Live broadcast e-commerce supply chain information monitoring method and platform

Abstract

The invention relates to the technical field of marketing monitoring, and particularly discloses a live broadcast e-commerce supply chain information monitoring method and platform, and the method comprises the steps: obtaining a monitoring port set of a live broadcast e-commerce supply chain, and collecting the real-time monitoring data of the live broadcast e-commerce supply chain in real time based on the monitoring port set, determining historical product live broadcast data, product prediction reference data and product presale data of each sales product in a sales product set of each live broadcast in the real-time live broadcast sales data in the real-time monitoring data; and determining real-time management data. According to the invention, collaborative acquisition and integration of full-link multi-port data can be realized, a precise association system of real-time data and historical data is established, scientificity and foresight of supply chain demand prediction are improved, targeted product presale data and real-time management data are generated, risks of inventory overstock, goods shortage and logistics mismatching are effectively reduced, and the system is suitable for large-scale popularization and application. The overall operation refinement level is improved, and the market competitiveness and the anti-risk capability of the live broadcast e-commerce supply chain are enhanced.

Description

Technical Field

[0001] This invention relates to the field of marketing monitoring technology, and in particular to a method and platform for monitoring supply chain information in live-streaming e-commerce. Background Technology

[0002] In the early stages of the live-streaming e-commerce industry, supply chain monitoring relied on manual statistics or single-stage data collection tools. Data from different points was isolated and lacked linkage, making it impossible to accurately correlate live-stream sessions with individual product data. Historical data reuse rates were extremely low, and there was a lack of effective sales trend prediction, leading to frequent inventory backlogs and stockouts. Mismatches between logistics capacity and sales demand were widespread, resulting in delayed after-sales response and overall low supply chain operational efficiency. As the live-streaming e-commerce market rapidly expanded, leading e-commerce platforms and large supply chain service providers were the first to upgrade, building end-to-end data collaborative monitoring systems. These systems enabled real-time data collection from multiple points, real-time correlation with historical data, and model-based predictive decision-making, adapting to the dynamic changes in live-streaming scenarios to some extent. However, insufficient cross-ecosystem data integration capabilities remain, failing to accurately pinpoint the entire lifecycle of each product sold, thus affecting prediction accuracy. The industry urgently needs a supply chain information monitoring method that can achieve end-to-end data collaboration and integrate real-time and historical data to overcome the limitations of traditional models.

[0003] Therefore, this invention proposes a method and platform for monitoring supply chain information in live-streaming e-commerce. Summary of the Invention

[0004] This invention provides a method and platform for monitoring supply chain information in live-streaming e-commerce. By acquiring a set of monitoring ports for the live-streaming e-commerce supply chain, real-time monitoring data, and historical live-streaming data for each live stream within the real-time sales data, the method determines historical live-streaming data, product forecasting reference data, and product pre-sale data for each product in the product set of each live stream within the real-time sales data, and also determines real-time management data. This enables collaborative collection and integration of multi-port data across the entire supply chain, establishes a precise correlation system between real-time and historical data, improves the scientific rigor and foresight of supply chain demand forecasting, generates targeted product pre-sale data and real-time management data, effectively reduces the risks of inventory backlog, stockouts, and logistics mismatches, enhances the overall operational sophistication, and strengthens the market competitiveness and resilience of the live-streaming e-commerce supply chain.

[0005] This invention provides a method for monitoring supply chain information in live-streaming e-commerce, comprising: S1: Obtain the set of monitoring ports for the live e-commerce supply chain, and collect real-time monitoring data of the live e-commerce supply chain based on the set of monitoring ports. The real-time monitoring data includes real-time live sales data, real-time supply data, real-time warehousing data, real-time after-sales data, and real-time logistics data. S2: Based on real-time monitoring data of the live-streaming e-commerce supply chain, obtain historical live-streaming data for each live stream in the real-time live-streaming sales data in the real-time monitoring data, and determine the historical live-streaming data of each product in the product set of each live stream in the real-time live-streaming sales data in the real-time monitoring data. S3: Based on the historical live streaming data of each product in the sales product set of each live stream in the real-time monitoring data and real-time live streaming sales data, determine the product prediction reference data and product pre-sale data of each product in the sales product set of each live stream in the real-time monitoring data and real-time live streaming sales data. S4: Based on real-time monitoring data and real-time live sales data, determine the pre-sale data of each product in the product set of each live broadcast.

[0006] Preferably, a method for monitoring supply chain information in live-streaming e-commerce includes a monitoring port set that includes the supplier end, warehousing end, logistics end, live-streaming sales end, and after-sales service end.

[0007] Preferably, a method for monitoring supply chain information in live-streaming e-commerce involves collecting real-time monitoring data of the live-streaming e-commerce supply chain based on a set of monitoring ports, including: The system collects real-time live sales data from the live e-commerce supply chain monitoring portals. This data includes a collection of products sold across multiple live streams, the product volume, weight, pricing, price changes, and order generation data for each product within that collection, as well as basic live stream data. Basic live stream data includes the live stream number, start time, estimated end time, and time stamp. Order generation data includes the order number, transaction price, quantity, and delivery address. Price change data includes adjusted prices, effective and ineffective times, and adjustment tags for multiple price adjustments. Adjustment tags include platform subsidies, store discounts, and streamer discounts. Live stream time stamps include peak promotional days, off-peak promotional days, peak non-peak promotional days, and others. Based on the sales product set of each live broadcast, real-time supply data, real-time warehousing data, and real-time after-sales data are collected from the monitoring ports of the live e-commerce supply chain, including the supplier side, warehousing side, and after-sales side. The real-time supply data includes the purchased quantity, purchase arrival time, transported quantity, transport arrival time, transferred quantity, and transferred arrival time of each product in the sales product set of each live broadcast. The real-time warehousing data includes the inventory of each product in the sales product set of each live broadcast. The real-time after-sales data includes the return rate, exchange rate, and negative review and complaint data of each product in the sales product set of each live broadcast. The product complaint data includes multiple negative reviews and multiple complaint records. Real-time logistics data is collected from the logistics side of the live-streaming e-commerce supply chain monitoring port set. The real-time logistics data includes the estimated transportation time, remaining transportation volume, and remaining transportation weight for each transportation trunk line. Based on real-time live sales data, real-time supply data, real-time warehousing data, real-time after-sales data, and real-time logistics data of the live-streaming e-commerce supply chain, real-time monitoring data of the live-streaming e-commerce supply chain is determined.

[0008] Preferably, a method for monitoring supply chain information in live-streaming e-commerce involves acquiring historical live-streaming data for each live stream from the real-time live-streaming sales data within the real-time monitoring data, based on real-time monitoring data of the live-streaming e-commerce supply chain. This includes: Based on the real-time monitoring data of the live-streaming e-commerce supply chain, the live-streaming number of each live-streaming session is obtained from the real-time live-streaming sales data. The historical live-streaming data includes sub-data of multiple historical live-streaming sessions. The sub-data of historical live-streaming sessions includes historical product sets, historical start time, historical end time, historical live-streaming tags, and historical sales product data for each product sold in the historical product set. The historical sales product data includes historical price trend charts and the adjusted price, effective time, and ineffective time of each price adjustment in the historical price trend charts, actual sales volume, historical order data, and adjustment tags.

[0009] Preferably, a method for monitoring supply chain information in live-streaming e-commerce involves determining the historical live-streaming data of each product in the product set of each live stream within the real-time live-streaming sales data of the real-time monitoring data, including: Based on the real-time monitoring data of the live-streaming e-commerce supply chain, for each product in the set of products sold in each live stream, the historical live stream data corresponding to each historical product set containing the products sold in the historical live stream data is extracted to determine the historical product live stream data of each product in the set of products sold in each live stream in the real-time monitoring data. The historical product live stream data includes historical live stream sub-data from multiple historical live streams.

[0010] Preferably, a method for monitoring supply chain information in live-streaming e-commerce, based on historical live-streaming data of each product in the product set of each live stream in real-time monitoring data and real-time live-streaming sales data, determines product prediction reference data and product pre-sale data for each product in the product set of each live stream in real-time monitoring data and real-time live-streaming sales data, including: Based on the price change data of each product in the product set of each live broadcast in the real-time live broadcast sales data of the live broadcast e-commerce supply chain, a product price trend chart of each product in the product set of each live broadcast in the real-time monitoring data of the live broadcast sales data is drawn. Based on the product price trend chart and basic live streaming data of each product in the sales product set of each live broadcast in the real-time monitoring data, as well as the historical live broadcast tags and historical price trend charts of all historical live broadcast sub-data of each product in the historical product live broadcast data, the product prediction reference data of each product in the sales product set of each live broadcast in the real-time monitoring data is calculated. The product prediction reference data includes the historical sales product data of the sales products corresponding to the historical product sets of multiple historical live broadcast sub-data. The product pricing and order generation data of each product in the sales product set of each live broadcast in the real-time monitoring data and the historical sales product data of the corresponding products in the historical product set of all historical live broadcast sub-data in the product prediction reference data are input into the live broadcast prediction model to determine the product pre-sale data of each product in the sales product set of each live broadcast in the real-time live broadcast sales data. The product pre-sale data includes the predicted selling price, the predicted sales duration, and the predicted sales volume.

[0011] Preferably, a method for monitoring supply chain information in live-streaming e-commerce determines real-time management data based on real-time monitoring data and the pre-sale data of each product in the product set of each live stream, including: Feature extraction is performed on each negative review and each complaint record in the product complaint data of each product in the real-time after-sales data set of the real-time monitoring data to determine the negative review feature vector and complaint feature vector of each negative review and each complaint record in the product complaint data of each product in the real-time after-sales data set of the real-time monitoring data. The return rate and exchange rate of each product in the sales product set of each live broadcast in the real-time monitoring data, the negative review feature vector of each complaint record and the complaint feature vector in the product complaint data are input into the product after-sales evaluation model to determine the after-sales evaluation value of each product in the sales product set of each live broadcast in the real-time monitoring data. Based on the order generation data of each product in the product set of each live broadcast in the real-time live sales data, the predicted sales volume in the product pre-sale data, the real-time supply data in the real-time monitoring data, and the real-time warehousing data, calculate the out-of-stock risk value of each product in the product set of each live broadcast in the real-time live sales data. Based on the product volume and weight of each product in the product set of each live broadcast in the real-time live sales data, the transaction quantity of all order numbers in the order generation data, and the predicted sales volume, determine the volume transportation capacity requirement and weight transportation capacity requirement of each product in the product set of each live broadcast in the real-time live sales data. The volume and weight capacity requirements of all products sold in the live sales data, as well as the order generation data, and the remaining volume and weight of all transportation trunk lines in the real-time supply data of the real-time monitoring data are input into the capacity risk model to determine the capacity risk value of all products sold in the live sales data. The platform's management model is input into the real-time live sales data, which includes the capacity risk value of all products in the product set of all live streams, the after-sales evaluation value of each product in the product set of each live stream, and the stockout risk value. This allows the platform to determine the decision suggestion label for each product in the product set of each live stream in the real-time monitoring data. Based on the decision suggestion tags of all sales products in the set of all live-streamed sales products in the real-time monitoring data, the real-time management data is determined.

[0012] This invention provides a live-streaming e-commerce supply chain information monitoring platform for executing any one of the live-streaming e-commerce supply chain information monitoring methods in embodiments 1 to 7, comprising: Acquisition Module: Acquires the set of monitoring ports for the live-streaming e-commerce supply chain, and collects real-time monitoring data of the live-streaming e-commerce supply chain based on the set of monitoring ports. The real-time monitoring data includes real-time live-streaming sales data, real-time supply data, real-time warehousing data, real-time after-sales data, and real-time logistics data. Determine the module: Based on real-time monitoring data of the live-streaming e-commerce supply chain, obtain historical live-streaming data for each live stream in the real-time live-streaming sales data in the real-time monitoring data, and determine the historical live-streaming data of each product in the product set of each live stream in the real-time live-streaming sales data in the real-time monitoring data; Prediction module: Based on the historical live streaming data of each product in the sales product set of each live stream in the real-time monitoring data and real-time live streaming sales data, determine the product prediction reference data and product pre-sale data of each product in the sales product set of each live stream in the real-time monitoring data and real-time live streaming sales data. Management module: Based on real-time monitoring data and real-time live sales data, the pre-sale data of each product in the product set of each live broadcast determines the real-time management data.

[0013] The beneficial effects of this invention compared to existing technologies are as follows: By acquiring a set of monitoring ports for the live-streaming e-commerce supply chain, real-time monitoring data, and historical live-streaming data for each live stream within the real-time live-streaming sales data, this invention determines the historical live-streaming data, product prediction reference data, and product pre-sale data for each product in the product set of each live stream within the real-time live-streaming sales data, and also determines real-time management data. This enables collaborative collection and integration of multi-port data across the entire supply chain, establishes a precise correlation system between real-time and historical data, improves the scientific rigor and foresight of supply chain demand forecasting, generates targeted product pre-sale data and real-time management data, effectively reduces the risks of inventory backlog, stockouts, and logistics mismatches, enhances the overall operational refinement level, and strengthens the market competitiveness and risk resistance of the live-streaming e-commerce supply chain.

[0014] Other features and advantages of the invention will be set forth in the description which follows, and will be apparent in part from the description, or may be learned by practicing the invention. The objects and other advantages of the invention may be realized and obtained by means of the structures particularly pointed out in this application.

[0015] The technical solution of the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. Attached Figure Description

[0016] The accompanying drawings are provided to further illustrate the invention and form part of the specification. They are used in conjunction with embodiments of the invention to explain the invention and do not constitute a limitation thereof. In the drawings: Figure 1 This is a flowchart of a live-streaming e-commerce supply chain information monitoring method according to an embodiment of the present invention; Figure 2 This is a schematic diagram of a live-streaming e-commerce supply chain information monitoring platform according to an embodiment of the present invention. Detailed Implementation

[0017] The preferred embodiments of the present invention will be described below with reference to the accompanying drawings. It should be understood that the preferred embodiments described herein are for illustration and explanation only and are not intended to limit the present invention. Example 1:

[0018] This invention provides a method for monitoring supply chain information in live-streaming e-commerce, with reference to... Figure 1 ,include: S1: Obtain the set of monitoring ports for the live e-commerce supply chain, and collect real-time monitoring data of the live e-commerce supply chain based on the set of monitoring ports. The real-time monitoring data includes real-time live sales data, real-time supply data, real-time warehousing data, real-time after-sales data, and real-time logistics data. S2: Based on real-time monitoring data of the live-streaming e-commerce supply chain, obtain historical live-streaming data for each live stream in the real-time live-streaming sales data in the real-time monitoring data, and determine the historical live-streaming data of each product in the product set of each live stream in the real-time live-streaming sales data in the real-time monitoring data. S3: Based on the historical live streaming data of each product in the sales product set of each live stream in the real-time monitoring data and real-time live streaming sales data, determine the product prediction reference data and product pre-sale data of each product in the sales product set of each live stream in the real-time monitoring data and real-time live streaming sales data. S4: Based on real-time monitoring data and real-time live sales data, determine the pre-sale data of each product in the product set of each live broadcast.

[0019] In this embodiment, it is clearly stated that a set of monitoring ports for the live-streaming e-commerce supply chain needs to be obtained. This set of monitoring ports is a full-link data collection entry point covering the live-streaming sales end, supply end, warehousing end, after-sales end, and logistics end. It can realize data access and real-time collection of each key link in the supply chain. The real-time collection work carried out based on this determined set of monitoring ports ultimately yields real-time monitoring data containing five core categories: real-time live-streaming sales data, real-time supply data, real-time warehousing data, real-time after-sales data, and real-time logistics data. These five types of data comprehensively cover the key information of the entire process of the live-streaming e-commerce supply chain from product supply to after-sales feedback.

[0020] In this embodiment, the real-time monitoring data of the live-streaming e-commerce supply chain is collected as the core foundation. The focus is on the real-time live-streaming sales data. The historical live-streaming data corresponding to each live-streaming session is accurately extracted from it. The historical live-streaming data is the full-dimensional operational data of the same live-streaming number before the current live-streaming session. By filtering and matching these historical live-streaming data, the corresponding historical product live-streaming data is determined one by one for each product in the product set of each live-streaming session. These historical product live-streaming data are the sales performance and supply chain response data of each product in various live-streaming sessions in the past.

[0021] In this embodiment, the acquired real-time monitoring data is fully integrated and deeply analyzed with the historical live-stream data of each sales product. By summarizing the trend prediction through the collaborative analysis of the two types of data, the product prediction reference data and product pre-sale data of each sales product in the live-stream sales data of each live broadcast are finally determined. The product prediction reference data is the benchmark data used to predict the subsequent supply chain response trend of the product in this live broadcast. The product pre-sale data is the pre-sale related data formed by prediction based on real-time sales and historical data.

[0022] In this embodiment, the collected real-time monitoring data is systematically integrated and comprehensively evaluated with the pre-sale data of each product. Through comprehensive analysis and scientific judgment of the two types of data, real-time management data is finally determined to guide the real-time operation and management of the live-streaming e-commerce supply chain. This real-time management data can directly affect each key link of the supply chain, providing clear decision-making basis for inventory adjustment, logistics scheduling, supplier preparation, and after-sales resource allocation, ensuring that the operation rhythm of the supply chain is highly adapted to the real-time demand of live-streaming sales.

[0023] The beneficial effects of the above technologies are as follows: By acquiring the monitoring port set of the live-streaming e-commerce supply chain, real-time monitoring data, and historical live-streaming data for each live stream in the real-time live-streaming sales data, the historical live-streaming data, product prediction reference data, and product pre-sale data for each product in the sales product set of each live stream in the real-time live-streaming sales data are determined, and real-time management data is also determined. This enables collaborative collection and integration of multi-port data across the entire supply chain, establishing a precise correlation system between real-time and historical data, improving the scientific and forward-looking nature of supply chain demand forecasting, generating targeted product pre-sale data and real-time management data, effectively reducing the risks of inventory backlog, stockouts, and logistics mismatches, improving the overall operational refinement level, and enhancing the market competitiveness and risk resistance of the live-streaming e-commerce supply chain. Example 2:

[0024] Based on Example 1, a method for monitoring supply chain information in live-streaming e-commerce is provided, wherein the monitoring port set includes the supplier end, warehousing end, logistics end, live-streaming sales end, and after-sales service end.

[0025] In this embodiment, the supplier side refers to the data interface of the main source of goods, such as brand owners, factories or distributors; the warehousing side refers to the system port of central warehouses, cloud warehouses or forward warehouses; the logistics side refers to the carrier interface of trunk lines, branch lines and last-mile delivery; the live sales side refers to the data outlet of the live room itself; and the after-sales service side refers to the aggregation port of the platform customer service, return center and evaluation system.

[0026] The beneficial effects of the above technologies are: obtaining a set of monitoring ports for the live-streaming e-commerce supply chain can provide data support for collecting real-time monitoring data. Example 3:

[0027] Based on Example 2, a method for monitoring supply chain information in live-streaming e-commerce, which collects real-time monitoring data of the live-streaming e-commerce supply chain based on a set of monitoring ports, includes: The system collects real-time live sales data from the live e-commerce supply chain monitoring portals. This data includes a collection of products sold across multiple live streams, the product volume, weight, pricing, price changes, and order generation data for each product within that collection, as well as basic live stream data. Basic live stream data includes the live stream number, start time, estimated end time, and time stamp. Order generation data includes the order number, transaction price, quantity, and delivery address. Price change data includes adjusted prices, effective and ineffective times, and adjustment tags for multiple price adjustments. Adjustment tags include platform subsidies, store discounts, and streamer discounts. Live stream time stamps include peak promotional days, off-peak promotional days, peak non-peak promotional days, and others. Based on the sales product set of each live broadcast, real-time supply data, real-time warehousing data, and real-time after-sales data are collected from the monitoring ports of the live e-commerce supply chain, including the supplier side, warehousing side, and after-sales side. The real-time supply data includes the purchased quantity, purchase arrival time, transported quantity, transport arrival time, transferred quantity, and transferred arrival time of each product in the sales product set of each live broadcast. The real-time warehousing data includes the inventory of each product in the sales product set of each live broadcast. The real-time after-sales data includes the return rate, exchange rate, and negative review and complaint data of each product in the sales product set of each live broadcast. The product complaint data includes multiple negative reviews and multiple complaint records. Real-time logistics data is collected from the logistics side of the live-streaming e-commerce supply chain monitoring port set. The real-time logistics data includes the estimated transportation time, remaining transportation volume, and remaining transportation weight for each transportation trunk line. Based on real-time live sales data, real-time supply data, real-time warehousing data, real-time after-sales data, and real-time logistics data of the live-streaming e-commerce supply chain, real-time monitoring data of the live-streaming e-commerce supply chain is determined.

[0028] In this embodiment, real-time data collection is carried out on the live sales terminals in the monitoring port set. The collected real-time live sales data includes multiple aspects: first, the set of products sold for each of the multiple live streams; second, the product volume and weight of each product in the set of products; third, the product price of each product; fourth, the price change data of each product; fifth, the order generation data generated by each live stream; and sixth, the basic data of each live stream. Among them, the price change data includes the adjusted price, adjustment effective time, adjustment expiration time, and adjustment tag for each product after multiple price adjustments. The adjustment tag is specifically divided into three categories: platform subsidy, store discount, and streamer discount. The order generation data includes the order number, transaction price, transaction quantity, and order delivery address for each order. The basic data of the live stream includes the live stream number, live stream start time, estimated live stream end time, and live stream time tag for each live stream. The live stream time tag is specifically divided into four categories: peak period on promotional days, off-peak period on promotional days, peak period on non-promotional days, and others.

[0029] In this embodiment, based on the set of sales products corresponding to each live broadcast in the real-time live sales data, real-time data collection is carried out on the supplier end, warehouse end and after-sales end in the monitoring port set. The collected real-time supply data includes the purchased quantity, purchase arrival time, transported quantity, transport arrival time, transferred quantity and transferred arrival time for each sales product in the set of sales products for each live broadcast. The collected real-time warehouse data includes the inventory quantity for each sales product in the set of sales products for each live broadcast. The collected real-time after-sales data includes the return rate, exchange rate and negative review complaint data for each sales product in the set of sales products for each live broadcast. The negative review complaint data includes multiple negative reviews and multiple complaint records.

[0030] In this embodiment, real-time data collection is carried out on the logistics end of the monitoring port set. The collected real-time logistics data includes the estimated transportation time, remaining transportation volume, and remaining transportation weight for each transportation trunk line.

[0031] In this embodiment, the collected real-time live sales data, real-time supply data, real-time warehousing data, real-time after-sales data, and real-time logistics data are comprehensively integrated and summarized to ultimately determine the real-time monitoring data of the live e-commerce supply chain.

[0032] The beneficial effects of the above technologies are as follows: real-time monitoring data of the live-streaming e-commerce supply chain can be collected based on the monitoring port set, enabling accurate collection of full-link data of live-streaming e-commerce and improving the integrity and relevance of supply chain data. Example 4:

[0033] Based on Example 3, a method for monitoring supply chain information in live-streaming e-commerce, based on real-time monitoring data of the live-streaming e-commerce supply chain, acquires historical live-streaming data for each live stream from the real-time live-streaming sales data within the real-time monitoring data, including: Based on the real-time monitoring data of the live-streaming e-commerce supply chain, the live-streaming number of each live-streaming session is obtained from the real-time live-streaming sales data. The historical live-streaming data includes sub-data of multiple historical live-streaming sessions. The sub-data of historical live-streaming sessions includes historical product sets, historical start time, historical end time, historical live-streaming tags, and historical sales product data for each product sold in the historical product set. The historical sales product data includes historical price trend charts and the adjusted price, effective time, and ineffective time of each price adjustment in the historical price trend charts, actual sales volume, historical order data, and adjustment tags.

[0034] In this embodiment, the live stream number corresponding to each live stream in the real-time live sales data of the live e-commerce supply chain real-time monitoring data is used to accurately retrieve all historical live stream data for each live stream. The live stream number serves as a unique identifier in this process. The retrieved historical live stream data includes sub-data from multiple historical live streams. Each sub-data item covers several key data dimensions. First, there's the historical product set, which summarizes all products actually sold in the corresponding historical live stream session. Second, there's the historical start and end times, recording the start and end times of the corresponding live stream. Then, there are historical live stream tags, which categorize and label the attributes of the historical live stream sessions. In addition, the historical live stream sub-data also includes historical sales product data for each product sold in the historical product set. This historical sales product data provides detailed sales records for individual products, including not only historical price trend charts, which visually reflect the overall price change trajectory of the corresponding product during the historical live streams, but also a series of specific data corresponding to each price adjustment in the historical price trend chart. Specifically, this includes the adjusted price (the actual selling price of the product during the corresponding price adjustment phase), the effective time of the price adjustment (the specific time when the price adjustment began), the adjustment expiration time (the specific time when the price adjustment ended), the actual sales volume (the actual number of products sold during the corresponding price adjustment phase), historical order data (all order-related information generated during the corresponding price adjustment phase), and adjustment tags (the annotations for the price adjustment).

[0035] The beneficial effects of the above technologies are as follows: obtaining historical live streaming data for each live stream from real-time live streaming sales data in real-time monitoring data can provide accurate reference for determining historical product live streaming data, enhance the value of data reuse, and strengthen the scientific and forward-looking nature of live streaming e-commerce supply chain operation analysis. Example 5:

[0036] Based on Example 4, a method for monitoring supply chain information in live-streaming e-commerce determines the historical product live-streaming data for each product in the product set of each live stream within the real-time live-streaming sales data of the real-time monitoring data, including: Based on the real-time monitoring data of the live-streaming e-commerce supply chain, for each product in the set of products sold in each live stream, the historical live stream data corresponding to each historical product set containing the products sold in the historical live stream data is extracted to determine the historical product live stream data of each product in the set of products sold in each live stream in the real-time monitoring data. The historical product live stream data includes historical live stream sub-data from multiple historical live streams.

[0037] In this embodiment, data extraction is performed on each product within the product set of each live stream in the real-time live sales data from the real-time monitoring data. First, the core anchor point for data extraction is clearly defined as the individual product in each live stream. This anchor point is the sole basis for all subsequent data extraction operations, ensuring that the extraction work always focuses on the single product dimension rather than the general session dimension. Next, a comprehensive screening of the historical live stream data for the corresponding live stream is conducted. The core screening condition is to determine whether the historical product set included in the historical live stream data covers the specific product being extracted. Only historical live stream data containing the specific product in the historical product set is included in the extraction scope. This strict screening condition effectively filters out redundant historical data that does not contain the target product, preventing invalid data from interfering with subsequent analysis. Then, all historical live stream data that meets the screening conditions is comprehensively extracted and summarized. This extracted historical live stream data is valid data directly related to the target product and can accurately reflect the product's performance in past live stream sessions. Finally, these filtered and summarized historical live streaming data were identified as the historical product live streaming data for each live streaming product set in the real-time live streaming sales data of the real-time monitoring data. The specific composition of this historical product live streaming data consists of historical live streaming sub-data corresponding to multiple historical live streams. These historical live streaming sub-data can fully present the relevant information of the target product in different historical live streaming sessions.

[0038] The beneficial effects of the above technologies are as follows: by determining the historical product live streaming data of each product in the set of products sold in each live stream of the real-time monitoring data, accurate historical data matching with individual products can be achieved, thereby improving the compatibility between historical data and individual products. Example 6:

[0039] Based on Example 5, a live-streaming e-commerce supply chain information monitoring method, based on historical live-streaming data of each product in the sales product set of each live stream in real-time monitoring data and real-time live-streaming sales data, determines product prediction reference data and product pre-sale data for each product in the sales product set of each live stream in real-time monitoring data and real-time live-streaming sales data, including: Based on the price change data of each product in the product set of each live broadcast in the real-time live broadcast sales data of the live broadcast e-commerce supply chain, a product price trend chart of each product in the product set of each live broadcast in the real-time monitoring data of the live broadcast sales data is drawn. Based on the product price trend chart and basic live streaming data of each product in the sales product set of each live broadcast in the real-time monitoring data, as well as the historical live broadcast tags and historical price trend charts of all historical live broadcast sub-data of each product in the historical product live broadcast data, the product prediction reference data of each product in the sales product set of each live broadcast in the real-time monitoring data is calculated. The product prediction reference data includes the historical sales product data of the sales products corresponding to the historical product sets of multiple historical live broadcast sub-data. The product pricing and order generation data of each product in the sales product set of each live broadcast in the real-time monitoring data and the historical sales product data of the corresponding products in the historical product set of all historical live broadcast sub-data in the product prediction reference data are input into the live broadcast prediction model to determine the product pre-sale data of each product in the sales product set of each live broadcast in the real-time live broadcast sales data. The product pre-sale data includes the predicted selling price, the predicted sales duration, and the predicted sales volume.

[0040] In this embodiment, each product in the live-stream sales data set of each live-stream session is taken as the core. Price change data corresponding to each product is extracted, and a product price trend chart is drawn based on this price change data. The price change data includes multiple price adjustments for the product during the live stream, covering the adjusted price, the effective time of the adjustment, and the expiration time of the adjustment. By sorting, integrating, and visualizing this dynamic data, the resulting product price trend chart can clearly reflect the price fluctuations of the product at different time points during the live stream, intuitively showing the frequency, magnitude, and corresponding time range of price adjustments, representing the price change trajectory during the current live stream.

[0041] In this embodiment, the horizontal axis of the product price trend chart represents the live stream start time to the expected end time in the basic live stream data, and the vertical axis represents the adjusted price in the price change data of the products sold in the product set of the corresponding live stream.

[0042] In this embodiment, based on the product price trend chart and basic live streaming data of each product in the sales product set of each live broadcast in the real-time monitoring data, and the historical live broadcast tags and historical price trend charts of all historical live broadcasts in the historical product live broadcast data of each product, the product prediction reference data of each product in the sales product set of each live broadcast in the real-time monitoring data is calculated. The calculation formula is expressed as follows: ; in, This represents the product prediction reference data for the j-th product in the product set of the i-th live stream in the real-time live sales data. This represents the live broadcast time tag in the basic data of the i-th live broadcast in the real-time live broadcast sales data. This represents the historical live stream tag within the historical live stream sub-data of the k-th historical live stream in the historical product live stream data of the j-th product in the i-th live stream sales data. This represents the live broadcast time tag in the basic live broadcast data of the i-th live broadcast in real-time live broadcast sales data, and is a first indicator function based on the historical live broadcast tag in the historical live broadcast sub-data of the k-th historical live broadcast in the historical product live broadcast data of the j-th sales product set in the i-th live broadcast. This represents the j-th product sold in the set of products sold in the i-th live stream of real-time sales data. Let $j_j$ represent the number of products sold in the historical product set of the historical product sub-data of the historical product set ... i-th live broadcast in the real-time live broadcast sales data, where 1 ≤ m ≤ $ij_kN^2$. $ij_kN^2$ represents the number of products sold in the historical product set of the historical product set of the historical product set of the i-th live broadcast in the real-time live broadcast sales data, and $ij_kN^2$ represents the number of products sold in the historical product set of the historical product set of the historical product set of the historical product set of the historical product set of the historical product set of the k-th live broadcast in the real-time live broadcast sales data. This represents the trend curve in the product price trend graph of the j-th product in the product set of the i-th live stream. This represents the trend curve in the historical price trend chart of the historical product data for the j-th product in the i-th live stream sales data, the historical product data for the m-th product in the historical product set of the k-th historical live stream sub-data, and the historical product data for the i-th live stream in the real-time live stream sales data. This represents the distance between the trend curve in the price trend chart of the j-th product in the product set of the i-th live stream and the trend curve in the historical price trend chart of the m-th product in the historical product set of the k-th historical live stream. This represents the curve distance sequence of the j-th product in the product set of the i-th live stream in real-time live sales data. Let q represent the q-quantile of the curve distance sequence of the j-th product in the product set of the i-th live stream in real-time live sales data. Let Nij represent the historical product data of the j-th product in the historical product data of the k-th historical live stream sub-data of the i-th live stream in the real-time live stream sales data, and let Nij represent the number of historical live stream sub-data of the j-th product in the historical product data of the i-th live stream in the real-time live stream sales data. This represents the start time and estimated end time of the i-th live stream in the real-time live sales data. This represents the historical start time and historical end time of the historical live stream sub-data of the k-th historical live stream in the historical product live stream data of the j-th product in the i-th live stream sales data set. tc represents the current time. This represents the historical mapping time of the trend curve in the historical price trend chart of the historical product data of the j-th product in the i-th live broadcast sales data, the historical product data of the k-th historical live broadcast sub-data, and the historical product data of the m-th product in the historical product data of the i-th live broadcast sales data. Let represent the weight of the historical live stream sub-data of the k-th historical live stream in the historical product live stream data of the j-th product in the i-th live stream sales data set.

[0043] In this embodiment, the trend curve of the product price trend chart of the j-th product in the product set of the i-th live broadcast in the real-time live broadcast sales data is used. The trend curve in the historical price trend chart of the historical product data of the j-th product in the i-th live broadcast sales data, the historical product data of the m-th product in the historical product set of the k-th historical live broadcast sub-data, and the historical product data of the i-th live broadcast. Perform path mapping to determine the optimal alignment path between the two trend curves, and select the trend curve from the product price trend chart of the j-th product in the product set of the i-th live broadcast in the real-time live sales data. Find the last adjusted price in the trend curve, and then find the corresponding historical adjusted price in the optimal alignment path. The most recent adjusted price in the trend curve is the effective time of the adjustment. The historical effective time of the historical adjusted price aligned in the optimal alignment path is the historical mapping time. If the current time is far from the trend curve... The most recent adjustment price in the data is the adjustment failure time, and the historical failure time of the historical adjustment price aligned in the optimal alignment path is the historical mapping time.

[0044] In this embodiment, This represents the product of the curve distance value and weight of the trend curve of the j-th product in the product price trend chart of the i-th live broadcast's sales product set in the real-time live broadcast sales data, and the trend curve of the historical price trend chart of the m-th product in the historical sales product data of all historical live broadcast sub-data, sorted from smallest to largest.

[0045] In this embodiment, the q quantile can be 30%.

[0046] In this embodiment, the pricing and order generation data of each product in each live-stream sales set are collected from the real-time monitoring data. Simultaneously, historical sales data of the corresponding products are extracted from the historical product sets of all historical live-stream sub-data in the product prediction reference data. These two types of real-time data, along with one type of historical data, are input into the live-stream prediction model. Product pricing serves as the basic price benchmark for individual products in the current live stream. Order generation data reflects the real-time transaction status of the current live stream, and historical sales data provides past sales performance and price matching patterns for individual products. Through comprehensive analysis and processing of these multi-source data by the model, the pre-sale data for each product is ultimately determined. This data specifically includes the predicted selling price, predicted sales duration, and predicted sales volume, providing clear guidance for the supply chain to plan inventory preparation and adjust operational strategies in advance.

[0047] In this embodiment, the live streaming prediction model conducts multi-dimensional collaborative correlation analysis based on input product pricing, order generation data, and historical sales product data. First, the three types of heterogeneous data are standardized to eliminate format and magnitude differences between data sources, ensuring data comparability and analyzability. Next, product pricing is matched with historical price trends in historical sales product data to filter out historical live streaming cases similar to the current pricing strategy. Simultaneously, core features such as real-time transaction growth rate and order distribution density are extracted from order generation data and correlated with corresponding sales trend changes in historical cases. The accuracy of feature matching is enhanced by combining the scene attributes corresponding to historical live streaming tags. Finally, based on the mined feature correlations... The system uses a pattern-based approach to extrapolate trends. For the predicted selling price, it references the price adjustment effects under similar historical pricing strategies and the real-time order feedback from the current live stream to extrapolate the optimal price range that can achieve a balance between sales volume and profit. For the predicted sales duration, it combines the complete sales cycle of the same product in historical cases and the order growth rate of the current live stream to determine the duration of the product's sales popularity in this live stream. For the predicted sales volume, it integrates the single-product sales volume data from similar historical scenarios, the real-time order conversion efficiency of the current live stream, and the predicted sales duration to extrapolate the total sales volume of the product within the predicted period. Finally, the extrapolated predicted selling price, predicted sales duration, and predicted sales volume are integrated and output to form complete product pre-sale data.

[0048] The beneficial effects of the above technologies are as follows: Based on the historical live streaming data of each product in the product set of each live stream in the real-time monitoring data and real-time live streaming sales data, the product prediction reference data and product pre-sale data of each product in the product set of each live stream in the real-time monitoring data and real-time live streaming sales data can be determined. This can achieve accurate generation of prediction reference data, improve the pertinence and accuracy of pre-sale prediction, strengthen data visualization and deep reuse of historical data, provide a scientific basis for precise inventory preparation and dynamic price adjustment in the supply chain, and improve the predictive ability of live streaming e-commerce supply chain operation. Example 7:

[0049] Based on Example 6, a live-streaming e-commerce supply chain information monitoring method determines real-time management data based on real-time monitoring data and the pre-sale data of each product in the product set of each live stream, including: Feature extraction is performed on each negative review and each complaint record in the product complaint data of each product in the real-time after-sales data set of the real-time monitoring data to determine the negative review feature vector and complaint feature vector of each negative review and each complaint record in the product complaint data of each product in the real-time after-sales data set of the real-time monitoring data. The return rate and exchange rate of each product in the sales product set of each live broadcast in the real-time monitoring data, the negative review feature vector of each complaint record and the complaint feature vector in the product complaint data are input into the product after-sales evaluation model to determine the after-sales evaluation value of each product in the sales product set of each live broadcast in the real-time monitoring data. Based on the order generation data of each product in the product set of each live broadcast in the real-time live sales data, the predicted sales volume in the product pre-sale data, the real-time supply data in the real-time monitoring data, and the real-time warehousing data, calculate the out-of-stock risk value of each product in the product set of each live broadcast in the real-time live sales data. Based on the product volume and weight of each product in the product set of each live broadcast in the real-time live sales data, the transaction quantity of all order numbers in the order generation data, and the predicted sales volume, determine the volume transportation capacity requirement and weight transportation capacity requirement of each product in the product set of each live broadcast in the real-time live sales data. The volume and weight capacity requirements of all products sold in the live sales data, as well as the order generation data, and the remaining volume and weight of all transportation trunk lines in the real-time supply data of the real-time monitoring data are input into the capacity risk model to determine the capacity risk value of all products sold in the live sales data. The platform's management model is input into the real-time live sales data, which includes the capacity risk value of all products in the product set of all live streams, the after-sales evaluation value of each product in the product set of each live stream, and the stockout risk value. This allows the platform to determine the decision suggestion label for each product in the product set of each live stream in the real-time monitoring data. Based on the decision suggestion tags of all sales products in the set of all live-streamed sales products in the real-time monitoring data, the real-time management data is determined.

[0050] In this embodiment, for each product in the real-time after-sales data set of the real-time monitoring data, feature extraction is performed on each negative review and each complaint record in its product complaint data. The feature extraction covers all the core content of the negative reviews and complaint records, including the description of the negative reviews, the key issues in the complaint records, the product links involved in the complaints, and other key information. Through the system's comprehensive feature extraction operation, the negative review feature vector corresponding to each negative review and the complaint feature vector corresponding to each complaint record in the product complaint data of each product in the real-time after-sales data set of the real-time monitoring data are finally determined. These feature vectors transform the originally scattered text-based negative reviews and complaint information into standardized, quantifiable, and analyzable data.

[0051] In this embodiment, the return rate and exchange rate of each product in the sales product set of each live broadcast in the real-time monitoring data, as well as the negative review feature vector and complaint feature vector corresponding to each negative review and complaint record in the product complaint data of that product, are all input into the product after-sales evaluation model. The return rate and exchange rate can intuitively reflect the overall return and exchange situation of the product, while the negative review feature vector and complaint feature vector can accurately present the specific after-sales problems of the product. The product after-sales evaluation model performs comprehensive analysis and quantitative evaluation based on these input multi-dimensional data, and finally determines the after-sales evaluation value corresponding to each product in the sales product set of each live broadcast in the real-time monitoring data. This after-sales evaluation value is a comprehensive quantitative indicator of the product's after-sales performance, which can clearly reflect the product's after-sales quality and user satisfaction.

[0052] In this embodiment, the product after-sales evaluation model performs multi-dimensional quantitative integration and weighted analysis on the input return rate, exchange rate, negative review feature vector, and complaint feature vector. First, basic weights are assigned to the return rate and exchange rate, which directly reflect the severity of after-sales problems. Then, the negative review feature vector and complaint feature vector are analyzed to extract core dimensions such as product quality issues, service response issues, and logistics and delivery issues. These dimensions are precisely matched with the preset after-sales problem weight system. Subsequently, a comprehensive weighted calculation is performed by combining the basic weights of the return rate and exchange rate with the matching weights of the negative review and complaint feature dimensions. Finally, an after-sales evaluation value that can intuitively reflect the quality of after-sales performance of each product is obtained. The level of the evaluation value is positively correlated with the severity of the product's after-sales problems.

[0053] In this embodiment, based on the order generation data of each product in the product set of each live broadcast in the real-time live sales data, the predicted sales volume in the product pre-sale data, the real-time supply data, real-time logistics data, and real-time warehousing data in the real-time monitoring data, the out-of-stock risk value of each product in the product set of each live broadcast in the real-time live sales data is calculated. The calculation formula is expressed as follows: ; in, This represents the stockout risk value of the j-th product in the product set of the i-th live stream in the real-time logistics data. This represents the predicted sales volume of the j-th product in the set of products sold during the i-th live stream of real-time logistics data, as monitored in real-time data. Let ijN2 represent the number of transactions for the j-th product in the sales product set of the i-th live stream in the real-time logistics data, with order number a. Let these represent the purchased quantity, transported quantity, and transferred quantity of the j-th product in the sales product set of the i-th live broadcast in the real-time logistics data, respectively. Let $\mathbf$ and $\mathbf$ respectively represent the purchase arrival time, transportation arrival time, and restocking arrival time of the $j$-th product in the $i$-th live stream's product set, as part of the real-time logistics data. This represents the estimated end time of the live stream in the basic data of the i-th live stream. This indicates the next shipping time from the nearest order cutoff point in the basic data of live streams that are later than the i-th live stream, based on the expected end time of the live stream. These represent the warehouse baseline times for the purchase arrival time, transportation arrival time, and transfer arrival time of the j-th product in the sales product set of the i-th live broadcast in the real-time logistics data, respectively. This represents the warehouse time consumption adjustment factor in the live broadcast time tag of the i-th live broadcast in the real-time live broadcast sales data. Let represent the purchase arrival quantity indicator function, transportation arrival quantity indicator function, and transfer arrival quantity indicator function, respectively, for the j-th sales product in the sales product set of the i-th live broadcast in the real-time logistics data. This represents the inventory of the j-th product in the set of products sold during the i-th live broadcast of real-time warehouse data, as monitored in real-time.

[0054] In this embodiment, When supply exceeds demand, the value is 0. When supply falls short of demand, the value is 1.

[0055] In this embodiment, data integration and analysis are carried out based on the product volume and weight of each product in the product set of each live broadcast in the real-time live sales data, the transaction quantity corresponding to all order numbers in the order generation data, and the predicted sales volume in the product pre-sale data. Product volume and weight are the basic parameters for calculating transportation capacity demand. The transaction quantity reflects the cargo volume corresponding to the real-time order, and the predicted sales volume provides the expected cargo volume that needs to be transported in the future. Through comprehensive statistics and system accounting of these data, the volume transportation capacity demand and weight transportation capacity demand corresponding to each product in the product set of each live broadcast in the real-time live sales data are finally determined. These two types of transportation capacity demand data can clearly present the specific needs of the product for transportation space and transportation load.

[0056] In this embodiment, the volumetric and weight capacity requirements of all products in the live-streamed sales product set from the real-time live-streamed sales data, along with the order generation data, and the remaining transportation volume and weight of all transportation trunk lines from the real-time supply data in the real-time monitoring data, are all input into the capacity risk model. The volumetric and weight capacity requirements reflect the total transportation demand of the goods, the order generation data provides the order delivery location, and the remaining transportation volume and weight reflect the current available capacity of the transportation trunk lines. Based on these input data, the capacity risk model performs supply and demand matching analysis and risk assessment, and finally determines the capacity risk value corresponding to all products in the live-streamed sales product set from the real-time live-streamed sales data. This capacity risk value can accurately predict whether the product will experience insufficient transportation capacity and be unable to be delivered in a timely manner.

[0057] In this embodiment, the capacity risk model is based on the input volume capacity demand, weight capacity demand, order generation data, and the remaining volume and weight of the transportation trunk lines to conduct a full-chain capacity supply and demand matching analysis. First, based on the order generation data, the total volume capacity demand and total weight capacity demand of each trunk line for all products in the live-streamed sales product set are calculated. Then, the total demand of each trunk line is compared with the remaining capacity data of each transportation trunk line to determine the rationality of the allocation of capacity resources in different regions and at different times. The transportation trunk lines with capacity gaps and the corresponding product categories are accurately identified. Then, the capacity risk value corresponding to each sales product is calculated and determined according to the size and scope of the supply and demand gap. The higher the risk value, the greater the risk of transportation and delivery delays faced by the product.

[0058] In this embodiment, the transportation risk value corresponding to all sales products in the sales product set of all live broadcasts in the real-time live sales data, the after-sales evaluation value and the stockout risk value corresponding to each sales product in the sales product set of each live broadcast are all input into the platform management model. The transportation risk value reflects the risk situation in the transportation link, the after-sales evaluation value reflects the quality situation in the after-sales link, and the stockout risk value reflects the risk situation in the inventory supply link. Based on these multi-dimensional risk and evaluation data, the platform management model conducts comprehensive judgment and overall analysis, and finally determines the decision suggestion label corresponding to each sales product in the sales product set of each live broadcast in the real-time monitoring data. These decision suggestion labels can provide clear guidance for the operational adjustments of each link in the supply chain.

[0059] In this embodiment, the decision suggestion labels include banning sales, freezing, limiting traffic, converting to pre-sale, price reduction, and price increase.

[0060] In this embodiment, the platform management model performs multi-dimensional risk assessment and classification on the input capacity risk value, after-sales evaluation value, and stockout risk value. First, based on the operational characteristics of the live-streaming e-commerce supply chain, it presets the influence weights of three types of indicators: stockout risk value relates to inventory supply stability, capacity risk value relates to delivery fulfillment efficiency, and after-sales evaluation value relates to user satisfaction. The weighting of these three indicators fully considers the impact of each link on the overall supply chain operation. Then, it compares the input values ​​with preset risk level thresholds, classifying them into low, medium, and high risk levels. Next, through collaborative analysis of the risk levels of the three types of indicators, it judges the comprehensive operational status of each sales product in terms of inventory supply, logistics, delivery, and after-sales service. Finally, it outputs decision-making suggestion labels that can guide the supply chain to accurately prepare inventory, optimize capacity allocation, and improve after-sales service. For example, when a product's after-sales evaluation value reaches an extremely high threshold, with negative reviews exceeding 50% and mostly related to quality and safety issues, a sales ban label is triggered; when a product's after-sales evaluation value is high, with negative reviews concentrated on product damage, functional malfunctions, or a medium-to-high stockout risk value, and inventory is only sufficient to support 2 When sales are conducted within an hour and the replenishment cycle exceeds 24 hours, and the capacity risk value is moderate, and some trunk line capacity is tight, a freeze label will be triggered.

[0061] In this embodiment, decision suggestion tags corresponding to all sales products in the set of all live-streamed sales products in real-time monitoring data are comprehensively integrated and systematically sorted out to finally determine real-time management data. This real-time management data can directly affect the daily operation and management of the live-streaming e-commerce supply chain, providing scientific and accurate decision-making basis for various aspects of the supply chain, such as inventory adjustment, transportation capacity scheduling, supplier collaboration, and after-sales optimization.

[0062] The beneficial effects of the above technologies are as follows: Based on real-time monitoring data and pre-sale data of each product in the product set of each live broadcast, real-time management data can be determined. This enables precise quantitative assessment of negative reviews and complaints, accurate calculation of stockout and shipping capacity risks, determination of targeted decision-making suggestion tags, and promotes the shift of supply chain risk management from passive response to proactive prediction. It also improves the efficiency and accuracy of risk identification and control across the entire chain, and strengthens the stability and resilience of the live-streaming e-commerce supply chain. Example 8:

[0063] This invention provides a live-streaming e-commerce supply chain information monitoring platform, used to execute any one of the live-streaming e-commerce supply chain information monitoring methods in embodiments 1 to 7, with reference to... Figure 2 ,include: Acquisition Module: Acquires the set of monitoring ports for the live-streaming e-commerce supply chain, and collects real-time monitoring data of the live-streaming e-commerce supply chain based on the set of monitoring ports. The real-time monitoring data includes real-time live-streaming sales data, real-time supply data, real-time warehousing data, real-time after-sales data, and real-time logistics data. Determine the module: Based on real-time monitoring data of the live-streaming e-commerce supply chain, obtain historical live-streaming data for each live stream in the real-time live-streaming sales data in the real-time monitoring data, and determine the historical live-streaming data of each product in the product set of each live stream in the real-time live-streaming sales data in the real-time monitoring data; Prediction module: Based on the historical live streaming data of each product in the sales product set of each live stream in the real-time monitoring data and real-time live streaming sales data, determine the product prediction reference data and product pre-sale data of each product in the sales product set of each live stream in the real-time monitoring data and real-time live streaming sales data. Management module: Based on real-time monitoring data and real-time live sales data, the pre-sale data of each product in the product set of each live broadcast determines the real-time management data.

[0064] The beneficial effects of the above technologies are as follows: By acquiring the monitoring port set of the live-streaming e-commerce supply chain, real-time monitoring data, and historical live-streaming data for each live stream in the real-time live-streaming sales data, the historical live-streaming data, product prediction reference data, and product pre-sale data for each product in the sales product set of each live stream in the real-time live-streaming sales data are determined, and real-time management data is also determined. This enables collaborative collection and integration of multi-port data across the entire supply chain, establishing a precise correlation system between real-time and historical data, improving the scientific and forward-looking nature of supply chain demand forecasting, generating targeted product pre-sale data and real-time management data, effectively reducing the risks of inventory backlog, stockouts, and logistics mismatches, improving the overall operational refinement level, and enhancing the market competitiveness and risk resistance of the live-streaming e-commerce supply chain.

[0065] Obviously, those skilled in the art can make various modifications and variations to this invention without departing from its spirit and scope. Therefore, if these modifications and variations fall within the scope of the claims of this invention and their equivalents, this invention also intends to include these modifications and variations.

Claims

1. A method for monitoring supply chain information in live-streaming e-commerce, characterized in that, include: S1: Obtain the set of monitoring ports for the live e-commerce supply chain, and collect real-time monitoring data of the live e-commerce supply chain based on the set of monitoring ports. The real-time monitoring data includes real-time live sales data, real-time supply data, real-time warehousing data, real-time after-sales data, and real-time logistics data. S2: Based on real-time monitoring data of the live-streaming e-commerce supply chain, obtain historical live-streaming data for each live stream in the real-time live-streaming sales data in the real-time monitoring data, and determine the historical live-streaming data of each product in the product set of each live stream in the real-time live-streaming sales data in the real-time monitoring data. S3: Based on the historical live streaming data of each product in the sales product set of each live stream in the real-time monitoring data and real-time live streaming sales data, determine the product prediction reference data and product pre-sale data of each product in the sales product set of each live stream in the real-time monitoring data and real-time live streaming sales data. S4: Based on real-time monitoring data and real-time live sales data, determine the pre-sale data of each product in the product set of each live broadcast.

2. The method for monitoring supply chain information in live-streaming e-commerce according to claim 1, characterized in that, The monitoring port set includes the supplier end, warehouse end, logistics end, live sales end, and after-sales end.

3. The method for monitoring supply chain information in live-streaming e-commerce according to claim 2, characterized in that, Real-time monitoring data of the live-streaming e-commerce supply chain is collected based on a set of monitoring ports, including: The system collects real-time live sales data from the live e-commerce supply chain monitoring portals. This data includes a collection of products sold across multiple live streams, the product volume, weight, pricing, price changes, and order generation data for each product within that collection, as well as basic live stream data. Basic live stream data includes the live stream number, start time, estimated end time, and time stamp. Order generation data includes the order number, transaction price, quantity, and delivery address. Price change data includes adjusted prices, effective and ineffective times, and adjustment tags for multiple price adjustments. Adjustment tags include platform subsidies, store discounts, and streamer discounts. Live stream time stamps include peak promotional days, off-peak promotional days, peak non-peak promotional days, and others. Based on the sales product set of each live broadcast, real-time supply data, real-time warehousing data, and real-time after-sales data are collected from the monitoring ports of the live e-commerce supply chain, including the supplier side, warehousing side, and after-sales side. The real-time supply data includes the purchased quantity, purchase arrival time, transported quantity, transport arrival time, transferred quantity, and transferred arrival time of each product in the sales product set of each live broadcast. The real-time warehousing data includes the inventory of each product in the sales product set of each live broadcast. The real-time after-sales data includes the return rate, exchange rate, and negative review and complaint data of each product in the sales product set of each live broadcast. The product complaint data includes multiple negative reviews and multiple complaint records. Real-time logistics data is collected from the logistics side of the live-streaming e-commerce supply chain monitoring port set. The real-time logistics data includes the estimated transportation time, remaining transportation volume, and remaining transportation weight for each transportation trunk line. Based on real-time live sales data, real-time supply data, real-time warehousing data, real-time after-sales data, and real-time logistics data of the live-streaming e-commerce supply chain, real-time monitoring data of the live-streaming e-commerce supply chain is determined.

4. The method for monitoring supply chain information in live-streaming e-commerce according to claim 3, characterized in that, Based on real-time monitoring data of the live-streaming e-commerce supply chain, historical live-streaming data for each live stream is obtained from the real-time live-streaming sales data within the real-time monitoring data, including: Based on the real-time monitoring data of the live-streaming e-commerce supply chain, the live-streaming number of each live-streaming session is obtained from the real-time live-streaming sales data. The historical live-streaming data includes sub-data of multiple historical live-streaming sessions. The sub-data of historical live-streaming sessions includes historical product sets, historical start time, historical end time, historical live-streaming tags, and historical sales product data for each product sold in the historical product set. The historical sales product data includes historical price trend charts and the adjusted price, effective time, and ineffective time of each price adjustment in the historical price trend charts, actual sales volume, historical order data, and adjustment tags.

5. A method for monitoring supply chain information in live-streaming e-commerce according to claim 4, characterized in that, Determine the historical live-stream data for each product in the product set of each live stream within the real-time monitoring data and live-stream sales data, including: Based on the real-time monitoring data of the live-streaming e-commerce supply chain, for each product in the set of products sold in each live stream, the historical live stream data corresponding to each historical product set containing the products sold in the historical live stream data is extracted to determine the historical product live stream data of each product in the set of products sold in each live stream in the real-time monitoring data. The historical product live stream data includes historical live stream sub-data from multiple historical live streams.

6. The method for monitoring supply chain information in live-streaming e-commerce according to claim 5, characterized in that, Based on the historical live-stream data of each product in the product set of each live stream in the real-time monitoring data and real-time live-stream sales data, the product forecast reference data and product pre-sale data of each product in the product set of each live stream in the real-time monitoring data and real-time live-stream sales data are determined, including: Based on the price change data of each product in the product set of each live broadcast in the real-time live broadcast sales data of the live broadcast e-commerce supply chain, a product price trend chart of each product in the product set of each live broadcast in the real-time monitoring data of the live broadcast sales data is drawn. Based on the product price trend chart and basic live streaming data of each product in the sales product set of each live broadcast in the real-time monitoring data, as well as the historical live broadcast tags and historical price trend charts of all historical live broadcast sub-data of each product in the historical product live broadcast data, the product prediction reference data of each product in the sales product set of each live broadcast in the real-time monitoring data is calculated. The product prediction reference data includes the historical sales product data of the sales products corresponding to the historical product sets of multiple historical live broadcast sub-data. The product pricing and order generation data of each product in the sales product set of each live broadcast in the real-time monitoring data and the historical sales product data of the corresponding products in the historical product set of all historical live broadcast sub-data in the product prediction reference data are input into the live broadcast prediction model to determine the product pre-sale data of each product in the sales product set of each live broadcast in the real-time live broadcast sales data. The product pre-sale data includes the predicted selling price, the predicted sales duration, and the predicted sales volume.

7. A method for monitoring supply chain information in live-streaming e-commerce according to claim 6, characterized in that, Based on real-time monitoring data and the pre-sale data of each product in the product set of each live broadcast in real-time live sales data, real-time management data is determined, including: Feature extraction is performed on each negative review and each complaint record in the product complaint data of each product in the real-time after-sales data set of the real-time monitoring data to determine the negative review feature vector and complaint feature vector of each negative review and each complaint record in the product complaint data of each product in the real-time after-sales data set of the real-time monitoring data. The return rate and exchange rate of each product in the sales product set of each live broadcast in the real-time monitoring data, the negative review feature vector of each complaint record and the complaint feature vector in the product complaint data are input into the product after-sales evaluation model to determine the after-sales evaluation value of each product in the sales product set of each live broadcast in the real-time monitoring data. Based on the order generation data of each product in the product set of each live broadcast in the real-time live sales data, the predicted sales volume in the product pre-sale data, the real-time supply data in the real-time monitoring data, and the real-time warehousing data, calculate the out-of-stock risk value of each product in the product set of each live broadcast in the real-time live sales data. Based on the product volume and weight of each product in the product set of each live broadcast in the real-time live sales data, the transaction quantity of all order numbers in the order generation data, and the predicted sales volume, determine the volume transportation capacity requirement and weight transportation capacity requirement of each product in the product set of each live broadcast in the real-time live sales data. The volume and weight capacity requirements of all products sold in the live sales data, as well as the order generation data, and the remaining volume and weight of all transportation trunk lines in the real-time supply data of the real-time monitoring data are input into the capacity risk model to determine the capacity risk value of all products sold in the live sales data. The platform's management model is input into the real-time live sales data, which includes the capacity risk value of all products in the product set of all live streams, the after-sales evaluation value of each product in the product set of each live stream, and the stockout risk value. This allows the platform to determine the decision suggestion label for each product in the product set of each live stream in the real-time monitoring data. Based on the decision suggestion tags of all sales products in the set of all live-streamed sales products in the real-time monitoring data, the real-time management data is determined.

8. A live-streaming e-commerce supply chain information monitoring platform, characterized in that, A method for monitoring supply chain information in live-streaming e-commerce as described in any one of claims 1 to 7, comprising: Acquisition Module: Acquires the set of monitoring ports for the live-streaming e-commerce supply chain, and collects real-time monitoring data of the live-streaming e-commerce supply chain based on the set of monitoring ports. The real-time monitoring data includes real-time live-streaming sales data, real-time supply data, real-time warehousing data, real-time after-sales data, and real-time logistics data. Determine the module: Based on real-time monitoring data of the live-streaming e-commerce supply chain, obtain historical live-streaming data for each live stream in the real-time live-streaming sales data in the real-time monitoring data, and determine the historical live-streaming data of each product in the product set of each live stream in the real-time live-streaming sales data in the real-time monitoring data; Prediction module: Based on the historical live streaming data of each product in the sales product set of each live stream in the real-time monitoring data and real-time live streaming sales data, determine the product prediction reference data and product pre-sale data of each product in the sales product set of each live stream in the real-time monitoring data and real-time live streaming sales data. Management module: Based on real-time monitoring data and real-time live sales data, the pre-sale data of each product in the product set of each live broadcast determines the real-time management data.

Images