Invoice creation support system, invoice creation support device, invoice creation support method and program

Abstract

This invention provides an invoice creation support system, an invoice creation support device, an invoice creation support method, and a program that can create invoice information appropriately and efficiently. [Solution] The system includes: an intermediate information generation unit 232 that generates intermediate information including determination material information for determining whether the product can be delivered to the destination based on order information including the product and information regarding the delivery of the product; a delivery feasibility determination unit 235 that determines whether the product can be delivered to the destination based on the determination material information and generates a determination result; and an invoice information generation unit 237 that outputs invoice information based on the intermediate information and the determination result.

Description

Technical Field

[0001] The present disclosure relates to an invoice creation support system, an invoice creation support device, an invoice creation support method, and a program.

Background Art

[0002] In recent years, with the spread of the Internet, so-called cross-border e-commerce in which individuals or businesses conduct transactions of goods across borders has become popular. When shipping goods overseas in cross-border e-commerce, it is essential to create an invoice (purchase order) for customs procedures. The invoice needs to accurately describe the item name, price, quantity, and HS code (Harmonized System Code) of the goods.

[0003] Conventionally, the creation of such an invoice has often been performed by a person in charge of a business operator visually checking information on goods and regulations of the destination country and using a manual or semi-automatic system. However, in cross-border e-commerce, there are many cases where various goods are shipped little by little to various countries, and the invoice creation work has become complicated, increasing the burden on the person in charge. In particular, although the HS code is a globally unified item classification number, its identification requires specialized knowledge, so there is a risk of entering an incorrect code.

[0004] In response to such problems, a technique for specifying the HS code from an image or category information of a product has been proposed (see, for example, Patent Document 1).

Prior Art Documents

Patent Documents

[0005]

Patent Document 1

Summary of the Invention

Problems to be Solved by the Invention

[0006] The technology described in Patent Document 1 presents candidate product names and HS codes for products using image recognition of the products, thereby reducing the burden of HS code identification when creating invoices.

[0007] However, the challenges of creating invoices for cross-border e-commerce extend beyond simply identifying HS codes. In reality, exports themselves may be prohibited or restricted depending on the type of product, material, ingredients, or the destination country or region. Furthermore, the types of goods that can be shipped, as well as the information and format required on the invoice, may differ depending on the shipping carrier and method used (e.g., air freight, sea freight).

[0008] The technology described in Patent Document 1 could not determine whether delivery was possible while considering the delivery destination and delivery method, and it was still necessary for the person in charge to check each item individually. As a result, there was still a risk of mistakenly processing orders for items that could not be delivered, or of items being stopped at customs or returned due to incomplete information on the invoice.

[0009] Therefore, this disclosure aims to provide an invoice creation support system, an invoice creation support device, an invoice creation support method, and a program that can create invoice information appropriately and efficiently. [Means for solving the problem]

[0010] The invoice creation support system relating to this disclosure includes: an intermediate information generation unit that generates intermediate information including determination material information for determining whether or not the goods can be delivered to a destination based on order information including goods and information regarding the delivery of the goods; a delivery feasibility determination unit that determines whether or not the goods can be delivered to the destination based on the determination material information and generates a determination result; and an invoice information generation unit that outputs invoice information based on the intermediate information and the determination result.

[0011] The invoice creation support device according to this disclosure comprises: an intermediate information generation unit that generates intermediate information including determination material information for determining whether or not a product can be delivered to a destination based on order information including a product and information regarding the delivery of the product; a delivery feasibility determination unit that determines whether or not a product can be delivered to the destination based on the determination material information and generates a determination result; and an invoice information generation unit that outputs invoice information based on the intermediate information and the determination result.

[0012] The invoice creation support method relating to this disclosure includes the steps of: a computer generating intermediate information including determination material information for determining whether the goods can be delivered to a destination based on order information including the goods and information regarding the delivery of the goods; a computer determining whether the goods can be delivered to the destination based on the determination material information and generating a determination result; and a computer outputting invoice information based on the intermediate information and the determination result.

[0013] The program relating to this disclosure is a program executed by a computer, which causes the computer to perform the following processes: generating intermediate information including determination material information for determining whether or not the goods can be delivered to a destination based on order information including goods and information regarding the delivery of the goods; determining whether or not the goods can be delivered to the destination based on the determination material information and generating a determination result; and outputting invoice information based on the intermediate information and the determination result. [Effects of the Invention]

[0014] According to this disclosure, which has the configuration described above, invoice information can be created appropriately and efficiently. [Brief explanation of the drawing]

[0015] [Figure 1] This is a schematic diagram showing an invoice creation support system according to an embodiment of the present disclosure. [Figure 2] This is a functional block diagram of the order receiving server according to the embodiment of this disclosure. [Figure 3] It is a functional block diagram of an invoice creation support device according to an embodiment of the present disclosure. [Figure 4] It is a diagram showing a configuration example of information stored in a database according to an embodiment of the present disclosure. [Figure 5] It is a functional block diagram of a user terminal according to an embodiment of the present disclosure. [Figure 6] It is a flowchart showing invoice information generation processing by an invoice creation support system according to an embodiment of the present disclosure. [Figure 7] It is a diagram showing an example of a screen displayed on a user terminal according to an embodiment of the present disclosure. [Figure 8] It is a diagram showing an example of a screen displayed on a user terminal according to an embodiment of the present disclosure. [Figure 9] It is a block diagram showing the basic configuration of a computer. [Figure 10] It is a functional block diagram of a modified example of an order receiving server according to an embodiment of the present disclosure. [Figure 11] It is a functional block diagram of a modified example of an invoice creation support device according to an embodiment of the present disclosure.

Mode for Carrying Out the Invention

[0016] Hereinafter, an invoice creation support system 1 according to an embodiment of the present disclosure will be described with reference to the drawings. In all the drawings for explaining the embodiment, the same reference numerals are given to common components, and repeated explanations are omitted.

[0017] <Overview of the System> FIG. 1 is a schematic diagram showing an invoice creation support system 1 according to an embodiment of the present disclosure.

[0018] The invoice creation support system 1 comprises an order server 100, an invoice creation support device 200, a first user terminal 300, and a second user terminal 400. The order server 100, the invoice creation support device 200, the first user terminal 300, and the second user terminal 400 are connected to a network NW and can communicate with each other via the network NW. The network NW includes the Internet. The network NW may also include a LAN (Local Area Network) and / or a WAN (Wide Area Network). Although Figure 1 shows one of each device, in reality, multiple units of each may be used.

[0019] The order server 100, the invoice creation support device 200, the first user terminal 300, and the second user terminal 400 are each composed of one or more computers 900. The basic configuration of the computers 900 will be described later.

[0020] Invoice Creation Support System 1 is a system that supports the entire process from ordering goods to creating invoices in cross-border e-commerce.

[0021] The first user is a customer who purchases goods using a cross-border e-commerce site. They operate a first user terminal 300, such as a PC (Personal Computer), smartphone, or tablet, to place an order for goods through a website provided by the order server 100.

[0022] The order server 100 is a web server that accepts product orders from the first user and is composed of a server for a cross-border e-commerce platform, such as a proxy purchasing service. The order server 100 obtains order information based on the order information transmitted from the first user terminal 300 and processes the order. Details of the order information will be described later. The order server 100 automatically or upon request transmits the order information to the invoice creation support device 200.

[0023] The invoice creation support device 200 performs a series of processes to support invoice creation based on order information obtained from the order server 100. The invoice creation support device 200 utilizes AI (Artificial Intelligence) technology, particularly machine learning models, to determine whether the ordered goods can be delivered to the destination country (exportable) (delivery risk), and generates the information necessary for invoice creation (hereinafter referred to as "invoice information"), taking the delivery risk into consideration. The generated invoice information is transmitted to the order server 100 or provided to the second user terminal 400. In this embodiment, the order server 100 and the invoice creation support device are operated by the same cross-border e-commerce service provider.

[0024] The second user is the invoice manager (hereinafter referred to as the operator or simply the manager) on the business side responsible for invoice creation, and operates the second user terminal 400, such as a PC. The second user terminal 400 displays information provided by the invoice creation support device 200, particularly invoice information, and provides a user interface for receiving confirmation and correction operations from the second user. For example, if the invoice creation support device 200 determines that confirmation by the manager is necessary due to reasons such as a low degree of confidence in the automatic AI determination, a warning or confirmation prompt message will be displayed on the screen of the second user terminal 400. The second user will check the contents and, if necessary, input instructions to correct or confirm the invoice information.

[0025] In this way, the invoice creation support system 1 automates and streamlines the invoice creation process in cross-border e-commerce by coordinating with the order server 100, the invoice creation support device 200, and user terminals 300 and 400. In particular, the invoice creation support device 200 handles a series of processes, including determining product attributes from order information, judging the risk of non-delivery and / or delivery feasibility, and then generating invoice information. This significantly reduces the complex decision-making tasks that previously relied on manual labor, contributing to reduced human error and improved processing speed.

[0026] <System Configurations> Next, the functional configuration of the main device according to this embodiment will be described.

[0027] Figure 2 is a functional block diagram of the order server 100. The order server 100 comprises a communication unit 110, a storage unit 120, and a processing unit 130.

[0028] The communication unit 110 is an interface for data communication with external devices such as the invoice creation support device 200 and the first user terminal 300 via the network NW.

[0029] The memory unit 120 stores the program 121 and the database 122.

[0030] The processing unit 130 implements functions such as the order acquisition unit 131 and the shipping instruction unit 132 by having the CPU (Central Processing Unit) or the like read and execute the program 121 from the storage unit 120.

[0031] The order acquisition unit 131 accepts an order for goods in response to an operation by the first user on the first user terminal 300, generates order information, and transmits it to the invoice creation support device 200. The order information includes at least information about the goods ordered by the first user. The information about the goods includes, for example, the product name, product description, product category, product image, etc. The order information also includes information about the delivery of the goods ordered by the first user. The information about the delivery destination includes, for example, the country, region, address, recipient name, etc. The information about the delivery may also include information about the delivery fee, delivery method, delivery period, and especially information about the customer's specifications or preferences regarding these (prioritizing low cost, prioritizing speed, etc.).

[0032] The shipping instruction unit 132 processes the shipment of goods based on the invoice information generated by the invoice creation support device 200.

[0033] Figure 3 is a functional block diagram of the invoice creation support device 200. The invoice creation support device 200 comprises a communication unit 210, a storage unit 220, and a processing unit 230.

[0034] The communication unit 210 is an interface for data communication with external devices such as the order server 100 and the second user terminal 400 via the network NW.

[0035] The memory unit 220 stores the program 221, the database 222, and the learning model 223.

[0036] Figure 4 shows an example of the structure of information stored in the database 222 used by the invoice creation support device 200. The database 222 is managed by a relational database management system (RDBMS) and consists of multiple tables. The database 222 includes a classification database and an invoice information database.

[0037] The classification database is a database that manages information for classifying products from the unique perspective of cross-border e-commerce businesses. The classification database includes classification tables.

[0038] A classification table is a table that defines combinations of products and classification identifiers used to classify and identify products based on one or more predetermined criteria. For example, a classification table stores product information (product name, product category, etc.) and its corresponding "classification identifier." Specifically, the classification table has columns such as "product keywords," "product category," and "classification identifier." For example, a rule can be defined that for products whose "product keywords" include "perfume" and "eau de cologne," and whose "product category" is "beauty / cosmetics," the "classification identifier" will be "FLM01" (flammable liquid). The classification identification unit 234, described later, searches this classification table based on the product name, category, etc., of the order information to identify the most suitable classification identifier.

[0039] The invoice information database is a collection of data that defines whether goods can be delivered and the rules for invoice information. It defines combinations of classification identifiers and the information to be included in the invoice. The database stores tables such as deliverable countries, delivery methods, and customs codes.

[0040] The deliverable countries table manages a list of countries to which delivery is possible for each classification identifier. The deliverable countries table has "classification identifier" and "country code" as columns. For example, for the classification identifier "BAT02" (lithium-ion battery only), only the country codes of countries to which delivery is permitted are registered. The deliverable country determination unit 235, described later, searches this deliverable countries table using the product's classification identifier and destination country code as keys, and determines whether delivery is possible based on whether a record exists or not.

[0041] The shipping method table manages the available shipping methods for each combination of classification identifier and country code. The shipping method table has "classification identifier," "country code," and "shipping method code" as columns. For example, for the combination of classification identifier "ALC01" (alcoholic beverages) and country code "US" (United States), the shipping method table restricts shipping methods by registering only "SurfaceMail" (sea mail) as the "shipping method code," and not air mail. The shipping method determination unit 236, described later, searches this shipping method table using the product's classification identifier and destination country code as keys and extracts the appropriate shipping method.

[0042] The customs code table manages HS codes. HS codes are codes established internationally or nationally to classify goods. Other names for HS codes include "tariff codes," "statistical commodity codes," and "tariff numbers." After the intermediate information generation unit 232 generates an HS code using the learning model 223, it checks the generated HS code by referring to this tariff code table. For example, it checks whether any non-existent HS codes have been generated.

[0043] These databases and tables in database 222 operate in an interconnected manner. Database 222 is updated as needed by a second user (person in charge) via the second user terminal 400. This allows for rapid response to changes in regulations in various countries and the introduction of new delivery services.

[0044] The learning model 223 includes multiple learning models. The learning model 223 includes a first learning model, a second learning model, and a third learning model. Note that the first, second, and third learning models may be combined into a single model. The first learning model is primarily used by the specific attribute determination unit 233. The first learning model takes order information as input and outputs attribute determination results that recognize the specific attributes of the products in the order information. In this embodiment, the first learning model is a plurality of learning models trained for each of a plurality of predetermined attributes. The second learning model is primarily used by the classification identification unit 234. The second learning model takes at least one of the order information and attribute determination results as input and identifies and outputs the classification identifier corresponding to the product in the order information. The third learning model is primarily used by the intermediate information generation unit 232. The third learning model takes at least one of the order information, attribute determination result, or classification identifier as input and outputs at least one of the product name and customs code (HS code). Each learning model in the 223 learning model can output a degree of confidence regarding its output. The degree of confidence is an indicator of how confident the learning model 223 is about its output. The degree of confidence is expressed as a probability value between 0 and 1, or as a percentage. Other terms for the degree of confidence include "confidence score," "confidence level," or "confidence score."

[0045] The learning model 223 refers to a mathematical model constructed to learn patterns from given data and to perform prediction, classification, or generation on unknown data. In this embodiment, various types of learning models can be used depending on the task. For example, for tasks that generate new text information from text information, such as item names and customs codes, the learning model 223, including Generative AI represented by Large-Scale Language Models (LLMs), can be utilized. This makes it possible to generate appropriate item names and customs codes from unstructured data such as natural language text of product names and product descriptions while understanding the context. As a result, it becomes possible to handle order information containing ambiguous or polysemous expressions, which would be difficult to handle with simple methods such as keyword matching, with high accuracy. Note that the large-scale language model is not limited to a model constructed within the invoice creation support device 200; it may also be configured to utilize external models developed and provided by OpenAI®, Google®, etc., via an API (Application Programming Interface). In this case, the memory unit 220 of the invoice creation support device 200 stores instructions (prompts) corresponding to the first learning model, the second learning model, and the third learning model in advance, and the processing unit 230 is configured to combine the prompts with the information to be input to the large-scale language model and send them to an external model. For example, instructions (prompts) for appropriately generating item names and customs codes are stored in advance, and the processing unit 230 combines these prompts with at least one of the order information, attribute determination results, and classification identifiers and sends them to an external model to obtain the item names and customs codes output by the large-scale language model. Furthermore, the learning model 223 is not limited to this, and may include various machine learning or deep learning models, such as image recognition models like convolutional neural networks (CNNs) for extracting product features from product images, or classification models like support vector machines (SVMs) and random forests for determining the presence or absence of specific attributes such as "battery contained" or "alcohol contained" from product descriptions.

[0046] Returning to Figure 3, we continue the explanation of the invoice creation support device 200.

[0047] The processing unit 230 implements various functional units, including the order information acquisition unit 231, the intermediate information generation unit 232, the specific attribute determination unit 233, the classification determination unit 234, the delivery feasibility determination unit 235, the delivery method determination unit 236, the invoice information generation unit 237, and the learning management unit 238, by having the CPU or other system read and execute the program 221 from the memory unit 220. Details of each of these functional units will be described later. The processing unit 230 executes at least some of the steps of the invoice information generation process. Details of these processes will be described later.

[0048] The order information acquisition unit 231 acquires order information, including information about the product and its delivery, from the order server 100. The order information acquisition unit 231 may acquire new order information by receiving order information transmitted by the order server 100. The order information acquisition unit 231 may access the database 122 of the order server 100 using an API (Application Programming Interface) or the like to acquire new order information. This makes it possible to automate the entire process from order generation to invoice creation without human intervention.

[0049] The intermediate information generation unit 232 generates intermediate information for use in generating invoice information based on the order information. The intermediate information generation unit 232 generates intermediate information using information generated by the intermediate information generation unit 232 itself, as well as information output by the specific attribute determination unit 233 and the classification determination unit 234, which will be described later. The intermediate information generation unit 232, the specific attribute determination unit 233, and the classification identification unit 234 analyze the natural language text (product name, product description) and image data (product image) contained in the order information to understand the specific attributes of the product. In this process, the learning model 223 (particularly large-scale language models and image recognition models) is utilized.

[0050] The intermediate information generation unit 232 first causes the specific attribute determination unit 233 to analyze the order information and output an attribute determination result indicating whether the product in the order information corresponds to a predetermined attribute, and a degree of certainty (first degree of certainty) related to the attribute determination result. Next, the intermediate information generation unit 232 causes the classification identification unit 234 to output a classification identifier corresponding to the product in the order information and its degree of certainty (second degree of certainty) based on the order information and attribute determination result. Then, the intermediate information generation unit 232 analyzes the order information and outputs the customs duty code (e.g., HS code), the item name, and their degrees of certainty (third degree of certainty). Furthermore, the intermediate information generation unit 232 sets a person in charge confirmation flag based on the attribute determination result and degree of certainty. Finally, the intermediate information generation unit 232 generates intermediate information including the item name, customs duty code, classification identifier, and the set value of the person in charge confirmation flag. The item name, customs duty code, classification identifier, and person in charge confirmation flag are determination material information for determining whether the product can be delivered to the destination. Following this flow, the specific attribute determination unit 233 and the classification determination unit 234 will be explained first.

[0051] The specific attribute determination unit 233 causes the first learning model of the learning model 223 to determine, based on the order information, whether the product falls under a predetermined attribute related to the feasibility of delivery, and outputs the determination result and a first degree of certainty related to the determination result. The predetermined attribute is a characteristic of the product that affects the feasibility of delivery. Examples of predetermined attributes include alcohol, batteries, adhesives, products covered by the Washington Convention, and other prohibited products as defined by each country. The specific attribute determination unit 233, for example, if a label is visible in the product image, reads the text from the image using OCR processing and detects strings such as "Alcohol 14%". Based on this detection result, it determines that the item has the alcohol attribute and calculates the degree of certainty. In addition, in the prohibited item determination, if the product description contains a phrase such as "ivory-like carving", it determines that this may violate the Washington Convention.

[0052] The specific attribute determination unit 233 may output a combination of attribute determination result and degree of certainty for each of a plurality of predetermined attributes. In this embodiment, these multiple attribute determinations are performed using a plurality of first learning models (for example, an alcohol determination model, a battery determination model, etc.) that have been trained to specialize in each attribute, thereby achieving higher accuracy in determination.

[0053] For example, the specific attribute determination unit 233 detects descriptions such as "alcohol content XX%" from the product description text in the order information, or recognizes an object that resembles a battery from the product image. The specific attribute determination unit 233 analyzes the order information (for example, product name "XX Company Red Wine 750ml", product description, product image, destination "United States", etc.). At this time, using the first learning model of the learning model 223, it extracts information such as whether the product has the attribute of "alcoholic beverage", its alcohol content, and volume. This generates important "determination information" for determining whether or not a product can be delivered, such as whether or not it falls under the category of hazardous materials or specific regulated items.

[0054] The classification identification unit 234 identifies a classification identifier for the product in the order information. For example, the classification identification unit 234 has the second learning model of the learning model 223 analyze the order information to output product keywords, and then, using those product keywords (e.g., "perfume"), it refers to the classification table in Figure 4 to identify the classification identifier "FLM01" (an internal code meaning flammable material) corresponding to the product category "Beauty & Cosmetics". This allows the system to automatically recognize that this product requires special handling as flammable material (for example, only specific delivery methods can be used). The classification identification unit 234 may output the second learning model along with a second degree of confidence.

[0055] The intermediate information generation unit 232 generates the product name, customs code, and country code based on the order information. Specifically, the intermediate information generation unit 232 causes the third learning model of the learning model 223 to generate (output) the product name, customs code, and country code based on the order information. For example, if the product description states "A handcrafted, warm chair made from cherry wood," the third learning model understands that this is a "wooden chair" and a type of furniture. It then generates appropriate English item names such as "Wooden Chair" or the more general "Wooden Furniture," and presents the corresponding HS code. Furthermore, if the destination is the United States, the third learning model may generate a more detailed code based on the Harmonized Tariff Schedule of the United States (HTSUS). For example, if the product name in the order information is "Nintendo Switch" in Japanese and the product description states "Home video game console," the third learning model will translate this into the English item name "Video Game Console" and identify the corresponding HS code "8528.72" (for television receivers; the code for game consoles varies depending on the item). In this way, automating language translation and the identification of specialized codes eliminates the need for operator intervention and further enhances the level of automation in invoice creation.

[0056] The intermediate information generation unit 232 also causes the third learning model to output a third degree of certainty regarding the output results of the item name and / or tariff code.

[0057] The intermediate information generation unit 232 causes the classification identification unit 234 to identify a classification identifier for the product in the order information, and includes the identified classification identifier in the intermediate information.

[0058] The intermediate information generation unit 232 sets a person in charge confirmation flag to prompt the person in charge to confirm the invoice information. The person in charge confirmation flag indicates whether confirmation by a human person in charge is required. The person in charge confirmation flag can also be referred to as a "confirmation required flag," "human check flag," or "review request indicator." An example screen including the display of the person in charge confirmation flag will be described later.

[0059] The intermediate information generation unit 232 sets the person in charge confirmation flag to enabled or disabled according to predetermined conditions.

[0060] The intermediate information generation unit 232 may set the person in charge confirmation flag based on the determination result of whether the product in the order information falls under a predetermined attribute related to the feasibility of delivery. For example, it may be set to enabled if the determination result is "applicable" and disabled if it is "not applicable". The intermediate information generation unit 232 may set the person in charge confirmation flag based on the logical OR (OR condition) of multiple determination results when there are multiple determination results corresponding to each of multiple predetermined attributes. For example, the person in charge confirmation flag may be set to enabled if there is a determination result of "applicable" for at least one of the multiple predetermined attributes. The intermediate information generation unit 232 may set the person in charge confirmation flag based on whether the product in the order information falls under prohibited items. The intermediate information generation unit 232 may set the person in charge confirmation flag based on the attribute determination result output by the specific attribute determination unit 233. The intermediate information generation unit 232 may set the person in charge confirmation flag based on the logical OR (OR condition) of multiple attribute determination results. For example, if there are three attribute determination results, such as an attribute determination result related to alcohol, an attribute determination result related to batteries, and an attribute determination result related to adhesives, the person in charge confirmation flag may be enabled if at least one of the determination results is "applicable". The intermediate information generation unit 232 may also set the person in charge confirmation flag based on the probability that the item is a prohibited item, calculated based on the degree of certainty of the attribute determination result output by the specific attribute determination unit 233.

[0061] The intermediate information generation unit 232 may set a responsible person confirmation flag based on the degree of confidence regarding the output of the learning model 223. The intermediate information generation unit 232 may enable the person in charge confirmation flag when the degree of confidence regarding the output of the learning model 223 is below a predetermined threshold, and disable it when the degree of confidence is above the predetermined threshold. The intermediate information generation unit 232 may enable the person in charge confirmation flag when the first degree of confidence is below a predetermined threshold, and disable it when the degree of confidence is above the predetermined threshold. The intermediate information generation unit 232 may enable the person in charge confirmation flag when the second degree of confidence is below a predetermined threshold, and disable it when the degree of confidence is above the predetermined threshold. The intermediate information generation unit 232 may enable the person in charge confirmation flag when the third degree of confidence is below a predetermined threshold, and disable it when the degree of confidence is above the predetermined threshold. The intermediate information generation unit 232 may set the person in charge confirmation flag based on the overall degree of confidence calculated based on each degree of confidence.

[0062] Finally, the intermediate information generation unit 232 generates intermediate information by adding judgment material information (item name, customs code, classification identifier of the product in the order information, and person in charge confirmation flag) to the order information.

[0063] The delivery feasibility determination unit 235 determines whether or not the product can be delivered to the destination based on the determination material information included in the intermediate information and generates a determination result.

[0064] The delivery feasibility determination unit 235 determines whether the goods can be delivered to the destination based on the classification identifier and country code included in the intermediate information and the delivery country table in the invoice information database, and generates a delivery determination result. Determining whether or not delivery is possible is equivalent to determining the possibility of delivery, and may generate a binary delivery determination result of "delivery is possible" or "delivery is not possible," or it may generate a delivery determination result that includes a calculated value indicating the probability (also called delivery risk) that delivery is possible. For example, the delivery feasibility determination unit 235 compares the information contained in the intermediate information, such as the item name "alcoholic beverage" and destination country "United States (country code: US)," with the database 222 (the table of deliverable countries in Figure 4, etc.). If the database defines "classification identifier: ALC01 (alcohol), country code: US, delivery not possible," it generates a delivery determination result of "delivery not possible." For example, the delivery feasibility determination unit 235 searches the delivery country table in Figure 4 using the classification identifier "ALC01" (alcohol) and destination country "FR" (France) included in the intermediate information as keys. If a matching record exists in the table, it determines that the item is "delivery possible"; otherwise, it determines that the item is "delivery impossible". In this way, by replacing complex conditional branching with database lookup operations, the decision-making process can be executed quickly and accurately.

[0065] The shipping method determination unit 236 extracts the shipping method for the product based on the intermediate information. For example, the shipping method determination unit 236 searches the shipping method table in Figure 4 using the classification identifier "BAT01" (battery) and destination country "DE" (Germany) included in the intermediate information as keys. As a result, it identifies and extracts "DHL®" and "FedEx®" as available shipping methods. The shipping method determination unit 236 may further extract the shipping method for the product by considering the shipping determination result.

[0066] The invoice information generation unit 237 outputs invoice information based on the intermediate information, the result of the delivery feasibility determination, and the extracted delivery method. Based on the determination that delivery is not possible, the invoice information generation unit 237 generates notification information stating, "This product cannot be delivered to the destination country," or, if it is determined that delivery is possible, it creates formal invoice information using the item name and customs code included in the intermediate information.

[0067] The invoice information generation unit 237 determines the format of the invoice information based on the intermediate information. The invoice information generation unit 237 may also determine the format of the invoice information based on the destination information included in the intermediate information. The invoice information generation unit 237 may also determine the format of the invoice information based on the determined delivery method. The invoice information generation unit 237 uses the item names and customs codes included in the intermediate information to assemble invoice information in an appropriate format according to the destination country and delivery method. The invoice information generation unit 237 may also be configured to determine the format of the invoice information as an internal logic. Conditional branching (for example, if statements or switch statements) based on the destination country code and delivery method code is written in advance in program 221. Then, the invoice information generation unit 237 determines the information to be added to the invoice or changes the format of the item names based on the conditional branching according to the destination country and determined delivery method included in the intermediate information. For example, if the delivery method is "UPS (registered trademark)" and the product includes a battery (classification identifier BAT01), the invoice information generation unit 237 adds the string "PI967-II," which is defined as "additional information BAT01," to the item name on the invoice. Also, if the destination is France (FR) and the product is alcohol (ALC01), the invoice template is dynamically changed to list the item name in the format "[brand name], [volume ml], [alcohol percentage %]".

[0068] The learning management unit 238 trains the learning model using combined data of order information, intermediate information, and invoice information. The learning management unit 238 periodically executes the retraining process when a certain amount of processed data has been accumulated. For example, if the learning model 223 initially determined the item name to be "Toy," but the person in charge corrected it to "Video Game Console," this pair of "order information" and "corrected invoice information" is added to the learning model 223 as new training data. As a result, in the future, similar products will be more likely to be correctly identified as "Video Game Console."

[0069] Figure 5 is a functional block diagram of the first user terminal 300 and the second user terminal 400 (hereinafter collectively referred to as user terminals 300 and 400). User terminals 300 and 400 are equipped with communication units 310 and 410, storage units 320 and 420, control units 330 and 430, display units 340 and 440, and operation units 350 and 450, which are connected via buses 360 and 460. The control units 330 and 430 are composed of a CPU, etc., and control the operation of the entire terminal by executing programs stored in the storage units 320 and 420. The display units 340 and 440 are liquid crystal displays, etc., and the operation units 350 and 450 are keyboards, mice, touch panels, etc.

[0070] The control unit 330 of the first user terminal 300 communicates with the order server 100 by executing an application program stored in the storage unit 320, displays a screen related to the cross-border e-commerce service on the display unit 340, and accepts user operations from the operation unit 350.

[0071] The control unit 430 of the second user terminal 400 communicates with the invoice creation support device 200 by executing an application program stored in the storage unit 420, displays an invoice-related screen on the display unit 440, and accepts user operations from the operation unit 450. The second user terminal 400 controls the display of the invoice information list screen (for example, screen D100 illustrated in Figure 7), the invoice information editing screen (for example, screen D200 illustrated in Figure 8), and other screens, as well as the transitions between screens, in response to the operations of the second user.

[0072] <System Operation> Next, the operation of the invoice creation support system 1, which has the configuration and functions described above, will be explained using a flowchart. The invoice creation support system 1 executes the invoice information generation process.

[0073] Figure 6 is a flowchart showing the invoice information generation process by the invoice creation support system 1 according to the embodiment of this disclosure.

[0074] In step S101, the order information acquisition unit 231 acquires order information from the order server 100 via the network NW. This order information includes information such as the product name, product description, product image, and destination country ordered by the first user.

[0075] In step S102, the specific attribute determination unit 233 determines, based on the acquired order information, whether the product falls under a predetermined attribute related to the feasibility of delivery. The predetermined attribute here refers to the attributes of items that require special attention or procedures for import and export, or are prohibited, such as flammable materials like alcohol, batteries, adhesives, and perfumes, and animal and plant products that may violate the Washington Convention. The specific attribute determination unit 233 makes the determination using multiple learning models 223 (first learning models) that have been specially trained for each attribute. For example, it may input the text of the product description into a natural language processing model to detect the alcohol content, or input the product image into an image recognition model to detect the presence of a battery. At this time, the learning model 223 outputs a "degree of certainty" indicating the certainty of the determination, along with the determination result of whether the item is applicable or not.

[0076] In step S103, the classification identification unit 234 identifies the business-specific "classification identifier" corresponding to the product based on the order information. This process uses the learning model 223 (second learning model). This is done by searching the aforementioned classification table.

[0077] In steps S104 and S105, the intermediate information generation unit 232 generates the "customs duty code (HS code)" and "item name" to be included in the invoice based on the order information. A learning model 223 (third learning model), such as a large-scale language model, is used for this process. The learning model 223 comprehensively interprets the product name, product description, category, and specific attributes determined in step S102, and generates the most appropriate pair of customs duty code and item name. At this time, the learning model 223 also outputs the "degree of certainty" for the generated result.

[0078] In step S106, the intermediate information generation unit 232 sets the "person in charge confirmation flag" based on the processing results up to this point.

[0079] In step S107, the intermediate information generation unit 232 generates intermediate information including the information generated or identified so far, namely the determination result of the specific attribute, classification identifier, customs code, item name, and person in charge confirmation flag. This intermediate information is temporarily stored in the storage unit 220 for use in subsequent processing.

[0080] In step S108, the delivery feasibility determination unit 235 determines whether the goods can be delivered to the destination based on the determination material information (particularly specific attributes and classification identifiers) and destination country information contained in the intermediate information generated in step S107. This determination is made by querying the aforementioned invoice information database (delivery possible country table, etc.). If it is determined that delivery is not possible, the process is interrupted and an error notification is sent to the order server 100 and the second user terminal 400. If it is determined that delivery is possible, the process proceeds to step S109.

[0081] In step S109, that is, in step S108, if it is determined that delivery is possible, the delivery means determination unit 236 extracts available delivery means. Multiple available delivery means may be extracted.

[0082] Finally, in step S110, the invoice information generation unit 237 generates and outputs the final "invoice information" based on the intermediate information (item name, customs code, etc.) corresponding to the goods determined to be deliverable and the extracted delivery method, thereby ending the flow. Possible output destinations include the order server 100, the customs system, or a data format for printing.

[0083] <Screen example> Next, an example of the screen display and user interface (UI) shown on the user terminal 400 will be described with reference to Figures 7 and 8. These screens are mainly displayed by the control unit 430 of the user terminal 400 drawing the data generated by the invoice information generation unit 237 of the invoice creation support device 200 onto the display unit 440.

[0084] Figure 7 shows an example of the invoice information list screen D100. This screen displays a list D110 of orders processed by the invoice creation support device 200. The list D110 displays basic information for each order, such as the "order number," "order date," "destination region," and "item name."

[0085] In particular, this list D110 includes a column indicating whether "person in charge confirmation" is required. For orders with the "person in charge confirmation flag" enabled, "Required" is displayed in this column, allowing the person in charge to quickly identify cases that require confirmation. By selecting the row of the order they wish to confirm from this list, the person in charge will proceed to the invoice information editing screen described later. Alternatively, as shown in Figure 7, a display button D120 may be provided adjacent to the "Required" display.

[0086] Figure 8 shows an example of the invoice information editing screen D200. The invoice information editing screen D200 displays detailed information of the order selected on the invoice information list screen D100, and allows the person in charge to confirm and edit the information.

[0087] The title area D221 at the top of the screen displays the title "Invoice Editing" along with the "Order Number." Additionally, the 담당자 (person in charge) confirmation area D222 clearly indicates that person in charge confirmation is "required."

[0088] In the invoice details area D230, the AI-generated item names and HS codes are listed for each item. These items are displayed as editable text boxes, allowing the person in charge to modify the content as needed.

[0089] The shipping address area D240 displays the shipping address and other details included in the order information.

[0090] The AI ​​verification results area D250 provides important information for personnel to use in their verification work. Here, the "certainty level" (e.g., HS code confirmation level 95%) and the level of risk (e.g., high likelihood of being a prohibited item) for each AI-generated judgment are displayed. Personnel use this information to check the confirmation checkboxes provided for each item. Comments from the AI ​​(e.g., "Please check for prohibited items.") are also displayed, providing specific instructions to the personnel on points that require particular attention.

[0091] After reviewing and correcting all items, the person in charge confirms the edits by pressing the "Save and Return to List" button D290. The confirmed information is collected by the learning management unit 238 and may be used to retrain the learning model 223.

[0092] In this way, orders with the "Responsible Person Confirmation Flag" set to "Enabled" will be displayed as "Requires Confirmation" on the Invoice Information List screen D100. When a representative selects this order, the Invoice Information Editing screen D200 will be displayed, and the reason why confirmation is required (e.g., "High likelihood of prohibited items") and a comment from the AI ​​(e.g., "Please check for prohibited items.") will be displayed in the AI ​​confirmation result area D250. This allows the representative to immediately understand the points that need to be checked and to perform the review work efficiently.

[0093] <Basic Computer Configuration> Figure 9 is a block diagram showing the basic hardware configuration of computer 900. Computer 900 includes a control unit 901, a storage unit 902, a communication unit 903, an input unit 904, and an output unit 905. The control unit 901, storage unit 902, communication unit 903, input unit 904, and output unit 905 are electrically connected to each other via a communication bus 910.

[0094] The control unit 901 includes a CPU (Central Processing Unit, also called a processor) and controls various parts of the computer 900, as well as reading and executing various programs stored in the storage unit 902.

[0095] The memory unit 902 includes a main memory such as DRAM (Dynamic Random Access Memory) and an auxiliary memory such as a hard disk, and is a device that stores various programs for running the operating system and various applications of the computer 900, as well as data used by these programs. The processes shown in the flowcharts described above are realized by the control units 901 of each computer constituting the order server 100, the invoice creation support device 200, and the user terminals 300 and 400 executing the programs stored in their respective memory units 902.

[0096] The communication unit 903 is a device for communicating with external devices and sends and receives data according to instructions from the control unit 901. Each computer comprising the order server 100, the invoice creation support device 200, and the user terminals 300 and 400 uses this communication unit 903 to communicate with other devices, including the network NW shown in Figure 2.

[0097] The input unit 904 is a device that receives input from an external source and supplies it to the control unit 901, and includes, for example, a keyboard, mouse, touch panel, and camera. The output unit 905 is a device that outputs the processing results of the control unit 901 to the outside, and includes, for example, a display and speaker.

[0098] <Program> Here, we will describe the programs for realizing each functional unit of the invoice creation support device 200 according to this embodiment.

[0099] The invoice creation support device 200 is implemented in the computer 900. The operation of each component of the invoice creation support device 200 is stored in the auxiliary storage device of the storage unit 902 in the form of a program. The control unit 901 reads the program from the auxiliary storage device of the storage unit 902, loads it into the main memory of the storage unit 902, and executes the above process according to the program. The control unit 901 also reserves a storage area in the main memory of the storage unit 902 corresponding to the storage unit 120 described above, according to the program.

[0100] Specifically, the program is a program to be executed by a computer 900 which includes a processor and a memory unit, and the program causes the computer to perform the following processes: generating intermediate information including determination material information for determining whether or not the goods can be delivered to the destination based on order information including information about the goods; determining whether or not the goods can be delivered to the destination based on the determination material information and generating a determination result; and outputting invoice information based on the intermediate information and the determination result.

[0101] The auxiliary storage device of the memory unit 902 is an example of a tangible medium that is not temporary. Other examples of tangible mediums that are not temporary include magnetic disks, magneto-optical disks, CD-ROMs, DVD-ROMs, and semiconductor memory connected via the input unit 904. Furthermore, if this program is distributed to the computer 900 via the network NW, the computer 900 that receives the program may load it into the main memory of the memory unit 902 and execute the above processing.

[0102] Furthermore, the program may be intended to implement some of the functions described above. In addition, the program may be a so-called differential file (differential program) that implements the functions described above in combination with other programs already stored in the auxiliary storage device of the memory unit 902.

[0103] According to the invoice creation support system 1 of this embodiment described above, based on order information acquired from the order server 100, the intermediate information generation unit 232 first extracts and generates information (judgment material information) necessary for determining whether delivery is possible from the order information. Next, the delivery feasibility determination unit 235 uses the judgment material information to mechanically determine whether delivery is possible based on a specialized knowledge database or the like. Then, the invoice information generation unit 237 combines the generated intermediate information and the delivery feasibility determination result to output the final invoice information. By performing the processing in this step-by-step manner, complex judgment logic can be systematically executed, making it possible to generate invoice information accurately and quickly without human judgment. This significantly improves the efficiency of invoice creation work in cross-border e-commerce. This configuration not only suggests item names, but can also make more practical and important judgments, such as "whether or not delivery is possible," based on the attributes of the product and destination restrictions.

[0104] This disclosure is not limited to the embodiments described above, and various modifications are possible within the scope of its technical concept.

[0105] (Variation 1) In the above-described embodiment, an example was explained in which the order receiving server 100 and the invoice creation support device 200, which assists in invoice creation, are configured as separate devices. However, these functions may be integrated into a single server device. By configuring it in this way, the overall system configuration can be simplified and the overhead associated with communication between devices can be reduced. Conversely, it is also possible to distribute the functions of the invoice creation support device 200 across multiple servers. For example, the invoice creation support device 200 may be configured to perform pre-processing to generate intermediate information from order information, and the order server 100 may be configured to perform post-processing such as determining whether delivery is possible and generating invoice information based on the intermediate information. Figures 10 and 11 show an example of such a configuration of the order server 100A and invoice creation support device 200A. In addition to the configuration in Figure 2, the order server 100A includes, within the processing unit 130A, a delivery feasibility determination unit 133, a delivery means determination unit 134, and an invoice information generation unit 135 (each function corresponds to the delivery feasibility determination unit 235, delivery means determination unit 236, and invoice information generation unit 237 in the above embodiment) as invoice creation support functions. Furthermore, the storage unit 120A stores a learning model 123 (its function corresponds to the learning model 223 in the above embodiment) in addition to the database 122. The invoice creation support device 200A in Figure 11 includes only an order information acquisition unit 231, an intermediate information generation unit 232, a specific attribute determination unit 233, and a classification identification unit 234 in the processing unit 230A. By distributing the computationally intensive learning model-based processing and the rule-based processing, which mainly involves database lookups, to separate servers, it becomes easier to allocate optimal server specifications according to the characteristics of each process and to ensure scalability.

[0106] (Modification 2) In the above embodiment, the intermediate information was the order information with judgment material information (item name, customs code, classification identifier of the product in the order information, and person in charge confirmation flag) added to it. However, the intermediate information may also be in a form that includes only the order information necessary for the invoice information in addition to the judgment material information.

[0107] (Variation 3) In the above embodiment, the order server 100 and the invoice creation support device 200 were assumed to be operated by the same cross-border e-commerce service provider, but the order server 100 and the invoice creation support device 200 may be operated by different service providers.

[0108] (Modification 4) In step S108 of the above embodiment, if it is determined that delivery is not possible, the process is interrupted and an error notification is sent to the order server 100 and the second user terminal 400, so invoice information is not generated. However, invoice information including the information "delivery not possible" may be generated and displayed in the invoice information list.

[0109] (Variation 5) In the embodiment described above, as shown in Figure 6, a flow was described in which delivery feasibility is determined after determining specific attributes and generating item names. However, the order of processing is not limited to this. For example, after obtaining order information (S101), a screening process may be performed first to refer to the database 222 based on the destination country and general product category, and to early exclude cases where delivery is clearly impossible (e.g., a complete ban on food imports to a specific country). Then, only orders that pass this screening can proceed to a detailed determination process using a computationally expensive learning model (S102-S105). This is expected to further improve the overall processing efficiency of the system.

[0110] <Other Embodiments> The above-described operation flows can be performed not only independently, but also in combination of two or more operation flows. For example, some steps of one operation flow may be added to another operation flow, or some steps of one operation flow may be replaced with some steps of another operation flow. It is not necessary to execute all steps in each flow; only some steps may be executed. Furthermore, the order of steps in each flow may be changed as appropriate.

[0111] A program may be provided that causes a computer to execute each of the processes according to the above embodiment. The program may be recorded on a computer-readable medium. Using a computer-readable medium, it is possible to install the program on a computer. Here, the computer-readable medium on which the program is recorded may be a non-transient storage medium. The non-transient storage medium is not particularly limited, but may be a storage medium such as a CD-ROM or DVD-ROM. Furthermore, the circuits that execute each of the processes performed by the device according to the above embodiment may be integrated, and at least a part of the device may be configured as a semiconductor integrated circuit (chipset, SoC).

[0112] The functions realized by the apparatus according to the above embodiments may be implemented in a circuit or processing circuitry, including a general-purpose processor, an application-specific processor, an integrated circuit, an ASIC (Application Specific Integrated Circuit), a CPU (a Central Processing Unit), conventional circuits, and / or a combination thereof, programmed to realize the described functions. A processor, including transistors and other circuits, is considered a circuit or processing circuitry. A processor may be a programmed processor that executes a program stored in memory. In this disclosure, circuitry, unit, and means are hardware programmed to realize or perform the described functions. Such hardware may be any hardware disclosed herein, or any hardware known to be programmed to realize or perform the described functions. If such hardware is a processor that is considered a type of circuitry, then such circuitry, means, or unit is a combination of hardware and software used to constitute such hardware and / or processor.

[0113] The terms "based on" and "depending on / in response to" used in this disclosure do not mean "based solely on" or "depending solely on" unless otherwise specified. "Based on" means both "based solely on" and "at least partially on." Similarly, "depending on" means both "at least partially on" and "at least partially on." The terms "include," "comprise," and their variations do not mean to include only the listed items, but may include only the listed items or may include additional items in addition to the listed items. Furthermore, the term "or" used in this disclosure is not intended to mean exclusive OR. Additionally, any reference to elements using designations such as "first," "second," etc., used in this disclosure does not generally limit the quantity or order of those elements. These designations may be used herein as a convenient way to distinguish between two or more elements. Therefore, references to the first and second elements do not imply that only two elements may be adopted therein, or that the first element must precede the second element in any way. In this disclosure, where articles are added by translation, such as a, an, and the in English, these articles shall be plural unless it is clearly indicated by the context that they are not.

[0114] Although the embodiments have been described in detail above with reference to the drawings, the specific configuration is not limited to those described above, and various design changes can be made without departing from the gist of the invention.

[0115] (Note) The features of the above-described embodiment are noted below. (Note 1) An intermediate information generation unit 232 generates intermediate information including determination material information for determining whether or not a product can be delivered to the destination, based on order information including information about the product. A delivery feasibility determination unit 235 determines whether or not the product can be delivered to the destination based on the judgment material information and generates a determination result, An invoice information generation unit 237 outputs invoice information based on intermediate information and the judgment result, Equipped with, Invoice creation support system 1. In this configuration, first, the intermediate information generation unit 232 extracts and generates information (determination material information) necessary for determining whether delivery is possible from the order information. Next, the delivery feasibility determination unit 235 uses this determination material information to mechanically determine whether delivery is possible based on a specialized knowledge database or the like. Finally, the invoice information generation unit 237 combines the generated intermediate information with the delivery feasibility determination result to output the final invoice information. By performing the processing in this step-by-step manner, complex determination logic can be executed systematically, making it possible to generate invoice information accurately and quickly without relying on human judgment. This significantly improves the efficiency of invoice creation operations in cross-border e-commerce. (Note 2) The intermediate information generation unit 232 generates intermediate information including the product name and customs code. Invoice creation support system 1 as described in Appendix 1. With this configuration, the intermediate information includes essential invoice items such as the item name and customs code (HS code, etc.), allowing the invoice information generation unit 237 to complete the invoice information using this information as is or with minimal processing. This simplifies the invoice information generation process and improves the overall system processing speed. (Note 3) The system further includes a storage unit 220 that stores a classification database that defines a combination of a product and a classification identifier for classifying and identifying the product from one or more predetermined perspectives. The invoice creation support device 200 further includes a classification identification unit 234 that identifies classification identifiers for products in order information, The intermediate information generation unit 232 generates intermediate information including a classification identifier. Invoice creation support system 1 as described in Appendix 1 or 2. This configuration allows for the management of product characteristics that cannot be expressed by HS codes alone, by introducing "classification identifiers" that classify products from the business's own perspective or based on delivery risks (e.g., hazardous materials, fragile items, etc.). Using these classification identifiers makes it possible to perform subsequent processes such as determining whether a product can be delivered and deciding on the delivery method more detailed and efficient. (Note 4) The storage unit 220 stores an invoice information database that defines combinations of classification identifiers and information to be included in the invoice information. The delivery feasibility determination unit 235 determines whether or not the goods can be delivered to the destination based on the classification identifier and the invoice information database. Invoice creation support system 1 as described in any of the appendices 1 to 3. This configuration allows for flexible adaptation to changes in the decision rules (for example, the addition of items prohibited from being delivered to specific countries) by defining the delivery eligibility logic in an external table as an invoice information database. By simply updating the database values ​​without modifying the program code, decisions can be made in line with the latest regulations, improving system maintainability. (Note 5) It further includes an order information acquisition unit 231 that acquires order information. Invoice creation support system 1 as described in any of the appendices 1 to 4. With this configuration, the invoice creation support device 200 itself has the function of proactively acquiring order information from the order server 100, which makes system coordination smoother. For example, it can acquire order information periodically or automatically in response to a trigger from the order server 100 and start the invoice creation process. (Note 6) The intermediate information generation unit 232 causes the learning model 223 to generate the product name and customs duty code based on order information including the product name, product description, and destination information. Invoice creation support system 1 as described in any of Appendix 1 to 5. This configuration utilizes a learning model 223 that includes generative AI such as a Large-Scale Language Model (LLM), enabling the generation of appropriate item names and customs codes from unstructured data such as natural language text of product names and product descriptions, while understanding the context. This makes it possible to handle order information containing ambiguous or polysemous expressions, which would be difficult to address with simple methods such as keyword matching, with high accuracy. (Note 7) The intermediate information generation unit 232 generates intermediate information including a person in charge confirmation flag to prompt the person in charge to confirm the invoice information. Invoice creation support system 1 as described in any of Appendix 1 to 5. (Note 8) The intermediate information generation unit 232 sets the person in charge confirmation flag to enabled or disabled according to predetermined conditions. Invoice creation support system 1 as described in Appendix 7. According to the configuration described in Appendix 7 and Appendix 8, if the accuracy of the AI's automatic judgment is low, or for products requiring special attention (e.g., products that may be prohibited), enabling (turning on) the person in charge confirmation flag encourages a double check by a human. This avoids the risks associated with full automation and ensures both system reliability and security. It functions as a fail-safe mechanism to prevent critical judgment errors while increasing the automation rate. (Note 9) The intermediate information generation unit 232 is: Based on order information including product name and product description, the first learning model generates the product item name and customs code. The learning model outputs the degree of certainty of the item name and / or tariff code it generated. The person in charge verification flag is set to enabled when the degree of certainty is below a predetermined threshold, and disabled when the degree of certainty is above a predetermined threshold. Invoice creation support system 1 as described in any of Appendix 8. (Note 10) Based on the order information, the learning model 223 further includes a specific attribute determination unit 233 that determines whether or not a product falls under a predetermined attribute related to the feasibility of delivery, and outputs the determination result and degree of certainty of the determination. The intermediate information generation unit 232 sets the person in charge confirmation flag to enabled when the degree of certainty is less than a predetermined threshold, and to disabled when the degree of certainty is equal to or greater than a predetermined threshold. Invoice creation support system 1 as described in any of Appendix 8. (Note 11) The system further includes a specific attribute determination unit 233 that causes the learning model 223 to determine whether or not the product is a prohibited item based on the order information, and outputs the determination result. The intermediate information generation unit 232 sets the person in charge confirmation flag to enabled if the judgment result is "applicable," and to disabled if the judgment result is "not applicable." The invoice creation support system described in Appendix 8. (Note 12) The specific attribute determination unit 233 outputs the determination result and degree of certainty to the learning model 223, which includes multiple learning models that have been trained for each of the multiple predetermined attributes. Invoice creation support system 1 as described in any of Appendix 10. According to the structure outlined in appendices 9, 10, and 11, using the AI's self-assessment ability, "certainty level," as a judgment criterion allows for an objective and dynamic determination of whether or not to set a flag for employee confirmation. By only seeking human intervention when the confidence level is low, i.e., when the AI ​​determines it "lacks confidence," the confirmation process can be narrowed down to cases that truly require it, minimizing the burden on employees while ensuring the reliability of the judgment. (Note 13) The system further includes a delivery method determination unit 236 that determines the delivery method of the product based on intermediate information. Invoice creation support system 1 as described in any of the appendices 1 to 12. This configuration allows for the automatic determination of the optimal shipping method by comprehensively considering factors such as product attributes (e.g., flammable materials, lithium-ion batteries), regulations in the destination country / region, and conditions like cost and delivery speed. This prevents users from mistakenly selecting unavailable shipping methods and ensures smooth delivery. (Note 14) The storage unit 220 stores a database of invoice format descriptions that defines combinations of destination regions and invoice information description formats. The invoice information generation unit 237 determines the format of the invoice information based on the intermediate information. Invoice creation support system 1 as described in any of the appendices 1 to 13. (Note 15) The system further includes a storage unit 220 that stores a database of invoice format definitions that define combinations of delivery methods and invoice information formatting, The invoice information generation unit 237 determines the format of the invoice information based on the determined delivery method and the format database. Invoice creation support system 1 as described in Appendix 13. According to the structure described in appendices 14 and 15, by creating a database of complex invoice formatting rules that differ from country to country or from carrier to carrier, it is possible to consistently generate invoices in the correct format. For example, it is possible to automatically add a field for the "EORI number" required for shipments to the EU, or to automatically insert additional information about batteries requested by specific carriers. This prevents customs clearance problems caused by incomplete information. (Note 16) The system further includes a learning management unit 238 that retrains the learning model 223 using a combination of order information, intermediate information, and the final output invoice information. Invoice creation support system 1 as described in Appendix 6 or 9. With this configuration, the AI's judgment accuracy improves the longer the system is in operation, as the learning model is retrained using the actual processing results (especially the corrected results if they were modified by the person in charge) as training data. This is expected to reduce the number of cases that require confirmation by a person in charge in the future, leading to further improvements in automation and reductions in operating costs. (Note 17) An intermediate information generation unit generates intermediate information including determination material information for determining whether or not the product can be delivered to the destination, based on order information including information about the product. A delivery feasibility determination unit determines whether the product can be delivered to the destination based on the aforementioned determination material information and generates a determination result, An invoice information generation unit outputs invoice information based on the aforementioned intermediate information and the aforementioned determination result, An invoice creation support device equipped with the following features. (Note 18) The computer generates intermediate information, which includes determination information for determining whether the product can be delivered to the destination, based on order information including the product and information regarding the delivery of the product. The computer determines whether the product can be delivered to the destination based on the judgment material information and generates a determination result, The computer outputs invoice information based on the intermediate information and the determination result, An invoice creation support method having the following features. (Note 19) A process that generates intermediate information including determination material information for determining whether the product can be delivered to the destination, based on order information including the product and information regarding the delivery of the said product, A process to determine whether the product can be delivered to the destination based on the aforementioned determination material information and to generate a determination result, A process to output invoice information based on the aforementioned intermediate information and the aforementioned determination result, A program that causes a computer to execute something. [Explanation of Symbols]

[0116] 1: Invoice creation support system 100,100A: Order server 110: Communications Department 120,120A: Storage section 121: Program 122: Database 130, 130A: Processing Unit 131: Order Acquisition Department 132: Shipping Instructions Department 200, 200A: Invoice creation support device 210: Communications Department 220: Storage section 221: Program 222: Database 223,123: Learning Model 230,230A: Processing Unit 231: Order Information Acquisition Department 232: Intermediate Information Generation Unit 233: Specific attribute determination section 234: Classification specific part 235,133:Delivery availability determination unit 236,134:Delivery method determination department 237,135: Invoice Information Generation Unit 238: Learning Management Department 300: First user terminal 400: Second user terminal 900: Computer NW: Network

Claims

1. A storage unit that stores a classification database that defines combinations of goods and classification identifiers for classifying and identifying the goods from one or more perspectives, and a delivery country database that defines combinations of classification identifiers and countries to which delivery is possible, An acquisition unit that acquires order information including product information and destination information including country or region, A classification identification unit analyzes the order information to identify the product in the order information, and refers to the classification database to identify a classification identifier corresponding to the product. An intermediate information generation unit generates determination material information including the identified classification identifier, and generates intermediate information including the order information and the determination material information, A delivery feasibility determination unit that refers to the delivery country database, determines whether the product can be delivered to the destination based on the classification identifier included in the determination material information, and generates a determination result, An invoice information generation unit outputs invoice information based on the aforementioned intermediate information and the aforementioned determination result, An invoice creation support system equipped with the following features.

2. The storage unit stores a learning model obtained by machine learning or deep learning that correlates data including order information and data including the product name and customs code of the goods in the order information, The intermediate information generation unit, based on the order information to be analyzed by the classification identification unit, causes the learning model to generate the item name and customs code of the product in the order information to be analyzed, and generates the determination material information including the item name and customs code of the product. The invoice creation support system according to claim 1.

3. The storage unit takes order information as input and stores prompts for the learning model, which is a generating AI, to output the product name and customs code of the product in the order information. The intermediate information generation unit, based on the prompt and the order information to be analyzed by the classification identification unit, causes the learning model to generate the item name and customs code of the product in the order information to be analyzed, and generates the determination material information including the item name and customs code of the product. The invoice creation support system according to claim 1.

4. The prompt includes instructions for outputting the degree of certainty for the item name and / or the degree of certainty for the customs code generated by the learning model, The intermediate information generation unit causes the degree of certainty to output to the learning model. The invoice creation support system according to claim 3.

5. The storage unit stores a learning model obtained by machine learning or deep learning that correlates order information with attribute determination results that determine whether or not the product in the order information falls under a predetermined attribute related to whether or not it can be delivered, The system further includes a specific attribute determination unit that, based on the order information to be analyzed by the classification identification unit, causes the learning model to determine whether the product in the order information to be analyzed falls under a predetermined attribute related to whether or not it can be delivered, and outputs the attribute determination result of the determination, The aforementioned determination information further includes a person in charge confirmation flag to prompt the person in charge to confirm the invoice information, The intermediate information generation unit sets the person in charge confirmation flag to enable or disable based on the attribute determination result. The invoice creation support system according to claim 1.

6. The storage unit takes order information as input and stores prompts for a learning model, which is a generating AI, to output an attribute determination result that determines whether or not the product in the order information falls under a predetermined attribute related to whether or not it can be delivered. The system further comprises a specific attribute determination unit that, based on the prompt and the order information to be analyzed by the classification identification unit, causes the learning model to determine whether the product in the order information to be analyzed falls under a predetermined attribute related to whether or not it can be delivered, and outputs the attribute determination result of the determination, The aforementioned determination information further includes a person in charge confirmation flag to prompt the person in charge to confirm the invoice information, The intermediate information generation unit sets the person in charge confirmation flag to enable or disable based on the attribute determination result. The invoice creation support system according to claim 1.

7. The predetermined attribute relating to whether or not delivery is possible includes prohibited items, The specified attribute determination unit causes the learning model to determine, based on the order information, whether the product falls under the category of prohibited items as a predetermined attribute, and outputs the attribute determination result of that determination. The invoice creation support system according to claim 5 or 6.

8. The intermediate information generation unit, when it determines in the attribute determination result that the product in the order information corresponds to a predetermined attribute related to the feasibility of delivery, sets the person in charge confirmation flag to enabled. The invoice creation support system according to claim 5 or 6.

9. The specific attribute determination unit causes the plurality of learning models, each trained for each of the plurality of predetermined attributes, to output the determination result. The invoice creation support system according to claim 5 or 6.

10. The storage unit further stores a delivery means table that manages the delivery means available for the combination of the classification identifier and the country code, The system further includes a delivery means determination unit that determines the delivery means of the product based on the intermediate information and the delivery means table. The invoice creation support system according to claim 1.

11. The storage unit further stores a description format database that defines combinations of destination country or region and invoice information description format, The invoice information generation unit determines the format of the invoice information based on the intermediate information and the format database. The invoice creation support system according to claim 1.

12. The storage unit further stores a description format database that defines combinations of delivery means and invoice information description formats, The invoice information generation unit determines the format of the invoice information based on the determined delivery method and the format database. The invoice creation support system according to claim 10.

13. The system further includes a learning management unit that causes the learning model to be retrained using data obtained by correcting the output of the learning model as training data. The invoice creation support system according to claim 2 or 5.

14. A storage unit that stores a classification database that defines combinations of goods and classification identifiers for classifying and identifying the goods from one or more perspectives, and a delivery country database that defines combinations of classification identifiers and countries to which delivery is possible, An acquisition unit that acquires order information including product information and destination information including country or region, A classification identification unit analyzes the order information to identify the product in the order information, and refers to the classification database to identify a classification identifier corresponding to the product. An intermediate information generation unit generates determination material information including the identified classification identifier, and generates intermediate information including the order information and the determination material information, A delivery feasibility determination unit that refers to the delivery country database, determines whether the product can be delivered to the destination included in the order information based on the classification identifier included in the determination material information, and generates a determination result, An invoice information generation unit outputs invoice information based on the aforementioned intermediate information and the aforementioned determination result, An invoice creation support device equipped with the following features.

15. A computer storing a classification database that defines combinations of goods and classification identifiers for classifying and identifying the goods from one or more perspectives, and a delivery country database that defines combinations of classification identifiers and countries to which delivery is possible, The computer obtains order information, which includes information about the product and information about the destination, including the country or region. The steps include: analyzing the order information to identify the product in the order information, and referring to the classification database to identify the classification identifier corresponding to the product; The steps include generating determination material information including the identified classification identifier, and generating intermediate information including the order information and the determination material information, The computer refers to the database of deliverable countries, determines whether the product can be delivered to the destination included in the order information based on the classification identifier included in the determination material information, and generates a determination result. The computer outputs invoice information based on the intermediate information and the determination result, An invoice creation support method having the following features.

16. A process in which a computer stores a classification database that defines combinations of goods and classification identifiers for classifying and identifying the goods from one or more perspectives, and a delivery country database that defines combinations of classification identifiers and countries to which delivery is possible, The process involves a computer acquiring order information, including information about the product and information about the destination, including the country or region. The process involves analyzing the order information to identify the product in the order information, and referring to the classification database to identify the classification identifier corresponding to the product. A process to generate determination material information including the identified classification identifier, and to generate intermediate information including the order information and the determination material information, The computer performs a process of referring to the database of deliverable countries, determining whether the product can be delivered to the destination included in the order information based on the classification identifier included in the determination material information, and generating a determination result. A process to output invoice information based on the aforementioned intermediate information and the aforementioned determination result, A program that causes a computer to execute something.

Images