Supply chain path optimization method and system for cross-border e-commerce
By extracting supply chain data from procurement documents, performing dimensionality reduction and feature extraction, building profit functions and objective functions, optimizing cross-border e-commerce supply chain paths, the problem of insufficient flexibility and accuracy of supply chain paths is solved, and more efficient supply chain management.
Patent Information
- Application Number
- CN202411448057.8
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-10-17
- Publication Date
- 2024-11-15
- Estimated Expiration
- Not applicable · inactive patent
AI Technical Summary
The cross-border e-commerce supply chain path has shortcomings in flexibility and accuracy, and it is difficult to adapt to the complex international trade environment and changing market demands.
By obtaining text characters in the procurement file, generating supply chain data, performing data dimensionality reduction and feature extraction, building profit functions, generating e-commerce supply scenario sets, and calculating the objective function to optimize the supply chain path.
It improves the flexibility and accuracy of the supply chain path, ensures more accurate optimization results, adapts to different e-commerce supply scenarios, and improves the efficiency and responsiveness of cross-border e-commerce supply chain management.
Smart Images

Figure CN118966477A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the technical field of supply chain path optimization, and in particular to a supply chain path optimization method and system for cross-border e-commerce. Background Art
[0002] With the development of globalization and Internet technology, cross-border e-commerce has risen rapidly and become an important part of international trade. It connects global buyers and sellers through e-commerce platforms and realizes the cross-border circulation of goods and services. However, the supply chain management of cross-border e-commerce faces many challenges, including logistics efficiency, tariff policies, cross-border payments, multilingual and cultural differences. Traditional supply chain management methods are inadequate in the field of cross-border e-commerce. At the same time, due to the involvement of multiple countries and markets, the supply chain path becomes complex and changeable, and traditional management techniques and tools are difficult to adapt to this complexity. In addition, there are many factors to consider in the optimization of the supply chain path of cross-border e-commerce, including laws and regulations of different countries, tax policies, logistics costs and efficiency, transparency and traceability of the supply chain, etc., which leads to the lack of flexibility and ability to respond to market changes in the supply chain path of cross-border e-commerce. Therefore, how to optimize the supply chain path of cross-border e-commerce and improve the flexibility and accuracy of the supply chain path has become an urgent problem to be solved. Summary of the invention
[0003] The present invention provides a supply chain path optimization method and system for cross-border e-commerce, the main purpose of which is to solve the problem of poor flexibility and accuracy of the supply chain path for cross-border e-commerce.
[0004] To achieve the above purpose, the present invention provides a supply chain path optimization method for cross-border e-commerce, comprising: Acquire a procurement document of a target e-commerce supply chain, identify text characters in the procurement document, and generate supply chain data in the procurement document according to the text characters; Performing data dimensionality reduction on the supply chain data to obtain a reduced-dimensionality data set, and extracting target data features from the reduced-dimensionality data set; Calculating price parameters of the target e-commerce supply chain according to the target data characteristics, and constructing a profit function of the target e-commerce supply chain according to the price parameters; Generating an e-commerce supply scenario set of the target e-commerce supply chain according to the profit function, and calculating an objective function of the target e-commerce supply chain according to the profit function; A target e-commerce supply scenario is calculated according to the objective function and the e-commerce supply scenario set, and an optimized supply chain path of the target e-commerce supply chain is determined according to the target e-commerce supply scenario.
[0005] Optionally, the identifying text characters in the procurement document includes: Detecting a text region in each of the procurement documents, and segmenting the procurement documents according to the text region to obtain segmented regions; Identify the text characters in each segmented area to obtain the regional text characters; The text characters of the regions are concatenated according to the position of each of the segmented regions in the procurement document to obtain the text characters in the procurement document.
[0006] Optionally, generating the supply chain data in the procurement document according to the text characters includes: Performing character encoding on the text characters to obtain a character vector sequence; Performing bidirectional semantic feature extraction on the character vector sequence to obtain semantic features of the text characters; Text entities in the text characters are identified according to the semantic features, and supply chain data are extracted from the text characters according to the text entities.
[0007] Optionally, performing data dimensionality reduction on the supply chain data to obtain a dimensionality-reduced data set includes: Dividing the supply chain data into a numerical attribute data set and a categorical attribute data set, and respectively calculating the data distance of the supply chain data in the numerical attribute data set and the categorical attribute data set; The data distance between the numerical attribute data set and the categorical attribute data set is calculated using the following formulas: in, Indicates Supply chain data and The data distance of the supply chain data in the numerical attribute data set, Indicates Supply chain data in The data value of a numeric attribute, Indicates Supply chain data in The data value of a numeric attribute, Represents the total number of numerical attributes in the numerical attribute dataset. Indicates Supply chain data and The data distance of the supply chain data in the classification attribute data set is Indicates Supply chain data in data values of categorical attributes, Indicates Supply chain data in data values of categorical attributes, Represents the total number of categorical attributes in the categorical attribute dataset; Calculating a mixed distance between the supply chain data according to the data distance; The hybrid distance between the supply chain data is calculated using the following formula: in, represents the mixing distance, Indicates the total number of numeric attributes, Indicates Supply chain data and The data distance of the supply chain data in the numerical attribute data set, represents the total number of categorical attributes, Indicates Supply chain data and The data distance of the supply chain data in the classification attribute data set; Calculating the data gradient of each supply chain data of the mixed distance in a preset low-dimensional space; Performing gradient descent on the data gradient using a gradient descent method until the data gradient is less than a preset gradient threshold, thereby obtaining the data position of each supply chain data in the low-dimensional space; A dimension reduction data set corresponding to the supply chain data is determined according to the data location.
[0008] Optionally, extracting target data features from the dimension reduction data set includes: Extracting attribute data of each dimension-reduced data in the dimension-reduced data set, and constructing an attribute data matrix according to the attribute data; Performing data normalization and averaging processing on the attribute data matrix to obtain a standard data matrix; Calculate the eigenvalue contribution rate of the dimension reduction data according to the standard data matrix; The target data features in the dimension reduction data set are determined according to the eigenvalue contribution rate.
[0009] Optionally, constructing a profit function of the target e-commerce supply chain according to the price parameter includes: Calculate the supply chain cost function and the supply chain expected value function of the target e-commerce supply chain according to the price parameter; The supply chain cost function and supply chain expected value function of the target e-commerce supply chain are calculated using the following formula: in, represents the supply chain cost function, Indicates Fixed rental costs for warehouses, Represents the total number of warehouses, No. A decision variable that is 1 if a warehouse is rented and 0 otherwise. Indicates Unit transportation cost per warehouse, Indicates that e-commerce products are delivered to The stock quantity of a warehouse, Represents a collection of customer points. represents the expected value function of the supply chain, Indicates Customer points, Indicates Warehouse to The unit delivery cost for delivering to each customer point, Indicates The exchange rate at the customer's location, Indicates Warehouse to The number of deliveries to each customer point, Indicates The import tariff rate at the customer location, Indicates The unit price of the product at the customer point, Indicates the first The replenishment quantity for each customer point, Express to the Unit delivery cost for delivery to each customer point; Calculating the revenue function of the target e-commerce supply chain according to the price parameter; The revenue function of the target e-commerce supply chain is calculated using the following formula: in, represents the income function, Represents a collection of customer points. Indicates Customer points, Indicates the commission rate of the platform where the target e-commerce supply chain is located, Represents the total number of warehouses, Indicates Warehouse to The number of deliveries to each customer point, Indicates The unit price of the product at the customer point, Indicates The exchange rate at the customer's location, Indicates the first The replenishment quantity for each customer point; A profit function of the target e-commerce supply chain is constructed based on the supply chain cost function, the supply chain expected value function and the revenue function.
[0010] Optionally, generating the e-commerce supply scenario set of the target e-commerce supply chain according to the profit function includes: Acquire historical scenario data of the target e-commerce supply chain, and perform sampling classification with replacement on the historical scenario data to obtain a classified data set; Determine a plurality of random forest classification tree nodes according to the classification data set; A plurality of decision trees are constructed according to the classification tree nodes, and an e-commerce supply scenario set is generated according to the leaf node values of the decision trees.
[0011] Optionally, calculating a target e-commerce supply scenario according to the objective function and the e-commerce supply scenario set includes: generating a particle population according to the e-commerce supply scenario set, and calculating an objective function value of each particle in the particle population according to the objective function; Calculating an inertia weight value of each particle according to the objective function value, and updating the velocity of the particle according to the inertia weight value to obtain a particle velocity; The inertia weight value of each particle is calculated using the following formula: in, Indicates The particle in The inertia weight value at the update iteration, Indicates the maximum value of the preset inertia weight value. Indicates the minimum value of the preset inertia weight value. Indicated in The maximum value of the objective function value at the update iteration is Indicated in The minimum value of the objective function value at the update iteration, Indicates The particle in The historical maximum value of the objective function value at the update iteration; Iterate the particle population according to the particle velocity to obtain an optimized particle population; The optimization objective function value of each particle in the optimized particle population is calculated, and the target e-commerce supply scenario is determined according to the optimization objective function value.
[0012] Optionally, the iteratively updating the particle population according to the particle speed to obtain an optimized particle population includes: Performing a velocity boundary limit on the particle velocity to obtain a target particle velocity; updating the position of the particle population according to the target particle speed to obtain updated particles; Generate an updated particle population according to the updated particles, and calculate the objective function value of each particle in the updated particle population; The updated particle population is iterated according to the objective function value until the number of iterations is greater than a preset number threshold, thereby obtaining an optimized particle population.
[0013] In order to solve the above problems, the present invention also provides a supply chain path optimization system for cross-border e-commerce, the system comprising: A supply chain data identification module is used to obtain a procurement document of a target e-commerce supply chain, identify text characters in the procurement document, and generate supply chain data in the procurement document according to the text characters; A target data feature extraction module is used to perform data dimensionality reduction on the supply chain data to obtain a reduced-dimensionality data set, and extract target data features from the reduced-dimensionality data set; A profit function construction module, used to calculate the price parameters of the target e-commerce supply chain according to the target data characteristics, and to construct the profit function of the target e-commerce supply chain according to the price parameters; An e-commerce supply scenario set and objective function construction module, used to generate an e-commerce supply scenario set of the target e-commerce supply chain according to the profit function, and calculate the objective function of the target e-commerce supply chain according to the profit function; The optimized supply chain path determination module is used to calculate the target e-commerce supply scenario according to the objective function and the e-commerce supply scenario set, and determine the optimized supply chain path of the target e-commerce supply chain according to the target e-commerce supply scenario.
[0014] The embodiment of the present invention can identify supply chain data through the procurement documents of the target e-commerce supply chain, and can extract supply chain data from the actual documents to ensure the accuracy of the supply chain data; perform data dimension reduction on the supply chain data, and extract target data features, which can remove irrelevant or redundant features, reduce the complexity of subsequent profit function calculations, and improve the efficiency and accuracy of supply chain path optimization; construct a profit function for the target e-commerce supply chain, generate a set of random e-commerce supply scenarios faced when making different decisions according to the profit function, improve the flexibility of optimizing the supply chain path, and calculate the objective function according to the profit function, and determine the optimized supply chain path when the profit of the target e-commerce supply chain is maximized according to the objective function and the e-commerce supply scenario set, so as to obtain a more accurate optimized supply chain path. Therefore, the supply chain path optimization method and system for cross-border e-commerce proposed by the present invention can solve the problem of poor flexibility and accuracy of the cross-border e-commerce supply chain path. BRIEF DESCRIPTION OF THE DRAWINGS
[0015] Figure 1 A schematic diagram of a flow chart of a supply chain path optimization method for cross-border e-commerce provided by an embodiment of the present invention; Figure 2 A schematic diagram of a process for performing dimensionality reduction on supply chain data provided by an embodiment of the present invention; Figure 3 A schematic diagram of a process for calculating a target e-commerce supply scenario according to an objective function and an e-commerce supply scenario set provided by an embodiment of the present invention; Figure 4 A functional module diagram of a supply chain path optimization system for cross-border e-commerce provided in one embodiment of the present invention.
[0016] The realization of the purpose, functional features and advantages of the present invention will be further explained in conjunction with embodiments and with reference to the accompanying drawings. DETAILED DESCRIPTION
[0017] It should be understood that the specific embodiments described herein are only used to explain the present invention, and are not used to limit the present invention.
[0018] The embodiment of the present application provides a supply chain path optimization method for cross-border e-commerce. The execution subject of the supply chain path optimization method for cross-border e-commerce includes but is not limited to at least one of the electronic devices such as a server, a terminal, etc. that can be configured to execute the method provided by the embodiment of the present application. In other words, the supply chain path optimization method for cross-border e-commerce can be executed by software or hardware installed on a terminal device or a server device, and the software can be a blockchain platform. The server includes but is not limited to: a single server, a server cluster, a cloud server or a cloud server cluster, etc. The server can be an independent server, or it can be a cloud server that provides cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communications, middleware services, domain name services, security services, content delivery networks (Content Delivery Network, CDN), and basic cloud computing services such as big data and artificial intelligence platforms.
[0019] Reference Figure 1 FIG. 1 is a flow chart of a supply chain path optimization method for cross-border e-commerce provided by an embodiment of the present invention. In this embodiment, the supply chain path optimization method for cross-border e-commerce includes: S1. Obtain a procurement document of a target e-commerce supply chain, identify text characters in the procurement document, and generate supply chain data in the procurement document according to the text characters.
[0020] In an embodiment of the present invention, the target e-commerce supply chain refers to a network structure composed of various links and participants involved in the entire process from production and circulation of goods to the hands of final consumers in cross-border e-commerce activities. The procurement documents include documents involved in various links and participants of the target e-commerce supply chain in previous e-commerce activities, such as procurement contracts, purchase orders, logistics documents, customs declaration documents, etc. The procurement documents can be used to determine the detailed information, quantity, price, delivery date, delivery location, transportation cost, fixed rental costs of each warehouse, unit inventory cost, storage capacity limit, exchange rate and other data of the goods in the target e-commerce supply chain in cross-border e-commerce transactions.
[0021] In an embodiment of the present invention, the identifying text characters in the procurement document includes: Detecting a text region in each of the procurement documents, and segmenting the procurement documents according to the text region to obtain segmented regions; Identify the text characters in each segmented area to obtain the regional text characters; The text characters of the regions are concatenated according to the position of each of the segmented regions in the procurement document to obtain the text characters in the procurement document.
[0022] In an embodiment of the present invention, a preset Paddle OCR model can be used to scan and detect text in a procurement document, and the portion of the procurement document that may contain text is segmented. Each line is marked as a detection box to obtain a segmented area, and OCR recognition is performed on each segmented area to obtain regional text characters. PaddleOCR is an ultra-lightweight OCR (Optical Character Recognition) tool library that can perform text detection, text recognition, and text direction classification, etc., to more accurately identify the regional text characters of each text area.
[0023] In detail, due to the layout of the procurement document and other factors, the recognized regional text characters may have problems such as uneven physical positions and mixed character sequences. Therefore, it is necessary to determine the layout of the text area based on the position information of each segmented area in the procurement document, and then combine the regional text characters to obtain the overall text characters.
[0024] Specifically, a two-dimensional coordinate system can be initialized in the procurement document, the coordinate points of the four vertices of each segmented area in the two-dimensional coordinate system can be determined, the x-axis coordinates of the four vertices of all segmented areas can be sorted in ascending order, and the average of the minimum and maximum x-axis coordinates after sorting can be taken as the x-axis coordinate of the centerline of the procurement document. If the identified segmented area falls on the center line, that is, the x-axis coordinate of the upper left corner of a segmented area is less than the x-axis coordinate of the center line and the x-axis coordinate of the upper right corner is greater than the x-axis coordinate of the center line of the image, then the segmented area distribution of the procurement document is determined to be arranged from top to bottom, otherwise the segmented areas are determined to be distributed from left to right; if the segmented areas of the procurement document are arranged from top to bottom, the identified text content is spliced according to the position order of each detection frame; if the segmented areas are arranged from left to right, the text area to which the segmented area belongs is judged: if the x-axis coordinate of the upper left corner of the segmented area is less than the x-axis coordinate of the center line, then the segmented area is determined to be located in the left half of the image; if the x-axis coordinate of the upper right corner of the identified segmented area is greater than the x-axis coordinate of the center line, then the segmented area is determined to be located in the right half of the procurement document; finally, the identified character sequence is spliced in the order of left to right to obtain the text characters of the procurement document.
[0025] In an embodiment of the present invention, the text characters may include irrelevant information in the procurement documents, such as advertisements for products or services, irrelevant legal terms, suppliers' internal operating procedures, irrelevant images or charts, and personal recommendations or opinions, etc. Therefore, it is necessary to focus on the information in the text characters that is directly related to the e-commerce activities, such as delivery date, delivery location, transportation cost, fixed rental costs of each warehouse, unit inventory cost, storage capacity limit, exchange rate, etc.
[0026] In an embodiment of the present invention, generating the supply chain data in the procurement document according to the text characters includes: Performing character encoding on the text characters to obtain a character vector sequence; Performing bidirectional semantic feature extraction on the character vector sequence to obtain semantic features of the text characters; Text entities in the text characters are identified according to the semantic features, and supply chain data are extracted from the text characters according to the text entities.
[0027] In an embodiment of the present invention, a pre-built embedding layer can be used to encode text characters into vectors to obtain a character vector sequence, and a BiLSTM (bidirectional long short-term memory network) is set after the embedding layer to obtain the semantic features of the context of the character vector sequence to obtain the semantic features of the text characters. Finally, a fully connected layer is used to identify text entities directly related to e-commerce activities in the text characters, such as quantity, price, delivery date, delivery location, transportation cost, fixed rental costs of each warehouse, unit inventory cost, storage capacity limit, exchange rate and other entity information, and then the data corresponding to each text entity is extracted from the text characters to obtain supply chain data, wherein the specific text entity can be a text entity in a pre-set text entity set that needs to be extracted.
[0028] In the embodiment of the present invention, the supply chain data of the target e-commerce supply chain is extracted through procurement documents. The supply chain data can be extracted from the actual documents of the target e-commerce supply chain in previous e-commerce activities, thereby ensuring the accuracy of the supply chain data.
[0029] S2. Performing data dimensionality reduction on the supply chain data to obtain a reduced-dimensionality data set, and extracting target data features from the reduced-dimensionality data set.
[0030] In the embodiment of the present invention, the supply chain data contains a mixture of multiple attributes and has a large dimension, so it is necessary to perform dimensionality reduction processing on it to obtain dimensionality reduction data to improve the efficiency of subsequent data calculation.
[0031] In the embodiment of the present invention, participation Figure 2 As shown, the supply chain data is subjected to data dimensionality reduction to obtain a dimensionality reduction data set, including: S21, dividing the supply chain data into a numerical attribute data set and a categorical attribute data set, and respectively calculating the data distance of the supply chain data in the numerical attribute data set and the categorical attribute data set; S22, calculating the mixed distance between the supply chain data according to the data distance; S23, calculating the data gradient of each supply chain data of the mixed distance in a preset low-dimensional space; S24, performing gradient descent on the data gradient using a gradient descent method until the data gradient is less than a preset gradient threshold, thereby obtaining the data position of each supply chain data in the low-dimensional space; S25. Determine a dimension reduction data set corresponding to the supply chain data according to the data location.
[0032] In an embodiment of the present invention, the supply chain data is divided into numerical data corresponding to numerical attributes and categorical data corresponding to categorical attributes according to the attributes of the supply chain data, to obtain a numerical data attribute set and a categorical data attribute set, and the data distance between the numerical attribute data and the data distance between the categorical attribute data are calculated respectively.
[0033] In the embodiment of the present invention, the data distance between the numerical attribute data set and the categorical attribute data set is calculated using the following formulas: in, Indicates Supply chain data and The data distance of the supply chain data in the numerical attribute data set, Indicates Supply chain data in The data value of a numeric attribute, Indicates Supply chain data in The data value of a numeric attribute, Represents the total number of numerical attributes in the numerical attribute dataset. Indicates Supply chain data and The data distance of the supply chain data in the classification attribute data set is Indicates Supply chain data in data values of categorical attributes, Indicates Supply chain data in data values of categorical attributes, Represents the total number of categorical attributes in the categorical attribute dataset.
[0034] In the embodiment of the present invention, the mixed distance between the supply chain data can be comprehensively calculated through the data distance of the supply chain data in the numerical attribute and the classification attribute.
[0035] Specifically, the mixed distance between the supply chain data is calculated using the following formula: in, represents the mixing distance, Indicates the total number of numeric attributes, Indicates Supply chain data and The data distance of the supply chain data in the numerical attribute data set, represents the total number of categorical attributes, Indicates Supply chain data and The data distance of supply chain data in the classification attribute data set.
[0036] In an embodiment of the present invention, the positions of supply chain data in a preset low-dimensional space are randomly initialized, and these positions can be regarded as model parameters. The mixed distance of the supply chain data in the low-dimensional space is calculated by a mixed distance calculation method, and the minimum value of the mixed distance in the low-dimensional space is used as the objective function. The gradient of the objective function relative to the position of the supply chain data in the low-dimensional space is calculated, which involves deriving the objective function to determine in which directions parameter updates will reduce the distance between data points. The parameters are updated using the gradient descent method, that is, the position of the supply chain data in the low-dimensional space is adjusted along the opposite direction of the gradient until the data gradient is less than the preset gradient threshold, and the data position of each supply chain data in the low-dimensional space is obtained, so that the high-dimensional supply chain data can be mapped to the low-dimensional space, and the local structure between the supply chain data can be maintained as much as possible.
[0037] In an embodiment of the present invention, the target data feature is to extract the most important feature in the dimension reduction data set and remove irrelevant or redundant features, which can reduce the complexity of subsequent profit function calculation and avoid overfitting of data calculation. At the same time, it can save computing resources and improve the efficiency of supply chain path optimization.
[0038] In the embodiment of the present invention, the step of extracting target data features from the dimension reduction data set includes: Extracting attribute data of each dimension-reduced data in the dimension-reduced data set, and constructing an attribute data matrix according to the attribute data; Performing data normalization and averaging processing on the attribute data matrix to obtain a standard data matrix; Calculate the eigenvalue contribution rate of the dimension reduction data according to the standard data matrix; The target data features in the dimension reduction data set are determined according to the eigenvalue contribution rate.
[0039] In an embodiment of the present invention, attribute data is data corresponding to each data attribute in the dimensionality reduction data set. For example, if the dimensionality reduction data is overseas warehouse data, it may include data on multiple attributes such as the import tariff rate of the country where the overseas warehouse is located, the exchange rate of the location of the overseas warehouse, and the inventory size of the overseas warehouse. A data matrix is constructed with the attribute data of each dimensionality reduction data as a row vector to obtain an attribute data matrix.
[0040] In the embodiment of the present invention, data normalization is to perform data normalization processing on each matrix element according to the maximum value and the minimum value in the attribute data matrix, and mean value processing is to subtract the data mean of each column of the attribute data matrix from the value of the column to obtain a standard data matrix.
[0041] Furthermore, the eigenvalue contribution rate of the reduced-dimensionality data is calculated based on the eigenvalues and eigenvectors of the standard data matrix calculated using the principal component analysis method, and then the ratio of the eigenvalue corresponding to each eigenvector to the sum of all eigenvalues is calculated to obtain the eigenvalue contribution rate. The eigenvalues whose eigenvalue contribution rates are greater than a preset threshold are taken as principal components, and the principal components are taken as the target data features in the reduced-dimensionality data set.
[0042] In the embodiment of the present invention, by calculating the target data features, the influence of irrelevant or redundant features on subsequent calculations can be further removed, the complexity of subsequent profit function calculations can be reduced, and the efficiency and accuracy of supply chain path optimization can be improved.
[0043] S3. Calculate price parameters of the target e-commerce supply chain according to the target data characteristics, and construct a profit function of the target e-commerce supply chain according to the price parameters.
[0044] In an embodiment of the present invention, the price parameter is a set of parameters corresponding to different target data features, such as the unit price of goods in the region, the import tariff rate of the country where the overseas warehouse is located, the commission rate of the platform, the exchange rate at the location of the overseas warehouse, the exchange rate at the location of the customer point, the transportation cost between different customer points, the unit unsalable cost and other parameters, which are used to calculate the profit function of the target e-commerce supply chain.
[0045] In detail, since cross-border e-commerce involves many links and regions in e-commerce activities, it is necessary to calculate the parameters in the e-commerce process based on the target feature data, such as the inventory of overseas inventory points in different e-commerce scenarios, unit inventory cost, unit transportation cost and other price parameters involved.
[0046] Specifically, data feature matching can be performed according to the target data features to obtain the price parameters of each target e-commerce supply chain. For example, the unit transportation cost from the wth border warehouse or overseas warehouse to the jth customer point, or the fixed rental cost of the wth overseas warehouse, the unit inventory cost, the exchange rate of the wth overseas warehouse and other price parameters can be calculated based on historical price data.
[0047] In the embodiment of the present invention, constructing the profit function of the target e-commerce supply chain according to the price parameter includes: Calculate the supply chain cost function and the supply chain expected value function of the target e-commerce supply chain according to the price parameter; Calculating the revenue function of the target e-commerce supply chain according to the price parameter; A profit function of the target e-commerce supply chain is constructed based on the supply chain cost function, the supply chain expected value function and the revenue function.
[0048] In the embodiment of the present invention, the supply chain cost function and the supply chain expected value function of the target e-commerce supply chain are calculated using the following formula: in, represents the supply chain cost function, Indicates Fixed rental costs for warehouses, Represents the total number of warehouses, No. A decision variable that is 1 if a warehouse is rented and 0 otherwise. Indicates Unit transportation cost per warehouse, Indicates that e-commerce products are delivered to The stock quantity of a warehouse, Represents a collection of customer points. represents the expected value function of the supply chain, Indicates Customer points, Indicates Warehouse to The unit delivery cost for delivering to each customer point, Indicates The exchange rate at the customer's location, Indicates Warehouse to The number of deliveries to each customer point, Indicates The import tariff rate at the customer location, Indicates The unit price of the product at the customer point, Indicates the first The replenishment quantity for each customer point, Express to the The unit delivery cost for delivering to each customer point.
[0049] In detail, the supply chain cost function and the supply chain expectation function can be used to calculate the cost information such as the rental fees of overseas warehouses, inventory costs, outbound costs, replenishment costs, etc. of cross-border e-commerce in conducting e-commerce activities, so as to accurately take into account the cost of the target e-commerce supply chain.
[0050] In the embodiment of the present invention, the revenue function is the sales revenue of the e-commerce enterprise corresponding to the target e-commerce supply chain.
[0051] Specifically, the revenue function of the target e-commerce supply chain is calculated using the following formula: in, represents the income function, Represents a collection of customer points. Indicates Customer points, Indicates the commission rate of the platform where the target e-commerce supply chain is located, Represents the total number of warehouses, Indicates Warehouse to The number of deliveries to each customer point, Indicates The unit price of the product at the customer point, Indicates The exchange rate at the customer's location, Indicates the first The replenishment quantity for each customer point.
[0052] In an embodiment of the present invention, the profit function of the target e-commerce supply chain is constructed by subtracting the supply chain cost function and the supply chain expected value function from the revenue function to clarify the profit of the target e-commerce supply chain, thereby providing a basis for the subsequent generation of an optimized supply chain path.
[0053] S4. Generate an e-commerce supply scenario set of the target e-commerce supply chain according to the profit function, and calculate the objective function of the target e-commerce supply chain according to the profit function.
[0054] In the embodiment of the present invention, the e-commerce supply scenario set is a random scenario faced by the target e-commerce supply chain when making different decisions. Under different e-commerce supply scenario sets, the target e-commerce supply chain will have different supply chain paths for different e-commerce activities. For example, under different e-commerce supply scenarios, the unit price of goods at the customer point, the replenishment quantity, the inventory quantity of the warehouse, the exchange rate at the customer point, etc. will have different values, and then a more adaptable and flexible optimized supply chain path can be calculated through multiple e-commerce supply scenarios.
[0055] In the embodiment of the present invention, the step of generating the e-commerce supply scenario set of the target e-commerce supply chain according to the profit function includes: Acquire historical scenario data of the target e-commerce supply chain, and perform sampling classification with replacement on the historical scenario data to obtain a classified data set; Determine a plurality of random forest classification tree nodes according to the classification data set; A plurality of decision trees are constructed according to the classification tree nodes, and an e-commerce supply scenario set is generated according to the leaf node values of the decision trees.
[0056] In an embodiment of the present invention, the historical scenario data is the supply chain data of the cross-border e-commerce corresponding to the target e-commerce supply chain within a preset time period, and may include data involved in various links and participants in different e-commerce scenarios in previous e-commerce activities. For example, in different e-commerce scenarios, the inventory, address and country of the overseas warehouse involved will be different. Therefore, the historical scenario data includes multiple historical e-commerce supply scenarios.
[0057] In detail, random forest bootstrap sampling can be used to generate multiple sub-datasets from historical scenario data to obtain classification data sets. Each classification data set is used to train a single decision tree, and then variable values are randomly extracted from the classification data set to determine the classification tree nodes, generating multiple decision trees. The leaf node values of each decision tree are used as e-commerce supply scenarios.
[0058] In the embodiment of the present invention, the objective function is a profit function that needs to be satisfied in order to optimize the supply chain path. The objective function can be used to maximize the profit of the target e-commerce supply chain and improve the accuracy of the supply chain path.
[0059] In the embodiment of the present invention, the objective function is expressed as:
[0060] In the embodiment of the present invention, the profit of the target e-commerce supply chain can be maximized through the objective function, and then various factors in e-commerce activities can be comprehensively considered to calculate a more accurate optimized supply chain path.
[0061] S5. Calculate a target e-commerce supply scenario according to the objective function and the e-commerce supply scenario set, and determine an optimized supply chain path of the target e-commerce supply chain according to the target e-commerce supply scenario.
[0062] In the embodiment of the present invention, the target e-commerce supply scenario is the e-commerce supply scenario corresponding to the target e-commerce supply chain when the profit function is at the maximum value.
[0063] In the embodiment of the present invention, refer to Figure 3 As shown, the calculating the target e-commerce supply scenario according to the objective function and the e-commerce supply scenario set includes: S31, generating a particle population according to the e-commerce supply scenario set, and calculating the objective function value of each particle in the particle population according to the objective function; S32, calculating the inertia weight value of each particle according to the objective function value, and updating the velocity of the particle according to the inertia weight value to obtain the particle velocity; S33, iterating and updating the particle population according to the particle velocity to obtain an optimized particle population; S34. Calculate the optimization objective function value of each particle in the optimized particle population, and determine the target e-commerce supply scenario according to the optimization objective function value.
[0064] In an embodiment of the present invention, parameters corresponding to each e-commerce supply scenario in the e-commerce supply scenario set are used as particles, for example, the unit price of the goods at the location, the import tariff rate of the country where the overseas warehouse is located, the commission rate of the platform, the exchange rate at the location of the overseas warehouse, the exchange rate at the location of the customer point, the transportation cost between different customer points, the unit unsalable cost and other parameters, which can be a multi-dimensional data, and then a particle population is obtained.
[0065] In the embodiment of the present invention, the inertia weight value can reflect the search ability of the particle population. When the particle population has a large inertia weight, if the particles in the population find some good positions, but because the particles have a large inertia weight, the iterative search step of the particles is large, the particles cannot perform a detailed search in the local position of the possible optimal solution, thereby missing the optimal solution. When the particle population has a small inertia weight, the iterative step of the particles in the population is small, which will cause the particles to only search within the local range of the current optimal solution. However, if the particle population does not perform a good search of the global in the early stage, the particle population will fall into the local extreme value too early, resulting in a large gap between the current optimal solution and the global optimal solution. Therefore, the accuracy of subsequent optimization of the particle population can be improved by the inertia weight value.
[0066] In the embodiment of the present invention, the inertia weight value of each particle is calculated using the following formula: in, Indicates The particle in The inertia weight value at the update iteration, Indicates the maximum value of the preset inertia weight value. Indicates the minimum value of the preset inertia weight value. Indicated in The maximum value of the objective function value at the update iteration is Indicated in The minimum value of the objective function value at the update iteration, Indicates The particle in The historical maximum value of the objective function value at the update iteration.
[0067] Specifically, the particle velocity is updated using the following formula to obtain the particle velocity: in, Indicates The particle in The particle velocity at the update iteration, Indicates The particle in The inertia weight value at the update iteration, , are the preset first impact factor and second impact factor respectively. , is a random number in the interval (0, 1), Indicates The individual historical optimal particle position of each particle, Indicates The particle position of a particle at the current moment, represents the global optimal position of the particle.
[0068] In an embodiment of the present invention, the particle position of each particle at the current moment is updated according to the particle speed to obtain an updated particle population, and then the objective function value of each particle in the updated particle population is calculated, and then the particle population can be updated and iterated until the number of particle population update iterations is greater than a preset threshold to obtain an optimized particle population, and the objective function value of each particle in the optimized particle population is calculated to obtain the optimized objective function value, and the e-commerce supply scenario corresponding to the particle with the maximum optimization objective function value is selected as the target e-commerce supply scenario.
[0069] In the embodiment of the present invention, the step of iteratively updating the particle population according to the particle velocity to obtain an optimized particle population includes: Performing a velocity boundary limit on the particle velocity to obtain a target particle velocity; updating the position of the particle population according to the target particle speed to obtain updated particles; Generate an updated particle population according to the updated particles, and calculate the objective function value of each particle in the updated particle population; The updated particle population is iterated according to the objective function value until the number of iterations is greater than a preset number threshold, thereby obtaining an optimized particle population.
[0070] In an embodiment of the present invention, the speed boundary limitation is to limit the particle speed within a speed range. When the particle speed exceeds the speed range, the boundary value of the speed range is used for replacement to obtain the target particle speed. The particle population is cross-operated according to the target particle speed to update the position of the particle population to obtain updated particles.
[0071] Specifically, it is calculated whether the objective function value of the updated particle is greater than the objective function value of each particle in the particle population. If it is greater, the position of the particle in the particle population is updated. If it is not greater, the position of the particle remains unchanged. Then, the updated particle of each particle in the particle population is obtained, and the updated particles are collected to obtain the updated particle population.
[0072] In the embodiment of the present invention, the objective function is used again to calculate the objective function value of each particle in the updated particle population, and the inertia weight value of the particle in each iteration can be calculated, and then the updated particle population is iteratively generated to obtain the optimized particle population.
[0073] In an embodiment of the present invention, a network structure composed of various links and participants involved in the entire process from production and circulation of goods to the hands of final consumers in a target e-commerce supply scenario is used as an optimized supply chain path. For example, the overseas warehouses, warehouse inventory, customer location, transportation costs, etc. involved can be comprehensively considered. The laws and regulations, tax policies, logistics costs and efficiency of different countries can be obtained to obtain a more accurate optimized supply chain path.
[0074] like Figure 4 , which is a functional module diagram of a supply chain path optimization system for cross-border e-commerce provided by an embodiment of the present invention.
[0075] The supply chain path optimization system 400 for cross-border e-commerce of the present invention can be installed in an electronic device. According to the functions implemented, the supply chain path optimization system 400 for cross-border e-commerce can include a supply chain data identification module 401, a target data feature extraction module 402, a profit function construction module 403, an e-commerce supply scenario set and an objective function construction module 404, and an optimized supply chain path determination module 405. The module of the present invention can also be referred to as a unit, which refers to a series of computer program segments that can be executed by an electronic device processor and can complete fixed functions, which are stored in the memory of the electronic device.
[0076] In this embodiment, the functions of each module / unit are as follows: The supply chain data identification module 401 is used to obtain the procurement documents of the target e-commerce supply chain, identify the text characters in the procurement documents, and generate the supply chain data in the procurement documents according to the text characters; The target data feature extraction module 402 is used to perform data dimensionality reduction on the supply chain data to obtain a reduced dimensionality data set, and extract target data features from the reduced dimensionality data set; The profit function construction module 403 is used to calculate the price parameters of the target e-commerce supply chain according to the target data characteristics, and construct the profit function of the target e-commerce supply chain according to the price parameters; The e-commerce supply scenario set and objective function construction module 404 is used to generate the e-commerce supply scenario set of the target e-commerce supply chain according to the profit function, and calculate the objective function of the target e-commerce supply chain according to the profit function; The optimized supply chain path determination module 405 is used to calculate the target e-commerce supply scenario according to the objective function and the e-commerce supply scenario set, and determine the optimized supply chain path of the target e-commerce supply chain according to the target e-commerce supply scenario.
[0077] In detail, each module described in the supply chain path optimization system 400 for cross-border e-commerce in the embodiment of the present invention is used in the same manner as described above. Figures 1 to 3 The same technical means are used as the supply chain path optimization method for cross-border e-commerce described in , and can produce the same technical effects, so I will not go into details here.
[0078] The present invention also provides an electronic device, which may include a processor, a memory, a communication bus and a communication interface, and may also include a computer program stored in the memory and executable on the processor, such as a supply chain path optimization method program for cross-border e-commerce.
[0079] In some embodiments, the processor may be composed of an integrated circuit, for example, it may be composed of a single packaged integrated circuit, or it may be composed of multiple packaged integrated circuits with the same or different functions, including one or more central processing units (CPU), microprocessors, digital processing chips, graphics processors and a combination of various control chips.
[0080] The memory includes at least one type of readable storage medium, and the readable storage medium includes a flash memory, a mobile hard disk, a multimedia card, a card-type memory (e.g., an SD or DX memory, etc.), a magnetic memory, a disk, an optical disk, etc. In some embodiments, the memory may be an internal storage unit of an electronic device, such as a mobile hard disk of the electronic device.
[0081] The communication bus may be a Peripheral Component Interconnect (PCI) bus or an Extended Industry Standard Architecture (EISA) bus, etc. The bus may be divided into an address bus, a data bus, a control bus, etc. The bus is configured to realize connection and communication between the memory and at least one processor, etc.
[0082] The communication interface is used for communication between the above-mentioned electronic device and other devices, including a network interface and a user interface. Optionally, the network interface may include a wired interface and / or a wireless interface (such as a WI-FI interface, a Bluetooth interface, etc.), which is usually used to establish a communication connection between the electronic device and other electronic devices. The user interface may be a display, an input unit (such as a keyboard), and optionally, the user interface may also be a standard wired interface or a wireless interface.
[0083] The figure only shows an electronic device with components. Those skilled in the art will understand that the structure shown in the figure does not constitute a limitation on the electronic device, and may include fewer or more components than shown in the figure, or combine certain components, or arrange the components differently.
[0084] Specifically, the specific implementation method of the processor for the above instructions can refer to the description of the relevant steps in the corresponding embodiment of the accompanying drawings, which will not be repeated here.
[0085] In the several embodiments provided by the present invention, it should be understood that the disclosed devices, systems and methods can be implemented in other ways. For example, the system embodiments described above are only schematic, for example, the division of the modules is only a logical function division, and there may be other division methods in actual implementation.
[0086] The modules described as separate components may or may not be physically separated, and the components shown as modules may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the solution of this embodiment.
[0087] In addition, each functional module in each embodiment of the present invention may be integrated into one processing unit, each unit may exist physically separately, or two or more units may be integrated into one unit. The above-mentioned integrated unit may be implemented in the form of hardware or in the form of hardware plus software functional modules.
[0088] It is obvious to those skilled in the art that the present invention is not limited to the details of the above exemplary embodiments, and that the present invention can be implemented in other specific forms without departing from the spirit or essential characteristics of the present invention.
[0089] Therefore, no matter from which point of view, the embodiments should be regarded as illustrative and non-restrictive, and the scope of the present invention is limited by the appended claims rather than the above description, so it is intended that all changes falling within the meaning and scope of the equivalent elements of the claims are included in the present invention. Any attached figure mark in the claims should not be regarded as limiting the claims involved.
[0090] The embodiments of the present application can acquire and process relevant data based on artificial intelligence technology. Artificial Intelligence (AI) is the theory, method, technology and application system that uses digital computers or machines controlled by digital computers to simulate, extend and expand human intelligence, perceive the environment, acquire knowledge and use knowledge to obtain the best results.
[0091] In addition, it is clear that the word "comprising" does not exclude other units or steps, and the singular does not exclude the plural. Multiple units or systems stated in a system claim can also be implemented by one unit or system through software or hardware. The words first, second, etc. are used to indicate names, and do not indicate any particular order.
[0092] Finally, it should be noted that the above embodiments are only used to illustrate the technical solution of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solution of the present invention can be modified or replaced by equivalents without departing from the spirit and scope of the technical solution of the present invention.
Claims
1. A supply chain path optimization method for cross-border e-commerce, characterized in that: The method comprises: Acquire a procurement document of a target e-commerce supply chain, identify text characters in the procurement document, and generate supply chain data in the procurement document according to the text characters; Performing data dimensionality reduction on the supply chain data to obtain a reduced-dimensionality data set, and extracting target data features from the reduced-dimensionality data set; Calculating price parameters of the target e-commerce supply chain according to the target data characteristics, and constructing a profit function of the target e-commerce supply chain according to the price parameters; Generating an e-commerce supply scenario set of the target e-commerce supply chain according to the profit function, and calculating an objective function of the target e-commerce supply chain according to the profit function; A target e-commerce supply scenario is calculated according to the objective function and the e-commerce supply scenario set, and an optimized supply chain path of the target e-commerce supply chain is determined according to the target e-commerce supply scenario.
2. The supply chain path optimization method for cross-border e-commerce according to claim 1, characterized in that: The identifying text characters in the procurement document comprises: Detecting a text region in each of the procurement documents, and segmenting the procurement documents according to the text region to obtain segmented regions; Identify the text characters in each segmented area to obtain the regional text characters; The text characters of the regions are concatenated according to the position of each of the segmented regions in the procurement document to obtain the text characters in the procurement document.
3. The supply chain path optimization method for cross-border e-commerce according to claim 1, characterized in that: The step of generating the supply chain data in the procurement document according to the text characters comprises: Performing character encoding on the text characters to obtain a character vector sequence; Performing bidirectional semantic feature extraction on the character vector sequence to obtain semantic features of the text characters; Text entities in the text characters are identified according to the semantic features, and supply chain data are extracted from the text characters according to the text entities.
4. The supply chain path optimization method for cross-border e-commerce according to claim 1, characterized in that: The performing data dimensionality reduction on the supply chain data to obtain a dimensionality reduction data set includes: Dividing the supply chain data into a numerical attribute data set and a categorical attribute data set, and respectively calculating the data distance of the supply chain data in the numerical attribute data set and the categorical attribute data set; The data distance between the numerical attribute data set and the categorical attribute data set is calculated using the following formulas: in, Indicates Supply chain data and The data distance of the supply chain data in the numerical attribute data set, Indicates Supply chain data in The data value of a numeric attribute, Indicates Supply chain data in The data value of a numeric attribute, Represents the total number of numerical attributes in the numerical attribute dataset. Indicates Supply chain data and The data distance of the supply chain data in the classification attribute data set is Indicates Supply chain data in data values of categorical attributes, Indicates Supply chain data in data values of categorical attributes, Represents the total number of categorical attributes in the categorical attribute dataset; Calculating a mixed distance between the supply chain data according to the data distance; The hybrid distance between the supply chain data is calculated using the following formula: in, represents the mixing distance, Indicates the total number of numeric attributes, Indicates Supply chain data and The data distance of the supply chain data in the numerical attribute data set, represents the total number of categorical attributes, Indicates Supply chain data and The data distance of the supply chain data in the classification attribute data set; Calculating the data gradient of each supply chain data of the mixed distance in a preset low-dimensional space; Performing gradient descent on the data gradient using a gradient descent method until the data gradient is less than a preset gradient threshold, thereby obtaining the data position of each supply chain data in the low-dimensional space; A dimension reduction data set corresponding to the supply chain data is determined according to the data location.
5. The supply chain path optimization method for cross-border e-commerce according to claim 1, characterized in that: The extracting target data features from the dimension reduction data set includes: Extracting attribute data of each dimension-reduced data in the dimension-reduced data set, and constructing an attribute data matrix according to the attribute data; Performing data normalization and averaging processing on the attribute data matrix to obtain a standard data matrix; Calculate the eigenvalue contribution rate of the dimension reduction data according to the standard data matrix; The target data features in the dimension reduction data set are determined according to the eigenvalue contribution rate.
6. The supply chain path optimization method for cross-border e-commerce according to claim 1, characterized in that: The constructing the profit function of the target e-commerce supply chain according to the price parameter includes: Calculate the supply chain cost function and the supply chain expected value function of the target e-commerce supply chain according to the price parameter; The supply chain cost function and supply chain expected value function of the target e-commerce supply chain are calculated using the following formula: in, represents the supply chain cost function, Indicates Fixed rental costs for warehouses, Represents the total number of warehouses, No. A decision variable that is 1 if a warehouse is rented and 0 otherwise. Indicates Unit transportation cost per warehouse, Indicates that e-commerce products are delivered to The stock quantity of warehouses, Represents a collection of customer points. represents the expected value function of the supply chain, Indicates Customer points, Indicates Warehouse to The unit delivery cost for delivering to customer points, Indicates The exchange rate at the customer's location, Indicates Warehouse to The number of deliveries to each customer point, Indicates The import tariff rate at the customer location, Indicates The unit price of the product at the customer point, Indicates the first The replenishment quantity for each customer point, Express to the Unit delivery cost for delivery to each customer point; Calculating the revenue function of the target e-commerce supply chain according to the price parameter; The revenue function of the target e-commerce supply chain is calculated using the following formula: in, represents the income function, Represents a collection of customer points. Indicates Customer points, Indicates the commission rate of the platform where the target e-commerce supply chain is located, Represents the total number of warehouses, Indicates Warehouse to The number of deliveries to each customer point, Indicates The unit price of the product at the customer point, Indicates The exchange rate at the customer's location, Indicates the first The replenishment quantity for each customer point; A profit function of the target e-commerce supply chain is constructed based on the supply chain cost function, the supply chain expected value function and the revenue function.
7. The supply chain path optimization method for cross-border e-commerce according to claim 1, characterized in that: The step of generating the e-commerce supply scenario set of the target e-commerce supply chain according to the profit function includes: Acquire historical scenario data of the target e-commerce supply chain, and perform sampling classification with replacement on the historical scenario data to obtain a classified data set; Determine a plurality of random forest classification tree nodes according to the classification data set; A plurality of decision trees are constructed according to the classification tree nodes, and an e-commerce supply scenario set is generated according to the leaf node values of the decision trees.
8. The supply chain path optimization method for cross-border e-commerce according to claim 1, characterized in that: The calculating the target e-commerce supply scenario according to the objective function and the e-commerce supply scenario set includes: generating a particle population according to the e-commerce supply scenario set, and calculating an objective function value of each particle in the particle population according to the objective function; Calculating an inertia weight value of each particle according to the objective function value, and updating the velocity of the particle according to the inertia weight value to obtain a particle velocity; The inertia weight value of each particle is calculated using the following formula: in, Indicates The particle in The inertia weight value at the update iteration, Indicates the maximum value of the preset inertia weight value. Indicates the minimum value of the preset inertia weight value. Indicated in The maximum value of the objective function value at the update iteration is Indicated in The minimum value of the objective function value at the update iteration, Indicates The particle in The historical maximum value of the objective function value at the update iteration; Iterate the particle population according to the particle velocity to obtain an optimized particle population; The optimization objective function value of each particle in the optimized particle population is calculated, and the target e-commerce supply scenario is determined according to the optimization objective function value.
9. The supply chain path optimization method for cross-border e-commerce according to claim 8, characterized in that: The updating and iterating the particle population according to the particle speed to obtain an optimized particle population includes: Performing a velocity boundary limit on the particle velocity to obtain a target particle velocity; updating the position of the particle population according to the target particle speed to obtain updated particles; Generate an updated particle population according to the updated particles, and calculate the objective function value of each particle in the updated particle population; The updated particle population is iterated according to the objective function value until the number of iterations is greater than a preset number threshold, thereby obtaining an optimized particle population.
10. A supply chain path optimization system for cross-border e-commerce, characterized in that: The system comprises: A supply chain data identification module is used to obtain a procurement document of a target e-commerce supply chain, identify text characters in the procurement document, and generate supply chain data in the procurement document according to the text characters; A target data feature extraction module is used to perform data dimensionality reduction on the supply chain data to obtain a reduced-dimensionality data set, and extract target data features from the reduced-dimensionality data set; A profit function construction module, used to calculate the price parameters of the target e-commerce supply chain according to the target data characteristics, and to construct the profit function of the target e-commerce supply chain according to the price parameters; An e-commerce supply scenario set and objective function construction module, used to generate an e-commerce supply scenario set of the target e-commerce supply chain according to the profit function, and calculate the objective function of the target e-commerce supply chain according to the profit function; The optimized supply chain path determination module is used to calculate the target e-commerce supply scenario according to the objective function and the e-commerce supply scenario set, and determine the optimized supply chain path of the target e-commerce supply chain according to the target e-commerce supply scenario.
Citation Information
Patent Citations
Method for extracting specific name and unit name attributes in multi-modal data
CN115410185A
Intelligent big data visualization processing method and device, equipment and storage medium
CN115510988A
Cross-modal picture text named entity recognition method and system and electronic equipment
CN116229493A
Supply chain data analysis method and device based on cloud computing, equipment and medium
CN116681462A
Data processing method and system based on product supply chain
CN118410922A
Cited By
Index system construction method based on cross-border e-commerce supply chain toughness evaluation
CN119624101A
