMA and LSTM) can be used to analyze historical sales data, market trend data, and inventory and in-transit data in the supply chain information view to generate "supply chain demand forecast results." When generating dynamic scheduling schemes, various factors can be considered. For example, optimization algorithms (such as linear programming and genetic algorithms) can be combined with real-time traffic conditions, warehouse capacity, vehicle availability, and other information to generate optimal transportation routes, cargo loading schemes, or production plans, thereby obtaining "dynamic scheduling schemes adapted to real-time operating conditions." For supply chain risk early warning, a risk assessment model can be constructed to continuously monitor key indicators in the supply chain information view (such as supplier delivery delay rates, abnormal inventory levels, and changes in macroeconomic indicators). Once an anomaly or potential risk is detected, an early warning mechanism is immediately triggered to generate "supply chain risk early warning information." For example, when a supplier of a key raw material experiences a production interruption, an early warning can be issued immediately.

[0043] Finally, in the process of transmitting the supply chain demand forecast results, dynamic scheduling plans, and supply chain risk warning information in real time to the execution systems or operators at each stage of the supply chain to drive their actions, and establishing mechanisms to ensure information security and privacy to support cross-enterprise information sharing within authorized scope, the methods of information transmission can be diversified. For example, the "supply chain demand forecast results" can be sent directly to the enterprise's ERP system via API, allowing the ERP system to automatically adjust production plans. Alternatively, the "dynamic scheduling plan" and "supply chain risk warning information" can be pushed to logistics operators via mobile applications or email for manual confirmation and execution. To ensure information security and privacy, data encryption technologies (such as TLS / SSL protocol for transmission encryption and AES algorithm for storage encryption), access control policies (such as role-based access control (RBAC)), data anonymization technologies (such as anonymizing sensitive customer information), and blockchain technology can be used to ensure the immutability and traceability of data. Through these mechanisms, "cross-enterprise information sharing within authorized scope" can be achieved; for example, suppliers can be allowed to view demand forecasts for the products they are responsible for, but cannot access data from other suppliers.

[0044] This application leverages the powerful data processing capabilities and intelligent analysis technology of cloud computing platforms to construct a complete closed loop from data collection, integration, analysis to decision execution. It effectively addresses core pain points in traditional supply chain management, such as data fragmentation, slow response, and delayed decision-making, providing an innovative solution for achieving efficient, intelligent, and secure supply chain management.

[0045] In some embodiments, the step of associating and integrating information based on preset business logic to construct a globally updated supply chain information view includes:

[0046] When the general fields in the supply chain data, combined with additional information, can be parsed into multiple business semantic interpretations, the semantic resolution process is triggered.

[0047] In the semantic resolution process, preset product risk and compliance rules are queried, and the latest operation logs and external scenario information associated with product batches in the supply chain data are obtained; based on the preset product risk and compliance rules, the latest operation logs, and the external scenario information, risk assessments are performed on each refined business semantic interpretation; the refined business semantic interpretation corresponding to the highest risk avoidance or the most stringent compliance requirements is selected, and standardization processing is completed based on it, and the standardization results are output.

[0048] If it can only be parsed into a single business semantic interpretation, then the standardization process is completed directly using that single business semantic interpretation and the standardization result is output.

[0049] The standardized results are linked and integrated according to preset business logic to construct a globally updated supply chain information view.

[0050] Specifically, when common fields in supply chain data, such as product name, batch number, or material code, combined with their additional information, such as unit of measurement, production date, or supplier code, can be parsed into two or more different business semantic interpretations, the semantic resolution process will be automatically triggered. For example, a material code may represent both raw materials and semi-finished products in different business scenarios, in which case semantic resolution is required.

[0051] Once the semantic resolution process is triggered, it first queries pre-defined product risk and compliance rules. These rules may include industry standards, national regulations, and internal corporate policies, aiming to ensure the safety and legality of products during production, storage, transportation, and use. Simultaneously, it retrieves the latest operation logs associated with specific product batches in the supply chain data, such as warehousing records, outbound records, and production processing records for that batch, as well as external contextual information, such as market regulatory updates, industry warnings, and environmental changes. This information collectively provides comprehensive background data for subsequent risk assessment.

[0052] Furthermore, based on the pre-defined product risk and compliance rules, the latest operational logs, and external scenario information, a risk assessment will be conducted for each refined business semantic interpretation. Refined business semantic interpretation refers to the in-depth analysis and definition of every specific business meaning that the raw data may represent. The risk assessment process will quantify the potential risks of each interpretation, such as financial losses, reputational damage, legal liability, or operational disruption, and evaluate its compliance level.

[0053] Based on this, a refined business semantic interpretation corresponding to the highest risk avoidance or the most stringent compliance requirements will be selected. This means that when multiple possible interpretations exist, those that can minimize risk or meet the most stringent regulatory requirements will be prioritized to ensure the robustness and legality of supply chain operations. Once selected, this interpretation will be used as the basis for standardizing supply chain data and outputting standardized results. Standardization aims to unify data from different sources and in different formats into a format that conforms to preset specifications, facilitating subsequent correlation, integration, and analysis.

[0054] If, during the initial parsing phase, the supply chain data can only be parsed into a single business semantic interpretation, i.e., without any ambiguity, then the standardization process will be completed directly using this single business semantic interpretation, and the standardized result will be output without triggering a complex semantic resolution process.

[0055] Finally, the standardized results are integrated and correlated according to preset business logic to construct a globally updated supply chain information view. The preset business logic defines the relationships and processing rules between different data elements in the supply chain, ensuring the integrity, consistency, and real-time nature of the information view.

[0056] Through the aforementioned technical solution, this application significantly improves the accuracy and reliability of semantic recognition in supply chain data, especially when dealing with complex or ambiguous data. By introducing a semantic resolution mechanism based on risk and compliance assessment, this solution effectively avoids misinterpretations and potential business risks caused by data ambiguity. Compared to basic solutions, this application ensures that the constructed supply chain information view more accurately reflects the actual business context, thereby providing a high-quality data foundation for subsequent intelligent analysis and computation, reducing operational and compliance risks caused by data misunderstandings, and ultimately improving the overall decision-making quality and response speed of the supply chain.

[0057] In some embodiments, the steps of performing multi-dimensional intelligent analysis and calculation based on the supply chain information view to generate supply chain demand forecasting results, dynamic scheduling schemes adapted to real-time operating conditions, and supply chain risk early warning information include:

[0058] It can receive and analyze macro-situational events from global news agencies, government announcements, and professional analysis reports in real time, and obtain professional judgment conclusions related to macro-situational events from authoritative expert databases.

[0059] The macro-scenario events and the professional judgment conclusions are correlated with the data in the supply chain information view to identify the geographical overlap and dependency between the macro-scenario events and key nodes in the supply chain, and the correlation analysis results are obtained.

[0060] Based on the correlation analysis results, the chain reaction effect of supply chain disruption is deduced, and the propagation path and impact of supply chain disruption under different macro-scenarios are simulated to obtain the deduction results;

[0061] Based on the simulation results, supply chain demand forecasts are generated, dynamic scheduling schemes adapted to real-time operating conditions are generated, and supply chain risk warning information is generated. The supply chain risk warning information includes the impact of the macro-scenario event on specific links of the supply chain, the expected duration, and the potential economic losses.

[0062] Specifically, the real-time reception and analysis of macroeconomic scenario events involves continuously monitoring and acquiring information on major events globally from various public or subscribed data sources. These data sources can include real-time news streams from global news organizations such as Reuters and the Associated Press, policy announcements, trade restrictions, or pandemic updates issued by government departments of various countries, and industry analysis reports and economic forecasts published by professional consulting firms such as McKinsey and PwC. Simultaneously, it also retrieves professional assessments related to these macroeconomic scenario events from authoritative expert databases, such as geopolitical experts' assessments of the impact of conflicts in a particular region, climate scientists' predictions of extreme weather events, and economists' analyses of global economic trends. This information is then analyzed using Natural Language Processing (NLP) technology to extract key information and event elements.

[0063] The purpose of this analysis is to establish a correlation between changes in the external macroeconomic environment and internal supply chain operations by linking macroeconomic scenario events and expert assessments with data in the supply chain information view. Specifically, it identifies geographical overlaps between the location of macroeconomic scenario events and the locations of key production bases, logistics hubs, suppliers, or customers within the supply chain. For example, if a natural disaster occurs in a certain area, it immediately identifies whether there are important suppliers or transportation routes in that region. Furthermore, it analyzes the dependence of macroeconomic scenario events on various links in the supply chain; for example, a change in trade policy may affect the import of specific raw materials, thereby affecting the production of downstream products. Through this correlation analysis, a preliminary assessment of the interaction between events and the supply chain can be obtained, forming the correlation analysis results.

[0064] In practical applications, extrapolating the chain reaction effects of supply chain disruptions based on correlation analysis aims to predict potential cascading impacts. For example, if a key supplier is affected by a macroeconomic event, the simulation will demonstrate its impact on direct customers (such as manufacturers), then on the manufacturer's customers (such as retailers), and ultimately on end consumers. This process simulates the propagation path of supply chain disruptions under different macroeconomic scenarios, such as whether it is a localized disruption or a global contagion, and the degree of impact on different stages, such as whether it leads to production stoppages, logistical delays, or soaring costs. Through sophisticated simulation models and algorithms, detailed extrapolation results can be obtained, revealing potential weaknesses and risk exposures.

[0065] Furthermore, based on the simulation results, supply chain demand forecasts, dynamic scheduling schemes adapted to real-time operating conditions, and supply chain risk warning information are generated to provide forward-looking decision support. Supply chain demand forecasts will consider the potential impact of macroeconomic scenarios on market demand; for example, a pandemic may cause a surge or plunge in demand for certain goods. Dynamic scheduling schemes will, based on the simulation results, recommend adjustments to production plans, inventory strategies, transportation routes, or supplier selection to mitigate risks or seize opportunities. Supply chain risk warning information will specifically indicate the degree of impact of macroeconomic scenarios on specific links in the supply chain (such as raw material supply, manufacturing, logistics, and sales channels); for example, a 50% disruption in the supply of a certain raw material is expected to last for three months, and the potential economic losses will be quantified, such as the potential loss of millions of dollars in sales.

[0066] Through the aforementioned technical solution, this application significantly enhances the foresight and robustness of supply chain big data management methods in addressing external macroeconomic risks. By integrating global macroeconomic scenarios and expert assessments, it can identify potential supply chain disruption risks earlier and conduct in-depth simulations of their propagation paths and impact levels, thus avoiding decision-making errors caused by information lag or a single perspective in traditional methods. Consequently, the generated supply chain demand forecasts will be more realistic, dynamic scheduling schemes can more effectively mitigate risks or seize opportunities, and detailed supply chain risk early warning information can provide enterprises with quantitative and specific risk assessments, enabling them to formulate response strategies in advance, reduce potential economic losses, and enhance the overall resilience of the supply chain.

[0067] In some embodiments, the step of transmitting the supply chain demand forecast results, dynamic scheduling schemes, and supply chain risk early warning information in real time to the execution systems or operators at each stage of the supply chain to drive actions at each stage of the supply chain includes:

[0068] Obtain the identity and current business context of the recipient of the supply chain demand forecast results, dynamic scheduling scheme, and supply chain risk warning information;

[0069] Based on the identity identifier, query the preset permission configuration policy to determine the range of information that the recipient can access;

[0070] Based on the identity identifier and the business context, select one or more suitable information presentation templates from the preset information template library;

[0071] Based on the supply chain demand forecast results, dynamic scheduling scheme, supply chain risk warning information, and the information range accessible to the recipient, the content is trimmed, reorganized, and formatted according to the information presentation template to generate a context-adaptive information carrier.

[0072] The information carrier is pushed to the recipient in real time, and an interactive interface is provided for the recipient to confirm or perform operations.

[0073] Specifically, obtaining the identity and current business context of the recipient of the supply chain demand forecast results, dynamic scheduling plans, and supply chain risk warning information is mainly achieved by first identifying the unique identity information of the target recipient before information is pushed, such as their user ID, department, role, and permissions, and simultaneously sensing their current business status or specific work scenario, such as the task they are handling, their work location, or the business process they are currently focusing on. The purpose is to provide foundational data for subsequent personalized information processing.

[0074] Specifically, based on the identified identity, a preset permission configuration policy is queried to determine the scope of information accessible to the recipient. This can be understood as searching for corresponding access permission rules in a pre-defined permission management module based on the identified recipient's identity. These rules clearly specify which types and levels of information the recipient can view, operate, or receive. For example, senior managers may have access to global risk warning information, while frontline operators may only have access to dispatch instructions directly related to their responsibilities. The purpose is to ensure the compliance and security of information transmission, prevent the leakage of sensitive information, and avoid pushing irrelevant information to unauthorized personnel.

[0075] In practical applications, based on the identity identifier and the business context, one or more suitable information presentation templates are selected from a pre-defined information template library. Specifically, this involves comprehensively considering the recipient's identity characteristics (such as role, department) and their current business context (such as urgency, device type), intelligently selecting the most suitable template from a template library containing various predefined information presentation formats. For example, for risk warnings requiring immediate response, a concise and clear pop-up or SMS template might be chosen; for demand forecasts requiring detailed analysis, a report template including charts and data tables might be selected. The aim is to optimize the information display format, making it more in line with the recipient's reading habits and usage scenarios, thereby improving the readability and comprehension efficiency of the information.

[0076] Furthermore, based on the supply chain demand forecast results, dynamic scheduling plans, supply chain risk warning information, and the information accessibility scope of the recipient, the content is trimmed, reorganized, and formatted according to the information presentation template to generate a context-adapted information carrier. Specifically, the original forecast results, scheduling plans, and warning information are filtered according to the recipient's access permissions, retaining only the content that the recipient has the right to access and is highly relevant to the current business context—this is content trimming. Subsequently, these trimmed information items are logically reorganized and sorted to highlight key points, for example, placing the most urgent warning information at the top—this is reorganization. Finally, based on the selected information presentation template, the information's layout, font, color, and other visual elements are adjusted to conform to the specific display medium (such as PC interface, mobile APP, email, etc.)—this is formatting. Thus, the generated information carrier is highly personalized and context-adapted, maximizing the satisfaction of the recipient's needs.

[0077] Therefore, the information carrier is pushed to the recipient in real time, and an interactive interface is provided for the recipient to confirm or perform operations. Specifically, the generated context-adapted information carrier is immediately sent to the terminal device used by the recipient, such as through an internal enterprise messaging system, email, or mobile application notifications. Simultaneously, this information carrier typically includes one or more interactive elements, such as buttons or links for "Confirm Read," "Execute this schedule," or "Request more details," enabling the recipient not only to receive information but also to directly provide feedback or trigger subsequent business processes, thus forming a closed-loop information processing mechanism. Its purpose is to ensure timely information delivery and promote rapid response and effective processing by the recipient.

[0078] Through the aforementioned technical solutions, this application significantly improves the accuracy, security, and efficiency of supply chain information transmission. Specifically, by deeply understanding the recipient's identity and business context, combined with access control and intelligent template selection, personalized customization and contextual adaptation of information are achieved, effectively avoiding information overload and misunderstandings that may result from traditional general information push. As a result, recipients can understand and respond to critical information more quickly with lower cognitive load, thereby accelerating the decision-making process and improving the execution efficiency of each link in the supply chain. Furthermore, strict access control policies ensure the security and compliance of information transmission, reducing the risk of data leakage and misoperation. This highly optimized information transmission mechanism makes the supply chain more agile in response and more precise in resource scheduling, ultimately contributing to enhancing the resilience and competitiveness of the entire supply chain.

[0079] In some embodiments, the steps of performing multi-dimensional intelligent analysis and calculation based on the supply chain information view to generate supply chain demand forecasting results, dynamic scheduling schemes adapted to real-time operating conditions, and supply chain risk early warning information include:

[0080] Real-time monitoring and analysis of text data streams from social media platforms, e-commerce comment sections, and industry forums; extraction of keywords, sentiment values, and discussion popularity indices related to customized products; and capture of competitors' real-time promotional information to construct quantitative indicators that represent instantaneous market sentiment and competitive landscape.

[0081] Combining the quantitative indicators, the current inventory data, in-transit logistics data, production plan data, and historical sales data of customized products in the supply chain information view, the demand fluctuation patterns, peak cycles, and emergency replenishment characteristics of customized products are dynamically identified as demand patterns, and supply chain demand forecast results are generated through multi-scenario extrapolation and evaluation.

[0082] Real-time collection of actual sales data for customized products; comparison of actual sales data with supply chain demand forecast results; when the deviation exceeds a preset threshold, adjustment of feature weights in demand forecast calculation based on quantitative indicators; and dynamic adjustment of inventory safety thresholds and replenishment points related to the customized products to improve demand forecast accuracy and reduce inventory backlog and stockout risks.

[0083] Based on the revised supply chain demand forecast and the dynamically adjusted inventory safety threshold and replenishment points, priority is given to responding to instantaneous demand changes, generating dynamic scheduling schemes adapted to real-time operating conditions and supply chain risk warning information to meet the emergency replenishment needs of the customized products.

[0084] When generating the supply chain demand forecast results and dynamic scheduling schemes adapted to real-time operating conditions, sales opportunities and inventory risks are balanced, and optimal benefit decisions are executed to maximize the overall benefits of the supply chain.

[0085] Specifically, real-time monitoring and parsing of text data streams from social media platforms, e-commerce review sections, and industry forums refers to continuously acquiring publicly available, unstructured text information related to customized products, such as user comments, discussion posts, and news reports, through technologies like web crawlers and API interfaces. Keywords refer to words or phrases that reflect product characteristics, user concerns, or market trends; sentiment index refers to the analysis of text content using Natural Language Processing (NLP) technology to determine whether it is positive, negative, or neutral, and to quantify its intensity; discussion popularity index measures the frequency and breadth of mentions and discussions of a specific topic or product on social media, for example, calculated by counting the number of related posts, reposts, and comments. Capturing competitors' real-time promotional information can be understood as monitoring competitors' official websites, e-commerce platform stores, and social media accounts to obtain information on their product launches, price adjustments, and promotional activities. This information is integrated and constructed into quantitative indicators representing instantaneous market sentiment and competitive landscape, aiming to provide immediate and detailed insights into market dynamics.

[0086] Furthermore, by combining the aforementioned quantitative indicators with the current inventory data, in-transit logistics data, production plan data, and historical sales data of customized products in the supply chain information view, the demand fluctuation patterns, peak cycles, and emergency replenishment characteristics of customized products are dynamically identified as demand patterns. Demand fluctuation patterns refer to the sales volume trends of customized products over different time periods; peak cycles refer to specific periods where demand is significantly higher than the average level, such as holidays or specific marketing campaigns; emergency replenishment characteristics refer to the pattern of needing to quickly replenish inventory when demand suddenly increases or inventory is rapidly depleted. The identification of these demand patterns can be achieved through machine learning algorithms, such as time series analysis and cluster analysis. Generating supply chain demand forecasting results through multi-scenario extrapolation and evaluation refers to predicting the future demand for customized products using simulation models or forecasting algorithms under multiple market scenarios (e.g., optimistic, pessimistic, and neutral scenarios) to improve the robustness of the forecast.

[0087] Furthermore, real-time collection of actual sales data for customized products and comparison with supply chain demand forecasts establishes a feedback loop mechanism. When the deviation exceeds a preset threshold—for example, when actual sales are significantly higher or lower than forecasts—the feature weights in the demand forecast calculation are adjusted based on quantitative indicators. This means that if market sentiment or competitive landscape indicators show strong changes, their influence in the next forecast calculation will be strengthened or weakened to better adapt the forecasting model to the current market environment. Simultaneously, the inventory safety threshold and replenishment point associated with the customized products are dynamically adjusted. The inventory safety threshold refers to the amount of inventory held to cope with demand uncertainty; the replenishment point is the level at which a replenishment order needs to be triggered when inventory levels drop to that point. These dynamic adjustments aim to improve the accuracy of demand forecasts and reduce the risk of inventory backlog (excess inventory) and stockouts (lost sales opportunities) caused by inaccurate forecasts.

[0088] Therefore, based on the revised supply chain demand forecast and dynamically adjusted inventory safety thresholds and replenishment points, priority can be given to responding to instantaneous demand changes, generating dynamic scheduling schemes adapted to real-time operating conditions, as well as supply chain risk warning information. Dynamic scheduling schemes can include adjustments to production plans, optimization of logistics routes, and emergency changes to supplier orders to meet the urgent replenishment needs of customized products. Supply chain risk warning information can alert to potential risks such as supply disruptions, sudden drops or surges in demand. In generating the supply chain demand forecast and the dynamic scheduling scheme adapted to real-time operating conditions, sales opportunities and inventory risks are balanced; for example, ensuring market demand is met while avoiding cost waste caused by excessive inventory. By executing optimal benefit decisions, such as using optimization algorithms, the overall profitability of the supply chain can be maximized, ensuring that the company obtains the best economic benefits while meeting consumer demand.

[0089] Through the aforementioned technical solutions, this application significantly improves the accuracy and responsiveness of demand forecasting in the customized product supply chain. Specifically, by capturing instantaneous market sentiment and competitive landscape in real time, demand forecasting results more accurately reflect actual market demand, effectively avoiding biases caused by information lag in traditional forecasting methods. Simultaneously, dynamically adjusting inventory safety thresholds and replenishment points effectively reduces inventory backlog and stockout risks, optimizing inventory turnover efficiency. Furthermore, prioritizing responses to instantaneous demand changes and generating dynamic scheduling schemes adapted to real-time operating conditions ensures timely replenishment of customized products, meeting urgent market demands and seizing sales opportunities. Ultimately, by balancing sales opportunities and inventory risks and executing optimal benefit decisions, this application maximizes the overall profitability of the supply chain, bringing significant economic benefits and market competitiveness to the enterprise.

[0090] In some embodiments, the steps of real-time monitoring and parsing of text data streams from social media platforms, e-commerce review sections, and industry forums, extracting keywords, sentiment values, and discussion popularity indices related to customized products, and capturing real-time promotional information from competitors to construct quantitative indicators representing instantaneous market sentiment and competitive landscape include:

[0091] Select the corresponding language processing module based on the source region and language type of the text data stream;

[0092] The language processing module performs character set conversion, encoding standardization, and processing of punctuation marks and special characters in specific regions on the text data stream to obtain a preprocessed text data stream.

[0093] The preprocessed text data stream is subjected to a vocabulary and grammar rules based on a specific cultural context, and word segmentation and part-of-speech tagging are performed to obtain the word segmentation and part-of-speech tagging results.

[0094] Based on the word segmentation and part-of-speech tagging results, combined with the sentiment dictionary for specific cultural backgrounds and slang and colloquialism recognition rules, keywords are extracted from the text data stream, and the sentiment tendency of the text data stream is judged to obtain the sentiment tendency value.

[0095] The discussion intensity of the keywords and sentiment values ​​is calculated, and combined with the real-time promotional information of competitors, a quantitative indicator representing the instantaneous market sentiment and competitive situation is generated.

[0096] The selection of the appropriate language processing module based on the source region and language of the text data stream enables intelligent identification of its language attributes when receiving text data streams from different geographical regions or using different languages. It then automatically loads or invokes pre-configured processing modules optimized for that language or region. For example, for Chinese text, modules supporting Chinese word segmentation, part-of-speech tagging, and sentiment analysis might be selected; for English text, the corresponding English processing module would be chosen. The aim is to ensure that subsequent text processing accurately adapts to the grammar, vocabulary, and expression habits of different languages.

[0097] Furthermore, the language processing module performs character set conversion, encoding standardization, and punctuation and special character processing on the text data stream to obtain a preprocessed text data stream. Specifically, character set conversion and encoding standardization aim to solve the encoding inconsistencies that may exist between different data sources. For example, all text is uniformly converted to UTF-8 encoding to avoid garbled characters or parsing errors. Punctuation and special character processing in specific regions standardizes the usage habits of punctuation and special characters in different language or cultural backgrounds, such as full-width / half-width conversion, emoticon recognition and standardization, thereby providing clean and consistent input data for subsequent language analysis.

[0098] Based on this, a vocabulary and grammatical rules based on a specific cultural context are applied to the preprocessed text data stream to perform word segmentation and part-of-speech tagging, yielding the results. Word segmentation divides a continuous text sequence into word units with independent semantics, while part-of-speech tagging identifies the grammatical category of each word (such as noun, verb, adjective, etc.). Employing a vocabulary and grammatical rules based on a specific cultural context enables more accurate processing of industry-specific terminology, internet slang, or regional expressions, avoiding misjudgments caused by the limitations of general rules, thereby improving the depth and accuracy of text understanding.

[0099] Subsequently, based on the word segmentation and part-of-speech tagging results, and combined with a sentiment dictionary and slang / idiom recognition rules specific to a particular cultural context, keywords are extracted from the text data stream, and the sentiment tendency of the text data stream is judged to obtain a sentiment tendency value. Keyword extraction helps identify the core theme and focus of the text. Sentiment tendency judgment is achieved by analyzing the sentiment words, modifiers, and slang / idioms contained in the text to assess whether the user's attitude towards the customized product is positive, negative, or neutral. Combining a sentiment dictionary and slang / idiom recognition rules specific to a particular cultural context allows for more accurate capture of implicit emotional information, such as words or expressions with positive or negative connotations in a specific culture, thereby improving the accuracy and detail of sentiment analysis.

[0100] Finally, the discussion heat of the keywords and sentiment values ​​is calculated, and combined with real-time promotional information from competitors, a quantitative indicator representing instantaneous market sentiment and competitive landscape is generated. Discussion heat calculation can be based on a comprehensive evaluation of factors such as keyword frequency, interaction volume of related posts (e.g., likes, comments, and reposts), and user activity in the discussion, to reflect the level of attention a topic or product receives. Combining these internal data analysis results with externally obtained real-time promotional information from competitors allows for the comprehensive and dynamic construction of quantitative indicators reflecting current market sentiment and competitive landscape, providing data support for subsequent decision-making.

[0101] Through the aforementioned technical solutions, this application can significantly improve the accuracy and depth of processing unstructured text data such as social media and e-commerce reviews. Especially when dealing with complex data from multilingual and multicultural backgrounds, it can effectively avoid misunderstandings caused by language and cultural differences. Therefore, the extracted keywords, sentiment values, and discussion popularity indices will more accurately reflect users' true intentions and market sentiment. Furthermore, by combining real-time promotional information from competitors, this application can generate more comprehensive and dynamic quantitative indicators, providing more reliable and timely data support for demand forecasting and supply chain management of customized products. This effectively reduces the risk of market misjudgment and improves the responsiveness and decision-making quality of the supply chain.

[0102] In some embodiments, the steps of real-time monitoring, parsing text data streams from social media platforms, e-commerce review areas, and industry forums, extracting keywords, sentiment tendency values, and discussion heat indexes related to customized products, and capturing real-time promotion information of competitors to construct quantitative indicators representing instantaneous market sentiment and competitive situation include:

[0103] Identify the specific industry field to which the text data stream belongs, and load the corresponding industry field vocabulary and professional term ontology based on the specific industry field;

[0104] Preprocess the text data stream by removing common stop words and correcting industry-specific spelling mistakes to obtain a standardized industry text data stream;

[0105] Based on the industry field vocabulary and professional term ontology, perform word segmentation and entity recognition on the standardized industry text data stream, and extract industry-specific keywords and technical concepts;

[0106] Combine the corpus annotated by industry experts to judge the sentiment tendency of the keywords and technical concepts to distinguish neutral descriptions, technical affirmations, or negative evaluations in professional discussions, and obtain sentiment tendency values;

[0107] Calculate the discussion heat for the keywords and the sentiment tendency values, and combine the real-time promotion information of competitors to generate quantitative indicators representing instantaneous market sentiment and competitive situation.

[0108] Specifically, when identifying the specific industry field to which the text data stream belongs, various technical means can be adopted. For example, by analyzing the keywords and professional terms that frequently appear in the text data stream, or by judging from the source of the text data stream (such as a specific industry forum or professional media). Once the industry field is identified, the industry field vocabulary and professional term ontology highly relevant to this field will be loaded. Among them, the industry field vocabulary can be understood as a professional dictionary containing common words, phrases, and their semantics within a specific industry, while the professional term ontology is a structured knowledge representation used to describe concepts, entities, and their relationships within the industry. For example, in the medical industry, the ontology may include various diseases, drugs, treatment methods, and their associations.

[0109] Furthermore, when preprocessing the text data stream, in addition to removing common stop words (such as "of", "is", "in", etc.), spelling mistakes will also be corrected according to industry characteristics. For example, in the IT industry, some technical terms may have multiple non-standard spellings, which are corrected through preset industry-specific rules or machine learning models to ensure the accuracy of the text content. After preprocessing, the text data stream is transformed into a standardized industry text data stream, laying a foundation for subsequent in-depth analysis.

[0110] Building upon this foundation, and based on the loaded industry-specific lexicons and technical ontology, standardized industry text data streams are segmented and entity recognized. Lexicon segmentation divides a continuous text sequence into semantically meaningful word units, while entity recognition identifies entities with specific meanings within the text, such as product names, company names, and technical concepts. By utilizing industry-specific lexicons and ontology, specialized terms and technical concepts within the industry can be more accurately identified. For example, in the automotive industry, terms such as "turbocharging" and "autonomous driving assistance system" can be accurately identified.

[0111] In determining sentiment, this application incorporates a corpus annotated by industry experts. This corpus contains a large amount of data on industry-specific texts annotated by industry experts, enabling the model to learn and understand subtle differences in sentiment within professional discussions. For example, in technical forums, the word "bug" is generally negative in general contexts, but in some technical discussions, it may simply be a neutral description of a technical problem to be solved. By training with the expert corpus, the model can distinguish between neutral descriptions, positive technical assessments (such as "significant performance optimization"), and negative assessments (such as "serious product defects") in such professional discussions, thus obtaining more accurate sentiment values.

[0112] Finally, these professionally processed keywords and sentiment values ​​are used to calculate discussion volume, and combined with competitors' real-time promotional information, to generate quantitative indicators representing instantaneous market sentiment and competitive landscape. Discussion volume can be calculated based on factors such as keyword frequency, the number of reposts and comments on related texts, and user activity.

[0113] Through the aforementioned technical solutions, this application significantly enhances the ability to understand and analyze text data streams within specialized fields. Specifically, by identifying industry sectors and loading specialized vocabulary, the accuracy of keyword and technical concept extraction is ensured. Furthermore, by combining expert corpora for sentiment analysis, potential misjudgments of specialized terms in general contexts are effectively avoided, making the sentiment values ​​more realistic. Consequently, the constructed quantitative indicators can more accurately capture market sentiment and competitive dynamics within specific industries, providing more reliable and refined data support for supply chain demand forecasting and dynamic scheduling. This enhances the scientific rigor and effectiveness of decision-making and reduces the risks associated with information misjudgments.

[0114] In some embodiments, the step of calculating the discussion popularity of the keywords and the sentiment index includes:

[0115] Identify the interaction types related to the keywords and the sentiment values, including asking technical questions, sharing solutions, and disputes over product performance.

[0116] Based on the identified interaction type, the corresponding interaction weight is obtained from the preset weight configuration;

[0117] Based on the interaction weights, the frequency of occurrence of the keywords and the sentiment values, the number of times the text is forwarded or commented on, and the activity level of the relevant user groups are weighted and calculated to obtain the discussion popularity index.

[0118] Specifically, interaction types refer to the classification of user communication behaviors on social media platforms, e-commerce review sections, or industry forums regarding specific products or topics. These interaction types can be further subdivided into several categories. For example, technical questioning typically refers to users raising questions about product functions, technical specifications, compatibility, or usage methods, reflecting users' need for in-depth exploration of the product; solution sharing refers to users proactively providing solutions, operational experiences, or usage tips for specific problems or needs, embodying the spirit of mutual assistance and knowledge sharing within the community; product performance disputes may involve positive or negative evaluations and discussions by users regarding product quality, functionality, service experience, or price, directly reflecting user satisfaction or potential dissatisfaction. These interaction types can be identified and classified using Natural Language Processing (NLP) technology, combined with pre-set rule bases, keyword matching, semantic analysis, or machine learning models.

[0119] The preset weight configuration is a database or configuration table that stores the weights corresponding to different interaction types. For example, technical questions might be given a higher weight because they usually represent a user's deep concern about the product or a potential pain point, and are of significant reference value for product improvement; sharing solutions might also have a higher weight because it reflects the activity of the community and the spirit of mutual assistance among users, which helps to improve user stickiness; while the weight of product performance disputes may be dynamically adjusted based on their emotional tendency (positive or negative) and potential impact. For example, negative disputes might be given a higher weight to trigger an emergency response. These weights can be set by domain experts based on experience, or they can be optimized and adjusted through historical data analysis and machine learning methods to ensure that they accurately reflect the actual impact of different interaction types on market sentiment and competitive landscape.

[0120] In practical applications, weighted calculation refers to comprehensively calculating data from multiple dimensions, such as the frequency of keyword and sentiment values, the number of times the text is forwarded or commented on, and the activity level of relevant user groups, combined with the identified interaction weights. For example, for an interaction identified as a "technical question," the frequency of its keywords, the number of times it is forwarded, and the activity level of the user asking the question will be multiplied by a higher weight; while for a typical "product review," it may be multiplied by a relatively lower weight. This weighted calculation can more accurately reflect the contribution of different interactions to the overall discussion's popularity, thus obtaining a more representative discussion popularity index. The activity level of user groups can be quantified based on indicators such as user posting frequency, number of replies, online time, and influence score to comprehensively assess their activity level and voice in the community.

[0121] Through the aforementioned technical solution, this application enables a more refined assessment of market discussion intensity. By differentiating different interaction types and assigning them corresponding weights, it avoids generalizing all interactions, thus allowing the generated discussion intensity index to more accurately reflect user focus, the severity of issues, and the popularity of solutions. This weighted calculation method effectively improves the accuracy and representativeness of the discussion intensity index, providing a more reliable quantitative indicator for subsequent supply chain demand forecasting, dynamic scheduling, and risk warning, thereby enhancing the decision support capabilities of supply chain big data management methods.

[0122] In some embodiments, the steps of real-time monitoring and parsing of text data streams from social media platforms, e-commerce review sections, and industry forums, extracting keywords, sentiment values, and discussion popularity indices related to customized products, and capturing real-time promotional information from competitors to construct quantitative indicators representing instantaneous market sentiment and competitive landscape include:

[0123] Source identification and user behavior pattern analysis are performed on text data streams to identify users or accounts with abnormal behavior characteristics;

[0124] The text data published by the user or account is evaluated for content authenticity to obtain the content authenticity evaluation result. The content authenticity evaluation includes consistency checks between the text content and historical data, cross-validation of multi-source information, and semantic consistency analysis.

[0125] Based on the content authenticity assessment results, false information, malicious comments, or paid commenters in the text data stream are identified, and the text data stream is filtered or downgraded according to a preset reliability threshold.

[0126] For filtered or downweighted text data streams, extract keywords, sentiment values, and discussion popularity indices related to customized products, and capture real-time promotional information from competitors to construct quantitative indicators that characterize instantaneous market sentiment and competitive landscape.

[0127] The purpose of this study is to identify the source of text data streams and analyze user behavior patterns, aiming to filter out potentially unreliable information sources from massive amounts of text data. Source identification can include analyzing the publisher's IP address, geographical location, registration time, account type, etc. User behavior pattern analysis can monitor users' posting frequency, interaction patterns, repetitive comments, and the presence of abnormal forwarding or liking behaviors to identify users or accounts with abnormal behavioral characteristics, such as "water army" accounts, malicious attackers, or information dissemination bots.

[0128] Furthermore, assessing the authenticity of text data posted by the user or account is a crucial step in ensuring data reliability. This assessment includes: checking the consistency between the text content and historical data, i.e., comparing the current text content with the user's or related topics' past posts to identify any contradictions or significant deviations; cross-validation of multi-source information, which involves comparing key information mentioned in the text with information from other independent and authoritative sources to verify its authenticity; and semantic consistency analysis, which focuses on checking the logical coherence of the text content, the existence of self-contradictions, and whether its expressed semantics conform to common sense or industry context. Through these assessments, a quantitative content authenticity evaluation result can be obtained, such as a confidence score.

[0129] Based on the content authenticity assessment results, false information, malicious comments, or paid ranking manipulation in the text data stream are identified, and the text data stream is filtered or downweighted according to a preset confidence threshold. This means that text data assessed as having low authenticity can be completely removed from the analysis dataset (filtering) or have its weight reduced in subsequent calculations (downweighting), thereby reducing its impact on the final quantitative indicators. For example, a confidence threshold can be set; text data below this threshold will be filtered, while data above the threshold will be retained, or weighted according to the confidence score.

[0130] The aforementioned technical solutions can significantly improve the accuracy and reliability of quantitative indicators of market sentiment and competitive landscape in supply chain big data management methods. By filtering or downweighting unreliable text data, the interference of false information on supply chain demand forecasting and dynamic scheduling schemes can be effectively avoided, thereby reducing the risk of inventory backlog, stockouts, or lost market opportunities due to erroneous decisions. Furthermore, this solution helps maintain the purity of the supply chain information view, providing enterprises with more valuable decision-making basis, and ultimately optimizing the overall responsiveness and operational efficiency of the supply chain.

[0131] This application also proposes a supply chain big data management system based on a cloud computing platform, such as... Figure 2 As shown, a supply chain big data management system 100 based on a cloud computing platform includes:

[0132] The data receiving and normalization module 10 is used to receive raw data streams from all parties in the supply chain in real time. The raw data streams include real-time sensing data, data from the enterprise's internal business system, and data from external platforms. The module performs format checks and type identification on the raw data streams to obtain standardized supply chain data.

[0133] The information view construction module 20 is used to perform semantic automatic recognition and standardized mapping on the supply chain data, and to associate and integrate it according to preset business logic to construct a globally updated supply chain information view in real time.

[0134] The intelligent decision analysis module 30 is used to perform multi-dimensional intelligent analysis and calculation based on the supply chain information view, generate supply chain demand forecast results, adapt dynamic scheduling schemes to real-time working conditions, and supply chain risk early warning information.

[0135] The decision execution drive module 40 is used to transmit the supply chain demand forecast results, dynamic scheduling schemes and supply chain risk warning information to the execution systems or operators of each link in the supply chain in real time, so as to drive the execution actions of each link in the supply chain and establish a mechanism to protect information security and privacy, so as to support cross-enterprise information sharing within the authorized scope.

[0136] The cloud computing-based supply chain big data management system proposed in this application provides an innovative and efficient solution to the problems of data silos, insufficient real-time response capabilities, and difficulty in global optimization in traditional supply chain management. Through its modular architecture and intelligent processing capabilities, the cloud computing-based supply chain big data management system constructs a complete closed loop from data collection, integration, analysis to decision execution, effectively solving the core pain points of traditional supply chain management such as data dispersion, slow response, and delayed decision-making, and providing an innovative solution for achieving efficient, intelligent, and secure supply chain management.

[0137] The above description is merely an embodiment of this application and is not intended to limit the scope of protection of this application. Various modifications and variations can be made to this application by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the scope of protection of this application.

Claims

1. A supply chain big data management method based on a cloud computing platform, characterized in that, The method includes: The system receives raw data streams from all parties in the supply chain in real time. These raw data streams include real-time sensing data, data from internal business systems, and data from external platforms. The system performs format checks and type identification on the raw data streams to obtain standardized supply chain data. The supply chain data is automatically semantically identified and standardized and mapped, and then associated and integrated according to preset business logic to construct a globally updated supply chain information view in real time. Based on the aforementioned supply chain information view, multi-dimensional intelligent analysis and calculation are performed to generate supply chain demand forecast results, dynamic scheduling schemes adapted to real-time operating conditions, and supply chain risk early warning information. The supply chain demand forecast results, dynamic scheduling schemes, and supply chain risk warning information are transmitted in real time to the execution systems or operators at each link of the supply chain to drive the execution actions at each link of the supply chain, and to establish a mechanism to ensure information security and privacy in order to support cross-enterprise information sharing within the authorized scope.

2. The supply chain big data management method based on a cloud computing platform according to claim 1, characterized in that, The steps for constructing a globally updated supply chain information view by associating and integrating information based on preset business logic include: When the general fields in the supply chain data, combined with additional information, can be parsed into multiple business semantic interpretations, the semantic resolution process is triggered. In the semantic resolution process, preset product risk and compliance rules are queried, and the latest operation logs and external scenario information associated with product batches in the supply chain data are obtained; based on the preset product risk and compliance rules, the latest operation logs, and the external scenario information, risk assessments are performed on each refined business semantic interpretation; the refined business semantic interpretation corresponding to the highest risk avoidance or the most stringent compliance requirements is selected, and standardization processing is completed based on it, and the standardization results are output. If it can only be parsed into a single business semantic interpretation, then the standardization process is completed directly using that single business semantic interpretation and the standardization result is output. The standardized results are linked and integrated according to preset business logic to construct a globally updated supply chain information view.

3. The supply chain big data management method based on a cloud computing platform according to claim 1, characterized in that, The steps of performing multi-dimensional intelligent analysis and calculation based on the supply chain information view to generate supply chain demand forecasting results, dynamic scheduling schemes adapted to real-time operating conditions, and supply chain risk early warning information include: It can receive and analyze macro-situational events from global news agencies, government announcements, and professional analysis reports in real time, and obtain professional judgment conclusions related to macro-situational events from authoritative expert databases. The macro-scenario events and the professional judgment conclusions are correlated with the data in the supply chain information view to identify the geographical overlap and dependency between the macro-scenario events and key nodes in the supply chain, and the correlation analysis results are obtained. Based on the correlation analysis results, the chain reaction effect of supply chain disruption is deduced, and the propagation path and impact of supply chain disruption under different macro-scenarios are simulated to obtain the deduction results; Based on the simulation results, supply chain demand forecasts are generated, dynamic scheduling schemes adapted to real-time operating conditions are generated, and supply chain risk warning information is generated. The supply chain risk warning information includes the impact of the macro-scenario event on specific links of the supply chain, the expected duration, and the potential economic losses.

4. The supply chain big data management method based on a cloud computing platform according to claim 1, characterized in that, The steps of transmitting the supply chain demand forecast results, dynamic scheduling schemes, and supply chain risk early warning information in real time to the execution systems or operators at each stage of the supply chain to drive the actions of each stage of the supply chain include: Obtain the identity and current business context of the recipient of the supply chain demand forecast results, dynamic scheduling scheme, and supply chain risk warning information; Based on the identity identifier, query the preset permission configuration policy to determine the range of information that the recipient can access; Based on the identity identifier and the business context, select one or more suitable information presentation templates from the preset information template library; Based on the supply chain demand forecast results, dynamic scheduling scheme, supply chain risk warning information, and the information range accessible to the recipient, the content is trimmed, reorganized, and formatted according to the information presentation template to generate a context-adaptive information carrier. The information carrier is pushed to the recipient in real time, and an interactive interface is provided for the recipient to confirm or perform operations.

5. The supply chain big data management method based on a cloud computing platform according to claim 1, characterized in that, The steps of performing multi-dimensional intelligent analysis and calculation based on the supply chain information view to generate supply chain demand forecasting results, dynamic scheduling schemes adapted to real-time operating conditions, and supply chain risk early warning information include: Real-time monitoring and analysis of text data streams from social media platforms, e-commerce comment sections, and industry forums; extraction of keywords, sentiment values, and discussion popularity indices related to customized products; and capture of competitors' real-time promotional information to construct quantitative indicators that represent instantaneous market sentiment and competitive landscape. Combining the quantitative indicators, the current inventory data, in-transit logistics data, production plan data, and historical sales data of customized products in the supply chain information view, the demand fluctuation patterns, peak cycles, and emergency replenishment characteristics of customized products are dynamically identified as demand patterns, and supply chain demand forecast results are generated through multi-scenario extrapolation and evaluation. Real-time collection of actual sales data for customized products; comparison of actual sales data with supply chain demand forecast results; when the deviation exceeds a preset threshold, adjustment of feature weights in demand forecast calculation based on quantitative indicators; and dynamic adjustment of inventory safety thresholds and replenishment points related to the customized products to improve demand forecast accuracy and reduce inventory backlog and stockout risks. Based on the revised supply chain demand forecast and the dynamically adjusted inventory safety threshold and replenishment points, priority is given to responding to instantaneous demand changes, generating dynamic scheduling schemes adapted to real-time operating conditions and supply chain risk warning information to meet the emergency replenishment needs of the customized products. When generating the supply chain demand forecast results and dynamic scheduling schemes adapted to real-time operating conditions, sales opportunities and inventory risks are balanced, and optimal benefit decisions are executed to maximize the overall benefits of the supply chain.

6. The supply chain big data management method based on a cloud computing platform according to claim 5, characterized in that, The steps involved in real-time monitoring and parsing of text data streams from social media platforms, e-commerce review sections, and industry forums, extracting keywords, sentiment values, and discussion popularity indices related to customized products, and capturing real-time promotional information from competitors to construct quantitative indicators representing instantaneous market sentiment and competitive landscape include: Select the corresponding language processing module based on the source region and language type of the text data stream; The language processing module performs character set conversion, encoding standardization, and processing of punctuation marks and special characters in specific regions on the text data stream to obtain a preprocessed text data stream. The preprocessed text data stream is subjected to a vocabulary and grammar rules based on a specific cultural context, and word segmentation and part-of-speech tagging are performed to obtain the word segmentation and part-of-speech tagging results; Based on the word segmentation and part-of-speech tagging results, combined with the sentiment dictionary for specific cultural backgrounds and slang and colloquialism recognition rules, keywords are extracted from the text data stream, and the sentiment tendency of the text data stream is judged to obtain the sentiment tendency value. The discussion intensity of the keywords and sentiment values ​​is calculated, and combined with the real-time promotional information of competitors, a quantitative indicator representing the instantaneous market sentiment and competitive situation is generated.

7. The supply chain big data management method based on a cloud computing platform according to claim 5, characterized in that, The steps involved in real-time monitoring and parsing of text data streams from social media platforms, e-commerce review sections, and industry forums, extracting keywords, sentiment values, and discussion popularity indices related to customized products, and capturing real-time promotional information from competitors to construct quantitative indicators representing instantaneous market sentiment and competitive landscape include: Identify the specific industry sector to which the text data stream belongs, and load the corresponding industry sector vocabulary and professional term ontology based on the specific industry sector; The text data stream is preprocessed to remove common stop words and correct industry-specific spelling errors, resulting in a standardized industry text data stream. Based on the industry-specific vocabulary and professional terminology ontology, the standardized industry text data stream is segmented and entity recognized to extract industry-specific keywords and technical concepts. By combining a corpus annotated by industry experts, the keywords and technical concepts are evaluated for sentiment tendencies to distinguish between neutral descriptions, positive or negative evaluations in professional discussions, and to obtain sentiment tendency values. The discussion intensity of the keywords and sentiment values ​​is calculated, and combined with the real-time promotional information of competitors, a quantitative indicator representing the instantaneous market sentiment and competitive situation is generated.

8. The supply chain big data management method based on a cloud computing platform according to claim 7, characterized in that, The step of calculating the discussion popularity of the keywords and the sentiment index includes: Identify the interaction types related to the keywords and the sentiment values, including asking technical questions, sharing solutions, and disputes over product performance. Based on the identified interaction type, the corresponding interaction weight is obtained from the preset weight configuration; Based on the interaction weights, the frequency of occurrence of the keywords and the sentiment values, the number of times the text is forwarded or commented on, and the activity level of the relevant user groups are weighted and calculated to obtain the discussion popularity index.

9. The supply chain big data management method based on a cloud computing platform according to claim 5, characterized in that, The steps involved in real-time monitoring and parsing of text data streams from social media platforms, e-commerce review sections, and industry forums, extracting keywords, sentiment values, and discussion popularity indices related to customized products, and capturing real-time promotional information from competitors to construct quantitative indicators representing instantaneous market sentiment and competitive landscape include: Source identification and user behavior pattern analysis are performed on text data streams to identify users or accounts with abnormal behavior characteristics; The text data published by the user or account is evaluated for content authenticity to obtain the content authenticity evaluation result. The content authenticity evaluation includes consistency checks between the text content and historical data, cross-validation of multi-source information, and semantic consistency analysis. Based on the content authenticity assessment results, false information, malicious comments, or paid commenters in the text data stream are identified, and the text data stream is filtered or downgraded according to a preset reliability threshold. For filtered or downweighted text data streams, extract keywords, sentiment values, and discussion popularity indices related to customized products, and capture real-time promotional information from competitors to construct quantitative indicators that characterize instantaneous market sentiment and competitive landscape.

10. A supply chain big data management system based on a cloud computing platform, characterized in that, The system includes: The data receiving and normalization module is used to receive raw data streams from all parties in the supply chain in real time. The raw data streams include real-time sensing data, data from internal business systems of the enterprise, and data from external platforms. The module performs format checks and type identification on the raw data streams to obtain standardized supply chain data. The information view construction module is used to automatically identify and standardize the semantics of the supply chain data, and to associate and integrate it according to preset business logic to build a globally updated supply chain information view in real time. The intelligent decision analysis module is used to perform multi-dimensional intelligent analysis and calculation based on the supply chain information view, generate supply chain demand forecast results, adapt dynamic scheduling schemes to real-time working conditions, and supply chain risk early warning information. The decision execution driving module is used to transmit the supply chain demand forecast results, dynamic scheduling schemes, and supply chain risk warning information to the execution systems or operators at each link of the supply chain in real time, so as to drive the execution actions at each link of the supply chain and establish a mechanism to ensure information security and privacy, so as to support cross-enterprise information sharing within the authorized scope.

Images