Supply chain platform data management system and method based on digital logistics

Abstract

The invention discloses a supply chain platform data management system and method based on digital logistics, relates to the technical field of big data analysis, and aims to comprehensively evaluate the commodity quality of merchants through accurate differential evaluation data analysis, analyze commodity problems in differential evaluation by using a natural language processing technology in combination with order data, and improve the commodity quality of the merchants. And the commodity quality hidden danger can be quickly identified, so that the problem processing efficiency and accuracy are improved. Based on historical data and time sequence analysis, the change trend of the number of commodity problems is predicted, the commodity quality risk value of a merchant is accurately calculated, potential quality problems are recognized in advance, risks are reduced, and warehouse resource allocation is optimized. Through comprehensive evaluation of the high-load processing risk value and the order returning risk value, the platform is supported to comprehensively understand the operation capability and the potential risk of the merchant, the load risk of the merchant is quantized through analysis of the order returning data, the platform is helped to recognize the operation bottleneck in advance, warehouse resource optimization under the high load is ensured, and the operation efficiency of the platform is improved.

Description

Technical Field

[0001] The present invention relates to the technical field of big data analysis, and in particular to a supply chain platform data management system and method based on digital logistics. Background Art

[0002] The digital logistics supply chain platform integrates technologies such as the Internet of Things, big data, and artificial intelligence to achieve intelligent and efficient warehouse management. The platform can monitor inventory status in real time, optimize inventory management, accurately predict demand fluctuations, and avoid excess or shortages of inventory. Through intelligent algorithms, the platform improves warehouse space utilization, optimizes resource allocation, and reduces storage and operating costs. At the same time, the platform makes warehouse operations more transparent, and real-time data sharing improves the collaborative efficiency of warehouses and other links in the supply chain, reducing the risk of inventory backlogs and delivery delays. With the help of these technologies, companies can quickly respond to market changes, improve the flexibility and responsiveness of the supply chain, and ultimately enhance the company's competitiveness and customer satisfaction.

[0003] The current digital logistics supply chain platform has a problem of insufficient accuracy in warehouse resource allocation. The main reason is that factors such as merchants' order return rate and product sales evaluation have not been fully incorporated into the risk assessment system, resulting in the platform's insufficient risk assessment of each merchant, which in turn affects the optimization configuration and allocation accuracy of warehouse resources. Summary of the invention

[0004] The purpose of the present invention is to provide a supply chain platform data management system and method based on digital logistics to solve the problems raised in the prior art.

[0005] To achieve the above object, the present invention provides the following technical solution: a supply chain platform data management method based on digital logistics, the supply chain platform data management method comprising the following steps: Step S1, setting a data collection period, collecting the merchant's order data and store rating data within the data collection period, extracting the merchant's negative review data within the data collection period according to the store rating data, the negative review data including logistics problems and product problems, and screening out product problems based on the negative review data analysis; Step S1-1, setting a data collection cycle T, collecting the merchant's order data and store rating data through a data collection interface according to the set data collection cycle T, wherein the data collection interface is represented by an interface connected to the merchant platform, and is used to obtain the merchant's order information and store rating information. The order data is represented by the number of orders of the merchant within the data collection cycle T, and the store rating data is represented by the evaluation of the merchant's product quality and logistics by consumers on the merchant platform; Step S1-2: extracting negative review data within the data collection period T based on the evaluation classification set by the merchant platform, wherein the negative review data includes logistics problems and product problems, and analyzing and screening the product problems from the negative review data in combination with natural language processing, wherein the analysis and screening is performed to identify the merchant's logistics problems and product problems; By setting a data collection cycle and collecting the merchant's order data and store rating data, we can accurately obtain the merchant's negative review information within the cycle, especially the feedback on product and logistics issues. Combining natural language processing technology to analyze and filter negative review data can effectively identify product problems and timely discover potential quality risks, thereby improving the efficiency and accuracy of problem handling.

[0006] Step S2: Obtain the negative review data of the merchant in each historical data collection cycle, analyze it in combination with the number of product problems in the current data collection cycle, and predict the trend of the number of product problems; Step S2-1, obtaining negative review data of the merchant in each historical data collection period T, and counting the number of negative reviews of product problems in each data collection period T; Step S2-2: Based on the statistical number of product problems, the number of product problems is predicted to change. The prediction calculation uses the following formula: ; In the formula, y t It is represented by the number of product problems of the merchant in the tth data collection period of the forecast; u is represented by the mean of the time series; Z i It is represented as the relationship between the commodity problem and the commodity problem in the past p data collection cycles; p is represented as the sum of the number of data collection cycles; y t-i It is represented by the number of product problems in the ti-th data collection cycle in history; X j It is expressed as the relationship between the number of product problems and the prediction error of the qth data collection period; F t-j It is represented as the error term of the historical data collection cycle tj; q is represented as the sum of the error terms; F t It is expressed as the error term of the tth data collection cycle.

[0007] By counting the number of negative reviews of product problems during the historical data collection period, we can accurately understand the development trend of merchants' product problems and provide data support for future product problem predictions. Combining time series analysis and error term adjustment can effectively predict changes in the number of product problems.

[0008] Step S3: Divide the merchant's annual time into multiple quarters with equal time periods, obtain the merchant's order data and store rating data in the same historical quarter, calculate the order saturation based on the order data in the same historical quarter combined with the order data in the current data collection cycle, and calculate the product problem ratio value in the current quarter based on the store rating data in the same historical quarter; Step S3-1, dividing the merchant's annual time, where the annual time is represented by a whole year of a certain year, and the annual time is divided into multiple quarters of equal duration in chronological order, and the quarters of each year are divided in the same way, the time period length of each quarter is equal, and each quarter contains multiple data collection cycles T of the same number; Step S3-2, obtaining the merchant's order data and store rating data in the same historical quarter, where the same historical quarter refers to the time period corresponding to the current quarter in each year in the past, extracting negative review data from the store rating data, and screening out product problems in each same historical quarter through natural language processing analysis; Step S3-3, calculate the average value of the number of product orders and the number of product problems in each historical year up to the same time period as the current quarter, and calculate the order saturation and product problem ratio values ​​by combining the number of orders and the number of product problems in the current year up to the current data collection period T; The calculation process of order saturation is as follows: divide the number of orders in the current year up to the current data collection period T by the average number of commodity orders in each historical year up to the same period as the current quarter; The calculation process of the product problem ratio value is: divide the number of product problems in the current year up to the current data collection period T by the average number of product problems in each historical year up to the same time period as the current quarter.

[0009] By dividing the year into multiple quarters, we can systematically analyze the business performance of merchants and track the changing trends between quarters. Combining historical data with current data, calculating order saturation and product problem ratios helps assess the merchant's operating status and product quality risks.

[0010] Step S4: Calculate the number of product problems in the next data collection cycle based on the predicted trend of the number of product problems, and calculate the product quality risk value of the merchant in combination with the product problem ratio; Step S5: Obtain the merchant's historical order data, analyze and extract the data collection period with the most orders in each year, record it as the best order period of the year, and calculate the merchant's high-load processing risk value based on the data of the best order period of the year; Step S6, set the merchant risk threshold, and collect the merchant's historical return order data, calculate the merchant return order risk value based on the merchant's historical return order data combined with the order saturation, calculate the merchant's comprehensive risk value through the product quality risk value, high-load processing risk value and merchant return order risk value, and evaluate the merchant's warehouse resource allocation based on the merchant risk threshold.

[0011] Step S6-1: Set the merchant risk threshold Q max , and collect the merchant's return order data in each historical year and combine it with the order saturation to calculate the merchant's return order risk value. The merchant return order risk value calculation uses the following formula: ; In the formula, A T It is represented by the merchant’s chargeback risk value in the current data collection period T; d T It is represented by the number of returned orders in the current data collection period T; D T It is represented by the total number of orders of the merchant in the current data collection cycle T; Davg is represented by the average value of orders in each data collection cycle T in the current year; a is represented by the influence coefficient of order saturation on return risk; Step S6-2: Calculate the merchant comprehensive risk value based on the product quality risk value, the high-load processing risk value and the merchant return risk value. The merchant comprehensive risk value is calculated using the following formula: ; In the formula, R Total It is represented by the comprehensive risk value of the merchant; max (HT) is represented by the historical maximum value of the high-load processing risk value; max (AT) is represented by the historical maximum value of the return risk value; r is represented by the interaction coefficient between the product quality risk value and the high-load processing risk value; c is represented by the correlation coefficient between the product quality risk value and the return risk value; Step S6-2: Based on the calculated merchant comprehensive risk value and merchant risk threshold Q max The warehouse risk judgment is conducted on merchants in the supply chain platform. The judgment process is as follows: When R Total ≥Q max When the merchant is deemed to have a risk in the use of warehouse resources, an early warning is issued; When R Total <Q max , it is judged that the merchant's utilization of warehouse resources is normal.

[0012] According to the predicted trend of the number of product problems, the number of product problems in the next data collection cycle is calculated, and the product quality risk value of the merchant is calculated in combination with the product problem ratio value of the current cycle. The product quality risk value is calculated using the following formula: ; In the formula, R T It is represented by the merchant's product quality risk value in the current data collection cycle T; y t It is expressed as the number of product problems in the tth data collection cycle T; k t It is represented as the proportion of commodity problems in the current data collection cycle T; wk is represented as the influence coefficient of the commodity problem proportion on the commodity problem risk value.

[0013] By predicting the number of product problems in the future based on the trend of product problem numbers and calculating the product quality risk value based on the product problem ratio in the current cycle, potential product quality problems can be identified in advance. This method helps merchants adjust strategies in a timely manner, reduce risks, and optimize product quality management.

[0014] The merchant's historical order data is obtained, and the data collection period with the most orders in each year is analyzed and extracted, which is recorded as the best order period of the year. The merchant's high-load processing risk value is calculated based on the data of the best order period of the year. The high-load processing risk value is calculated using the following formula: ; In the formula, H T It is represented by the high-load processing risk value of the merchant in the current data collection cycle T; Represents the maximum order quantity in the historical order data; R T It is expressed as the order ratio value in the current data collection cycle T; V n It is expressed as the volatility coefficient of the order; Gn is expressed as The standard deviation of order fluctuations in the corresponding year; S T Expressed as the seasonal adjustment coefficient in the current data collection period T.

[0015] By analyzing the merchant's historical order data, identifying the best order cycle of the year, and calculating the high-load processing risk value, it helps to assess the merchant's processing capacity under high-load conditions. This can help merchants plan and deploy resources in advance during high-demand periods and reduce potential operational risks.

[0016] Further, a supply chain platform data management system based on digital logistics, the supply chain platform data management system includes a data collection module, a negative review analysis module, a product problem prediction module, a product order analysis module, a risk value calculation module and a comprehensive risk assessment module; The data collection module is used to collect the merchant's order data and evaluation data; the negative review analysis module is used to extract and filter the merchant's negative review data and identify product problems; the product problem prediction module is used to predict the changing trend of the number of product problems based on historical data; the product order analysis module is used to calculate the order saturation and product problem ratio to evaluate the merchant's load and problem incidence rate; the risk value calculation module is used to calculate the product quality and high-load processing risk value based on product problems and order load; the comprehensive risk assessment module is used to comprehensively calculate the merchant's comprehensive risk value; The output end of the data acquisition module is electrically connected to the input end of the negative review analysis module; the output end of the negative review analysis module is electrically connected to the input end of the product problem prediction module; the output end of the product problem prediction module is electrically connected to the input end of the product order analysis module; the output end of the product order analysis module is electrically connected to the input end of the risk value calculation module; the output end of the risk value calculation module is electrically connected to the input end of the comprehensive risk assessment module.

[0017] The data collection module includes an order collection unit and an evaluation collection unit; the order collection unit is used to collect the number of orders from the merchant within the data collection period; the evaluation collection unit is used to collect store evaluation data on the merchant platform; The negative review analysis module includes a negative review extraction unit and a product problem screening unit; the negative review extraction unit is used to extract negative review data from the evaluation data; the product problem screening unit is used to screen out reviews with product quality problems by combining natural language processing technology; The product problem prediction module includes a historical data analysis unit and a product problem trend prediction unit; the historical data analysis unit is used to analyze the historical negative review data of the merchant and count the number of product problems; the product problem trend prediction unit is used to predict the changing trend of product problems based on the historical data and the number of product problems; The commodity order analysis module includes an order saturation calculation unit and a commodity problem ratio calculation unit; the order saturation calculation unit is used to calculate the order saturation of the current data collection cycle and the historical quarter; the commodity problem ratio calculation unit is used to calculate the commodity problem ratio of the current cycle; The risk value calculation module includes a commodity quality risk value calculation unit and a high load risk value calculation unit; the commodity quality risk value calculation unit is used to calculate the commodity quality risk value according to the forecast of the number of commodity problems and the proportion of commodity problems; the high load risk value calculation unit is used to calculate the merchant's high load processing risk value based on order fluctuations and historical data; The comprehensive risk assessment module includes a chargeback risk value calculation unit and a comprehensive risk value calculation unit; the chargeback risk value calculation unit is used to calculate the merchant's chargeback risk value based on historical chargeback data and order saturation; the comprehensive risk value calculation unit is used to calculate the merchant's comprehensive risk value by combining the product quality risk value, high load processing risk value and chargeback risk value.

[0018] Compared with the prior art, the present invention has the following beneficial effects: 1. The present invention can comprehensively evaluate the merchant's product quality through accurate data collection and negative review data analysis. By combining the merchant's order data and store ratings, natural language processing is performed on the product problems in the negative reviews, and product quality risks are quickly identified to improve the efficiency and accuracy of problem handling. This process helps the platform to dynamically adjust warehouse resource allocation based on actual product quality risk conditions, optimize resource utilization, reduce resource waste, and improve platform operating efficiency.

[0019] 2. The present invention can scientifically predict the changing trend of the number of product problems through a product problem prediction method based on historical data and time series analysis. Combined with the product problem ratio and impact coefficient, the merchant's product quality risk value is accurately calculated, so as to identify potential product quality problems in advance, help merchants take timely countermeasures and reduce quality risks. It effectively improves the accuracy of warehouse resource allocation and avoids mismatch or over-allocation of warehouse resources caused by fluctuations in product quality.

[0020] 3. The present invention helps the platform to fully understand the merchant's operational capabilities and the potential risks they face by combining the comprehensive evaluation of the high-load processing risk value and the return order risk value. By analyzing the historical order data, the merchant's order processing capabilities during high-demand periods are identified, and the merchant's load risk is quantified by the return order data. This helps to identify the merchant's operational bottlenecks and risk points in advance, supports the platform to carry out scientific warehouse resource allocation, and ensures that platform resources can be efficiently scheduled and optimized under high load conditions. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] Figure 1 It is a flow chart of a supply chain platform data management method based on digital logistics of the present invention; Figure 2 This is a structural schematic diagram of a supply chain platform data management system based on digital logistics in the present invention. DETAILED DESCRIPTION

[0022] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0023] Embodiment 1: Figure 1 As shown, the present invention provides a technical solution, a supply chain platform data management method based on digital logistics, and the supply chain platform data management method includes the following steps: Step S1, setting a data collection period, collecting the merchant's order data and store rating data within the data collection period, extracting the merchant's negative review data within the data collection period according to the store rating data, the negative review data including logistics problems and product problems, and screening out product problems based on the negative review data analysis; Step S1-1, setting a data collection cycle T, collecting the merchant's order data and store rating data through a data collection interface according to the set data collection cycle T, wherein the data collection interface is represented by an interface connected to the merchant platform, and is used to obtain the merchant's order information and store rating information. The order data is represented by the number of orders of the merchant within the data collection cycle T, and the store rating data is represented by the evaluation of the merchant's product quality and logistics by consumers on the merchant platform; Step S1-2: based on the evaluation grading set by the merchant platform, extract the negative review data within the data collection period T, the negative review data includes logistics problems and product problems, and analyze the negative review data in combination with natural language processing to filter out product problems. The analysis and screening is represented by identifying the merchant's logistics problems and product problems.

[0024] Step S2: Obtain the negative review data of the merchant in each historical data collection cycle, analyze it in combination with the number of product problems in the current data collection cycle, and predict the trend of the number of product problems; Step S2-1, obtaining negative review data of the merchant in each historical data collection period T, and counting the number of negative reviews of product problems in each data collection period T; Step S2-2: Based on the statistical number of product problems, the number of product problems is predicted to change. The prediction calculation uses the following formula: ; In the formula, y t It is represented by the number of product problems of the merchant in the tth data collection period of the forecast; u is represented by the mean of the time series; Z iIt is represented as the relationship between the commodity problem and the commodity problem in the past p data collection cycles; p is represented as the sum of the number of data collection cycles; y t-i It is represented by the number of product problems in the ti-th data collection cycle in history; X j It is expressed as the relationship between the number of product problems and the prediction error of the qth data collection period; F t-j It is represented as the error term of the historical data collection cycle tj; q is represented as the sum of the error terms; F t It is expressed as the error term of the tth data collection cycle.

[0025] Step S3: Divide the merchant's annual time into multiple quarters with equal time periods, obtain the merchant's order data and store rating data in the same historical quarter, calculate the order saturation based on the order data in the same historical quarter combined with the order data in the current data collection cycle, and calculate the product problem ratio value in the current quarter based on the store rating data in the same historical quarter; Step S3-1, dividing the merchant's annual time, where the annual time is represented by a whole year of a certain year, and the annual time is divided into multiple quarters of equal duration in chronological order, and the quarters of each year are divided in the same way, the time period length of each quarter is equal, and each quarter contains multiple data collection cycles T of the same number; Step S3-2, obtaining the merchant's order data and store rating data in the same historical quarter, where the same historical quarter refers to the time period corresponding to the current quarter in each year in the past, extracting negative review data from the store rating data, and screening out product problems in each same historical quarter through natural language processing analysis; Step S3-3, calculate the average value of the number of product orders and the number of product problems in each historical year up to the same time period as the current quarter, and calculate the order saturation and product problem ratio values ​​by combining the number of orders and the number of product problems in the current year up to the current data collection period T; The calculation process of order saturation is as follows: divide the number of orders in the current year up to the current data collection period T by the average number of commodity orders in each historical year up to the same period as the current quarter; The calculation process of the product problem ratio value is: divide the number of product problems in the current year up to the current data collection period T by the average number of product problems in each historical year up to the same time period as the current quarter.

[0026] Step S4: Calculate the number of product problems in the next data collection cycle based on the predicted trend of the number of product problems, and calculate the product quality risk value of the merchant in combination with the product problem ratio; According to the predicted trend of the number of product problems, the number of product problems in the next data collection cycle is calculated, and the product quality risk value of the merchant is calculated in combination with the product problem ratio value of the current cycle. The product quality risk value is calculated using the following formula: ; In the formula, R T It is represented by the merchant's product quality risk value in the current data collection cycle T; y t It is expressed as the number of product problems in the tth data collection cycle T; k t It is represented as the proportion of commodity problems in the current data collection cycle T; wk is represented as the influence coefficient of the commodity problem proportion on the commodity problem risk value.

[0027] Step S5: Obtain the merchant's historical order data, analyze and extract the data collection period with the most orders in each year, record it as the best order period of the year, and calculate the merchant's high-load processing risk value based on the data of the best order period of the year; The merchant's historical order data is obtained, and the data collection period with the most orders in each year is analyzed and extracted, which is recorded as the best order period of the year. The merchant's high-load processing risk value is calculated based on the data of the best order period of the year. The high-load processing risk value is calculated using the following formula: ; In the formula, H T It is represented by the high-load processing risk value of the merchant in the current data collection cycle T; Represents the maximum order quantity in the historical order data; R T It is expressed as the order ratio value in the current data collection cycle T; V n It is expressed as the volatility coefficient of the order; Gn is expressed as The standard deviation of order fluctuations in the corresponding year; S T Expressed as the seasonal adjustment coefficient in the current data collection period T.

[0028] Step S6, set the merchant risk threshold, and collect the merchant's historical return order data, calculate the merchant return order risk value based on the merchant's historical return order data combined with the order saturation, calculate the merchant's comprehensive risk value through the product quality risk value, high-load processing risk value and merchant return order risk value, and evaluate the merchant's warehouse resource allocation based on the merchant risk threshold.

[0029] Step S6-1: Set the merchant risk threshold Q max , and collect the merchant's return order data in each historical year and combine it with the order saturation to calculate the merchant's return order risk value. The merchant return order risk value calculation uses the following formula: ; In the formula, A T It is represented by the merchant’s chargeback risk value in the current data collection period T; d T It is represented by the number of returned orders in the current data collection period T; D T It is represented by the total number of orders of the merchant in the current data collection cycle T; Davg is represented by the average value of orders in each data collection cycle T in the current year; a is represented by the influence coefficient of order saturation on return risk; Step S6-2: Calculate the merchant comprehensive risk value based on the product quality risk value, the high-load processing risk value and the merchant return risk value. The merchant comprehensive risk value is calculated using the following formula: ; In the formula, R Total It is represented by the comprehensive risk value of the merchant; max (HT) is represented by the historical maximum value of the high-load processing risk value; max (AT) is represented by the historical maximum value of the return risk value; r is represented by the interaction coefficient between the product quality risk value and the high-load processing risk value; c is represented by the correlation coefficient between the product quality risk value and the return risk value; Step S6-2: Based on the calculated merchant comprehensive risk value and merchant risk threshold Q max The warehouse risk judgment is conducted on merchants in the supply chain platform. The judgment process is as follows: When R Total ≥Q max When the merchant is deemed to have a risk in the use of warehouse resources, an early warning is issued; When R Total <Q max , it is judged that the merchant's utilization of warehouse resources is normal.

[0030] For example, if we divide the year into four quarters, we have the following data: Merchant order data (historical data): Quarter 1 (historical): Order quantity = 10000; Q2 (historical): Order quantity = 12000; Q3 (historical): Order quantity = 15000; Q4 (historical): Order quantity = 13500; Store rating data (historical data): Q1 (historical): Number of negative reviews due to product issues = 200; Q2 (historical): Number of negative reviews due to product issues = 180; Q3 (historical): Number of negative reviews due to product issues = 250; Q4 (historical): Number of negative reviews due to product issues = 220; Current year data: Q1 (current): Order quantity = 11,500 Q1 (current): Number of negative reviews due to product issues = 210; After natural language processing analysis of the product ratings and negative reviews (current quarter), it was determined that there were 210 product issues; Historical negative review statistics: Q1 (historical): Number of negative reviews due to product issues = 200; Q2 (historical): Number of negative reviews due to product issues = 180; Q3 (historical): Number of negative reviews due to product issues = 250; Q4 (historical): Number of negative reviews due to product issues = 220; According to the forecast calculation, the following formula can be used to calculate: y t+1 =0.4*210+0.68212.5=84+127.5=211.5; According to the prediction calculation formula, the number of negative reviews due to product problems in the next quarter will be 211.5.

[0031] Calculating the average of the number of historical orders yields: (10000+12000+15000+13500) / 4=12875; The order saturation of the current quarter is calculated as: 11500 / 12875≈0.89; Calculate the product problem ratio and insert the data to get: (200+180+250+220) / 4≈0.99; Calculation of commodity quality risk value: R T =240*0.99=237.6; High load handling risk value calculation: H T =15000*0.89*0.1*1500*1.1=1464750; Calculation of chargeback risk value: A T = (150 / 11500) * (1 + 0.05 * 0.89) ≈ 0.0142; Set the merchant risk threshold Q max=15000000, the interaction coefficient between product quality risk value and high load processing risk value r=0.5; the correlation coefficient between product quality risk value and return order risk value c=0.2; when put into the merchant comprehensive risk value calculation formula, we can get: R Total =1464750+0.0142+0.5*237.6+0.2*0.0142=1464868.8; 15000000>1464868.8, which means that the merchant’s utilization of warehouse resources is normal.

[0032] Embodiment 2, as Figure 2 As shown, the present invention provides a supply chain platform data management system based on digital logistics, which includes a data collection module, a bad review analysis module, a commodity problem prediction module, a commodity order analysis module, a risk value calculation module and a comprehensive risk assessment module; The data collection module is used to collect the merchant's order data and evaluation data; the negative review analysis module is used to extract and filter the merchant's negative review data and identify product problems; the product problem prediction module is used to predict the changing trend of the number of product problems based on historical data; the product order analysis module is used to calculate the order saturation and product problem ratio to evaluate the merchant's load and problem incidence rate; the risk value calculation module is used to calculate the product quality and high-load processing risk value based on product problems and order load; the comprehensive risk assessment module is used to comprehensively calculate the merchant's comprehensive risk value; The output end of the data acquisition module is electrically connected to the input end of the bad review analysis module; the output end of the bad review analysis module is electrically connected to the input end of the product problem prediction module; the output end of the product problem prediction module is electrically connected to the input end of the product order analysis module; the output end of the product order analysis module is electrically connected to the input end of the risk value calculation module; the output end of the risk value calculation module is electrically connected to the input end of the comprehensive risk assessment module; The data collection module includes an order collection unit and an evaluation collection unit; the order collection unit is used to collect the number of orders from the merchant within the data collection period; the evaluation collection unit is used to collect store evaluation data on the merchant platform; The negative review analysis module includes a negative review extraction unit and a product problem screening unit; the negative review extraction unit is used to extract negative review data from the evaluation data; the product problem screening unit is used to screen out reviews with product quality problems by combining natural language processing technology; The product problem prediction module includes a historical data analysis unit and a product problem trend prediction unit; the historical data analysis unit is used to analyze the historical negative review data of the merchant and count the number of product problems; the product problem trend prediction unit is used to predict the changing trend of product problems based on the historical data and the number of product problems; The commodity order analysis module includes an order saturation calculation unit and a commodity problem ratio calculation unit; the order saturation calculation unit is used to calculate the order saturation of the current data collection cycle and the historical quarter; the commodity problem ratio calculation unit is used to calculate the commodity problem ratio of the current cycle; The risk value calculation module includes a commodity quality risk value calculation unit and a high load risk value calculation unit; the commodity quality risk value calculation unit is used to calculate the commodity quality risk value according to the forecast of the number of commodity problems and the proportion of commodity problems; the high load risk value calculation unit is used to calculate the merchant's high load processing risk value based on order fluctuations and historical data; The comprehensive risk assessment module includes a chargeback risk value calculation unit and a comprehensive risk value calculation unit; the chargeback risk value calculation unit is used to calculate the merchant's chargeback risk value based on historical chargeback data and order saturation; the comprehensive risk value calculation unit is used to calculate the merchant's comprehensive risk value by combining the product quality risk value, high load processing risk value and chargeback risk value.

[0033] It will be apparent to those skilled in the art that the invention is not limited to the details of the exemplary embodiments described above and that the invention can be implemented in other specific forms without departing from the spirit or essential features of the invention. Therefore, the embodiments should be considered exemplary and non-limiting in all respects, and the scope of the invention is defined by the appended claims rather than the foregoing description, and it is intended that all variations falling within the meaning and scope of the equivalent elements of the claims be included in the invention. Any reference numeral in a claim should not be considered as limiting the claim to which it relates.

Claims

1. A supply chain platform data management method based on digital logistics, characterized by: The supply chain platform data management method comprises the following steps: Step S1, setting a data collection period, collecting the merchant's order data and store rating data within the data collection period, extracting the merchant's negative review data within the data collection period based on the store rating data, the negative review data including logistics problems and product problems, and screening out product problems based on the negative review data analysis; Step S2: Obtain the negative review data of the merchant in each historical data collection cycle, analyze it in combination with the number of product problems in the current data collection cycle, and predict the trend of the number of product problems; Step S3: Divide the merchant's annual time into multiple quarters with equal time periods, obtain the merchant's order data and store rating data in the same historical quarter, calculate the order saturation based on the order data in the same historical quarter combined with the order data in the current data collection cycle, and calculate the product problem ratio value in the current quarter based on the store rating data in the same historical quarter; Step S4: Calculate the number of product problems in the next data collection cycle based on the predicted trend of the number of product problems, and calculate the product quality risk value of the merchant in combination with the product problem ratio; Step S5: Obtain the merchant's historical order data, analyze and extract the data collection period with the most orders in each year, record it as the best order period of the year, and calculate the merchant's high-load processing risk value based on the data of the best order period of the year; Step S6, set the merchant risk threshold, and collect the merchant's historical return order data, calculate the merchant return order risk value based on the merchant's historical return order data combined with the order saturation, calculate the merchant's comprehensive risk value through the product quality risk value, high-load processing risk value and merchant return order risk value, and evaluate the merchant's warehouse resource allocation based on the merchant risk threshold.

2. A supply chain platform data management method based on digital logistics according to claim 1, characterized in that: The specific steps of step S1 are as follows: Step S1-1, setting a data collection cycle T, collecting the merchant's order data and store rating data through a data collection interface according to the set data collection cycle T, wherein the data collection interface is represented by an interface connected to the merchant platform, and is used to obtain the merchant's order information and store rating information. The order data is represented by the number of orders of the merchant within the data collection cycle T, and the store rating data is represented by the evaluation of the merchant's product quality and logistics by consumers on the merchant platform; Step S1-2: based on the evaluation grading set by the merchant platform, extract the negative review data within the data collection period T, the negative review data includes logistics problems and product problems, and analyze the negative review data in combination with natural language processing to filter out product problems. The analysis and screening is represented by identifying the merchant's logistics problems and product problems.

3. A supply chain platform data management method based on digital logistics according to claim 2, characterized in that: The specific steps of step S2 are as follows: Step S2-1, obtaining negative review data of the merchant in each historical data collection period T, and counting the number of negative reviews of product problems in each data collection period T; Step S2-2: Based on the statistical number of product problems, the number of product problems is predicted to change. The prediction calculation uses the following formula: ； In the formula, y t It is represented by the number of product problems of the merchant in the tth data collection period of the forecast; u is represented by the mean of the time series; Z i It is represented as the relationship between the commodity problem and the commodity problem in the past p data collection cycles; p is represented as the sum of the number of data collection cycles; y t-i It is represented by the number of product problems in the ti-th data collection cycle in history; X j It is expressed as the relationship between the number of product problems and the prediction error of the qth data collection period; F t-j It is represented as the error term of the historical data collection cycle tj; q is represented as the sum of the error terms; F t It is expressed as the error term of the tth data collection cycle.

4. A supply chain platform data management method based on digital logistics according to claim 3, characterized in that: The specific steps of step S3 are as follows: Step S3-1, dividing the merchant's annual time, where the annual time is represented by a whole year of a certain year, and the annual time is divided into multiple quarters of equal duration in chronological order, and the quarters of each year are divided in the same way, the time period length of each quarter is equal, and each quarter contains multiple data collection cycles T of the same number; Step S3-2, obtaining the merchant's order data and store rating data in the same historical quarter, where the same historical quarter refers to the time period corresponding to the current quarter in each year in the past, extracting negative review data from the store rating data, and screening out product problems in each same historical quarter through natural language processing analysis; Step S3-3, calculate the average value of the number of product orders and the number of product problems in each historical year up to the same time period as the current quarter, and calculate the order saturation and product problem ratio values ​​by combining the number of orders and the number of product problems in the current year up to the current data collection period T; The calculation process of order saturation is as follows: divide the number of orders in the current year up to the current data collection period T by the average number of commodity orders in each historical year up to the same period as the current quarter; The calculation process of the product problem ratio value is: divide the number of product problems in the current year up to the current data collection period T by the average number of product problems in each historical year up to the same time period as the current quarter.

5. A supply chain platform data management method based on digital logistics according to claim 4, characterized in that: In step S4, the number of product problems in the next data collection cycle is calculated based on the predicted trend of the number of product problems, and the product quality risk value of the merchant is calculated in combination with the product problem ratio value of the current cycle. The product quality risk value is calculated using the following formula: ； In the formula, R T It is represented by the merchant's product quality risk value in the current data collection cycle T; y t It is expressed as the number of product problems in the tth data collection cycle T; k t It is represented as the proportion of commodity problems in the current data collection cycle T; wk is represented as the influence coefficient of the commodity problem proportion on the commodity problem risk value.

6. A supply chain platform data management method based on digital logistics according to claim 5, characterized in that: In step S5, the merchant's historical order data is obtained, and the data collection period with the most orders in each year is analyzed and extracted, which is recorded as the best order period of the year. The merchant's high-load processing risk value is calculated based on the data of the best order period of the year. The high-load processing risk value is calculated using the following formula: ； In the formula, H T It is represented by the high-load processing risk value of the merchant in the current data collection cycle T; Represents the maximum order quantity in the historical order data; R T It is expressed as the order ratio value in the current data collection cycle T; V n It is expressed as the volatility coefficient of the order; Gn is expressed as Standard deviation of order fluctuations in the corresponding year; S T Expressed as the seasonal adjustment coefficient in the current data collection period T.

7. A supply chain platform data management method based on digital logistics according to claim 6, characterized in that: The specific steps of step S6 are as follows: Step S6-1: Set the merchant risk threshold Q max , and collect the merchant's return order data in each historical year and combine it with the order saturation to calculate the merchant's return order risk value. The merchant return order risk value calculation uses the following formula: ； In the formula, A T It is represented by the merchant’s chargeback risk value in the current data collection period T; d T It is represented by the number of returned orders in the current data collection period T; D T It is represented by the total number of orders of the merchant in the current data collection cycle T; Davg is represented by the average value of orders in each data collection cycle T in the current year; a is represented by the influence coefficient of order saturation on return risk; Step S6-2: Calculate the merchant comprehensive risk value based on the product quality risk value, the high-load processing risk value and the merchant return risk value. The merchant comprehensive risk value is calculated using the following formula: ； In the formula, R Total It is represented by the comprehensive risk value of the merchant; max (HT) is represented by the historical maximum value of the high-load processing risk value; max (AT) is represented by the historical maximum value of the return risk value; r is represented by the interaction coefficient between the product quality risk value and the high-load processing risk value; c is represented by the correlation coefficient between the product quality risk value and the return risk value; Step S6-2: Based on the calculated merchant comprehensive risk value and merchant risk threshold Q max The warehouse risk judgment is conducted on merchants in the supply chain platform. The judgment process is as follows: When R Total ≥Q max When the merchant is deemed to have a risk in the use of warehouse resources, an early warning is issued; When R Total <Q max , it is judged that the merchant's utilization of warehouse resources is normal.

8. A supply chain platform data management system based on digital logistics, which is applied to a supply chain platform data management method based on digital logistics according to any one of claims 1 to 7, characterized in that: The supply chain platform data management system includes a data collection module, a negative review analysis module, a product problem prediction module, a product order analysis module, a risk value calculation module and a comprehensive risk assessment module; The data collection module is used to collect the merchant's order data and evaluation data; the negative review analysis module is used to extract and filter the merchant's negative review data and identify product problems; the product problem prediction module is used to predict the trend of the number of product problems based on historical data; The commodity order analysis module is used to calculate order saturation and commodity problem ratio to evaluate the merchant's load and problem occurrence rate; The risk value calculation module is used to calculate the commodity quality and high-load processing risk values ​​according to commodity problems and order load; The comprehensive risk assessment module is used to comprehensively calculate the comprehensive risk value of the merchant; The output end of the data acquisition module is electrically connected to the input end of the negative review analysis module; the output end of the negative review analysis module is electrically connected to the input end of the product problem prediction module; the output end of the product problem prediction module is electrically connected to the input end of the product order analysis module; the output end of the product order analysis module is electrically connected to the input end of the risk value calculation module; the output end of the risk value calculation module is electrically connected to the input end of the comprehensive risk assessment module.

9. A supply chain platform data management system based on digital logistics according to claim 8, characterized in that: The data collection module includes an order collection unit and an evaluation collection unit; the order collection unit is used to collect the number of orders from the merchant within the data collection period; the evaluation collection unit is used to collect store evaluation data on the merchant platform; The negative review analysis module includes a negative review extraction unit and a product problem screening unit; the negative review extraction unit is used to extract negative review data from the evaluation data; The product problem screening unit is used to screen out evaluations of product quality problems by combining natural language processing technology; The product problem prediction module includes a historical data analysis unit and a product problem trend prediction unit; the historical data analysis unit is used to analyze the merchant's historical negative review data and count the number of product problems; the product problem trend prediction unit is used to predict the changing trend of product problems based on historical data and the number of product problems.

10. A supply chain platform data management system based on digital logistics according to claim 8, characterized in that: The commodity order analysis module includes an order saturation calculation unit and a commodity problem ratio calculation unit; the order saturation calculation unit is used to calculate the order saturation of the current data collection cycle and the historical quarter; The commodity problem ratio calculation unit is used to calculate the commodity problem ratio of the current cycle; The risk value calculation module includes a commodity quality risk value calculation unit and a high load risk value calculation unit; The commodity quality risk value calculation unit is used to calculate the commodity quality risk value according to the commodity problem quantity prediction and commodity problem ratio; The high-load risk value calculation unit is used to calculate the high-load processing risk value of the merchant based on order fluctuations and historical data; The comprehensive risk assessment module includes a chargeback risk value calculation unit and a comprehensive risk value calculation unit; The chargeback risk value calculation unit is used to calculate the merchant's chargeback risk value based on historical chargeback data and order saturation; The comprehensive risk value calculation unit is used to calculate the comprehensive risk value of the merchant by combining the product quality risk value, the high-load processing risk value and the return order risk value.

Images