Information processing systems, information processing methods, and programs

Abstract

Regarding documents, the aim is to reflect the diverse ways in which business partners provide documents or alternative information, while suppressing an increase in manual labor, and generating data that is easy to manage and process by subsequent systems. [Solution] The information processing server 2 includes a server information processing unit 12 which has the function of acquiring report data relating to a report, where the data type is not limited to a specific type, and even if the data type is the same, the manner in which information about the report is recorded may differ, identifying the data type of the report data, performing information processing on the report data according to the data type, and generating unified format data that holds item values ​​for one or more specific items relating to the report in a specific data structure.

Description

Technical Field

[0001] The present invention relates to an information processing system that executes processing related to forms, an information processing method by the information processing system, and a program executed by a computer of the information processing system.

Background Art

[0002] Currently, there are many enterprises, individuals, and other entities (hereinafter referred to as "activity entities") in the world that have business partners (a concept including customers here) and conduct economic activities by conducting transactions with business partners. In most cases, activity entities have multiple business partners and conduct transactions with multiple business partners simultaneously. And basically, in the process of transactions with business partners, activity entities receive forms such as purchase orders and invoices, or information substituting for forms (hereinafter, forms and information substituting for forms are collectively referred to as "form-related information") from business partners. It is rare for all business partners to provide form-related information in a unified manner, and each business partner provides form-related information to activity entities in various ways. Hereinafter, the method by which a business partner provides form-related information is referred to as a "form providing method". For example, focusing on a purchase order, which is one type of form. Regarding a purchase order, a business partner may provide the purchase order by fax, provide PDF data related to the purchase order via email, a chat application, or other communication tools, provide CSV data related to the purchase order, use the recording function of an answering machine to provide information related to the order as a voice message, or provide information related to the order as an input message via a communication tool. Also, the format of the purchase order may vary depending on the business partner, and the data structure of the CSV data may vary depending on the business partner. Also, voice messages and input messages basically have no format, and the content mode can be freely determined by the creator.

[0003] Regarding forms, Patent Document 1 describes the following technology: FAX management system A aggregates FAX data B (including FAX data corresponding to purchase orders) received by FAX devices at each location, processes the aggregated FAX data B according to user instructions, and then transmits the processed FAX data C to a FAX device at a designated location according to user instructions. The above technology is described. According to Patent Document 1, since FAX data sent to each location is aggregated and managed, the processing of received FAXes can be carried out without any geographical constraints between each location and each person in charge. [Prior art documents] [Patent Documents]

[0004] [Patent Document 1] Japanese Patent Publication No. 2024-089891 [Overview of the project] [Problems that the invention aims to solve]

[0005] As mentioned above, a key characteristic is the diversity of methods for providing reports. However, the organizations involved had a need to reflect this characteristic regarding reports, to generate data that is easy to manage, easy to process by subsequent systems, and that minimizes the increase in manual labor.

[0006] This invention was made to solve these problems, and with regard to forms, it aims to generate data that is easy to manage and easy to process by subsequent systems, while suppressing an increase in manual labor, reflecting the characteristic that there are various ways in which trading partners provide forms or alternative information to forms. [Means for solving the problem]

[0007] To solve the above-mentioned problems, the present invention has the following configuration. That is, the information processing system includes an information processing unit which has the function of acquiring report data relating to a report, the data type of which is not limited to a specific type, and even if the data type is the same, the manner in which information about the report is recorded may differ, identifying the data type of the report data, performing information processing on the report data according to the data type, and generating unified format data that holds item values ​​for one or more specific items relating to the report in a specific data structure. [Effects of the Invention]

[0008] Data corresponding to form-related information (hereinafter referred to as "form data") has various data types due to the diverse methods of providing forms (for example, PDF type for PDF data, CSV type for CSV data, audio type for audio data, text type for text data, etc.), and even if the data type is the same, the manner in which information about the form is recorded varies. According to the present invention configured as described above, regardless of the differences in data types and regardless of the differences in the manner in which information about the form is recorded for the same data type, the information processing unit's automatic information processing means generates unified format data from the form data that holds item values ​​for specific items related to the form in a specific data structure. Here, the unified format data has a unified specific data structure for all forms regardless of the form provision method, and therefore can be managed and processed using a unified method regardless of the form provision method. In other words, the data is easy to manage and easy to process by subsequent systems. According to the present invention, since such data that is easy to manage and easy to process by subsequent systems is generated by the information processing unit's automatic information processing means, the increase in human effort in generating the data is suppressed. In other words, according to the present invention, in light of the fact that there are various ways in which business partners provide forms or information that substitutes for forms, it is possible to generate data that is easy to manage and easy to process by subsequent systems, while suppressing an increase in human labor. [Brief explanation of the drawing]

[0009] [Figure 1] This figure shows an example configuration of a control system according to one embodiment. [Figure 2] This block diagram shows an example of the functional configuration of an information processing server and a user terminal according to one embodiment. [Figure 3] This is a diagram used to explain how to provide purchase orders. [Figure 4] This figure shows an example of a purchase order. [Figure 5] This figure shows an example of the contents of a CSV file. [Figure 6] It is a diagram showing an example of a candidate list screen. [Figure 7] It is a flowchart showing an information processing method of an information processing server. [Figure 8] It is a flowchart showing an information processing method of an information processing server. [Figure 9] It is a diagram showing an example of the content of intermediate data. [Figure 10] It is a diagram showing an example of the content of one record of unified format data. [Figure 11] It is a diagram showing an example of the content of unified format data. [Figure 12] It is a flowchart showing an information processing method of an information processing server. [Figure 13] It is a flowchart showing an information processing method of an information processing server. [Figure 14] It is a diagram showing an example of a response sentence. [Figure 15] It is a flowchart showing an information processing method of an information processing server. [Figure 16] It is a diagram showing an example of a processing result screen. [Figure 17] It is a diagram showing an example of a processing result screen. [Figure 18] It is a diagram showing an example of a processing result screen. [Figure 19] It is a diagram showing an example of a data integration table. [Figure 20] It is a diagram showing a configuration example of an information processing system.

Embodiments for Carrying Out the Invention

[0010] Hereinafter, an embodiment of the present invention will be described based on the drawings. FIG. 1 is a diagram showing a configuration example of a control system 1 (information processing system) according to this embodiment. As shown in FIG. 1, the control system 1 includes an information processing server 2 (information processing system, server, computer) and a user terminal 3 (information processing system, terminal, computer). Both the information processing server 2 and the user terminal 3 can be connected to a network N including the Internet, a telephone network, and other communication networks. A large language model server 4 is connected to the network N.

[0011] The information processing server 2 is a server device connected to the network N. The information processing server 2 has a function of providing various services to be described later to the user. Hereinafter, the service provided by the information processing server 2 is referred to as the "present service". The user is registered for the present service and can log in to the account of the present service using his or her login information (for example, a combination of a login ID and a password). In FIGS. 1 and 2, the information processing server 2 is represented by one block. However, this does not mean that the information processing server 2 is composed of a single device. For example, the information processing server 2 may include virtual servers. Also, for example, the information processing server 2 may be composed of a plurality of devices. In this case, the devices constituting the information processing server 2 may include a web server, a web application server, or a database server. Also, in this case, the information processing server 2 may be composed of a plurality of server devices that are load-balanced by a load balancer. When the information processing server 2 is composed of a plurality of devices, each of the information processing parts of the plurality of devices, or a plurality of combinations, functions as an "information processing part".

[0012] User terminal 3 is a terminal used by the user. In this embodiment, a user refers to anyone who can use the Service. However, a user is assumed to be an entity that engages in some kind of economic activity, conducts some kind of transaction with a business partner (here, a concept that includes customers) in the course of that economic activity, and receives documents (e.g., purchase orders, order forms, or invoices) or information equivalent to documents from the business partner in the course of that transaction (hereinafter referred to as "activity entity"). Activity entities are, for example, organizations or individuals that engage in economic activities related to retail, wholesale, service, food and beverage, construction, or other industries. As mentioned above, a user may be an organization, but for the sake of explanation below, not only users as organizations but also individuals belonging to an organization will be simply referred to as "users". User terminal 3 can be of any type. For example, a desktop computer, a notebook computer, a tablet computer (including so-called smartphones), or a wearable device can function as user terminal 3.

[0013] The large-scale language model server 4 is a server equipped with the large-scale language model 5 (language model). The provider, type, and machine learning method of the large-scale language model 5 are not limited. The large-scale language model server 4 has the function of receiving prompts (question sentences) from external devices, generating answer sentences to the prompts using the functions of the large-scale language model 5, and responding with the answer sentences to the external devices. The information processing server 2 is appropriately granted the authority to use the large-scale language model server 4.

[0014] Figure 2 is a block diagram showing an example of the functional configuration of the information processing server 2 and the user terminal 3. As shown in Figure 2, the information processing server 2 has a functional configuration comprising a server information processing unit 10 (information processing unit), a server communication unit 11, and a server storage unit 12. The user terminal 3 has a functional configuration comprising a terminal information processing unit 13 (information processing unit), a terminal communication unit 14, a terminal display unit 15 (display unit), a terminal input unit 16, and a terminal storage unit 17.

[0015] The server information processing unit 10 comprises a processing unit including a processor and a primary storage device. The server information processing unit 10 performs processing by having the processing unit read programs stored in the storage area of ​​the server storage unit 12 (or other storage areas) into the primary storage device and execute them. In other words, the server information processing unit 10 performs processing through the cooperation of hardware and software. The server communication unit 11 comprises a communication device including a communication control device and a network interface. The server communication unit 11 communicates with external devices via the communication device under the control of the server information processing unit 10. Hereinafter, it will be assumed that communication by the information processing server 2 is appropriately performed by the server communication unit 11, and explanations regarding communication will be omitted. The server storage unit 12 stores data in non-volatile memory. Non-volatile memory is, for example, a hard disk drive (or other magnetic storage device), ROM, or flash memory.

[0016] As shown in Figure 2, the server storage unit 12 stores a user management database 18. The user management database 18 contains records for each user. Each record is associated with a user ID and contains various information about the user. The configuration of the user management database 18 is not limited to the configuration illustrated in this embodiment; for example, it may be composed of a combination of multiple databases.

[0017] The terminal information processing unit 13 comprises a processing unit including a processor and a primary storage device. The terminal information processing unit 13 performs processing by having the processing unit read a program stored in the storage area of ​​the terminal storage unit 17 (or other storage area) into the primary storage device and execute it. In other words, the terminal information processing unit 13 performs processing through the cooperation of hardware and software. The terminal communication unit 14 comprises a communication device including a communication control device and a network interface. The terminal communication unit 14 communicates with external devices via the communication device under the control of the terminal information processing unit 13. Hereinafter, communication by the user terminal 3 will be appropriately performed by the terminal communication unit 14, and explanations regarding communication will be omitted. The terminal display unit 15 comprises a liquid crystal panel, an organic EL panel, or other display device. The terminal display unit 15 displays an image on the display device under the control of the terminal information processing unit 13. The terminal input unit 16 detects input to an input device and outputs the detection result to the terminal information processing unit 13. Input devices include, for example, a keyboard, mouse, touch panel, and camera. The terminal storage unit 17 stores data in non-volatile memory.

[0018] Next, we will explain the operation of control system 1. In the following explanation, we will assume that one of the users is the "User of Interest." This User of Interest is a company that has multiple business partners and receives orders from multiple business partners. Each business partner provides the User of Interest with an order form (document) or information that serves as an alternative to an order form (document) as appropriate. Hereinafter, the ordering party and the information that serves as an alternative to an order form will be referred to as "order form-related information." The method of providing order form-related information varies depending on the business partner. Hereinafter, the method by which a business partner provides order form-related information (document-related information) will be referred to as the "order form provision method." Note that it is common for the order form provision method (document provision method) to vary depending on the business partner.

[0019] Figure 3 is a diagram used to explain how each trading partner provides purchase orders for the user of interest. As shown in Figure 3, trading partners T1 and T2 of the user of interest place orders by faxing them to the user of interest. Figure 4(A) shows purchase order H1 for trading partner T1, and (B) shows purchase order H2 for trading partner T2. As shown in Figures 4(A) and (B), purchase order H1 for trading partner T1 and purchase order H2 for trading partner T2 have different formats, and due to the difference in format, the manner in which information related to the document is recorded differs. Different purchase order formats mean that the layout of the purchase order is different, and when focusing on a single piece of information that has a common meaning, the relative position in which that piece of information is recorded on the purchase order, the way that piece of information is expressed, the relationship between that piece of information and other information, and other manners in which that piece of information is recorded may differ. Thus, it is common for purchase order formats to differ depending on the trading partner (= the manner in which information related to the document is recorded differs). However, generally speaking, for a given client, the order form format will not change, and the same order form format will continue to be used. This also applies to the data structure of the CSV data related to the order forms.

[0020] Let's focus on purchase order H1 in Figure 4(A). Purchase order H1 records item values ​​for various items. Hereafter, the items of the purchase order will be referred to as "purchase order items". Purchase order H1 has a detail area 20. In the detail area 20, detail lines 21 are recorded for each detail item. The detail items are typically goods ordered by a trading partner (however, they are not limited to goods). As shown in Figure 4(A), purchase order H1 includes the following purchase order items for each detail line: product name <purchase order item>, quantity <purchase order item>, unit <purchase order item>, unit price <purchase order item>, subtotal <purchase order item>, and remarks <purchase order item>. In addition, purchase order H1 includes the following purchase order items, excluding the line items: Order date and time <Purchase order item>, Order ID <Purchase order item>, Desired delivery date <Purchase order item>, Delivery address <Purchase order item>, Delivery address phone number <Purchase order item>, Ordering party <Purchase order item>, Ordering party address <Purchase order item>, Ordering party phone number <Purchase order item>, Subtotal (excluding tax) <Purchase order item>, and Total (including tax) <Purchase order item>.

[0021] When an order is placed by fax, the paper order form printed from the fax machine is scanned by a scanner, generating a PDF version of the order form, which is then stored in the terminal storage unit 17 of the user terminal 3. Hereinafter, the PDF version of the order form will be referred to as "order form PDF data." The data type of the order form PDF data (form PDF data) is the PDF type corresponding to PDF data.

[0022] Referring to Figure 3, trading partners T3 and T4 place orders by sending CSV data, in which the information to be included in the purchase order is recorded according to a specific data structure, to the user of interest. The method of sending the CSV data can be anything; for example, it may be attached to an email, or it may be uploaded to the user of interest's system by a predetermined means. Hereinafter, the CSV data related to the purchase order will be referred to as "purchase order CSV data". Purchase order CSV data (document CSV data) is information that replaces the purchase order (document), and its data type is the CSV type corresponding to CSV data. When the trading partner sends the purchase order CSV data, the purchase order CSV data is stored in the terminal storage unit 17 of the user terminal 3 of the user of interest. Figure 5(A) shows the contents of purchase order CSV data C1 related to trading partner T3, and (B) shows the contents of purchase order CSV data C2 related to trading partner T4. As shown in Figures 5(A) and (B), the purchase order CSV data C1 for customer T3 and the purchase order CSV data C2 for customer T4 have different data structures, and due to these differences in data structure, the way in which information about the documents is recorded differs. It is common for the data structure of CSV data related to purchase orders to differ depending on the customer (i.e., the way in which information about the documents is recorded differs). Hereafter, the items in the purchase order CSV data will be referred to as "CSV items".

[0023] Referring to Figure 3, customer T5 places an order by leaving a voice message (a message spoken by the caller) using the answering machine's recording function. The voice message is information that replaces the purchase order and contains information related to the form. When an order is placed via voice message, the system of the user in question, or other systems (which may include human processing), stores the voice message in a predetermined file format (e.g., MP3 or WAVE) as audio data in the terminal storage unit 17 of the user terminal 3. Hereinafter, the audio data in which the voice message is recorded will be referred to as "purchase order voice data". The data type of the purchase order voice data is the voice type corresponding to the audio data. There is no format for the voice message, and the content can be freely determined by the caller. Therefore, the manner in which information related to the form is recorded differs for each voice message. For example, suppose the user in question is a company that delivers beverages. In this case, one example of the content of the spoken message is, "This is T5 Corporation. Tomorrow, please deliver 10 cans of XX 350ml canned beer and 10 cans of XX 350ml canned juice. The delivery address is the T5 office. The address of the T5 office is Saitama Prefecture..." Another example is, "Thank you for your continued support. This is T5 Corporation. Tomorrow, as usual, please deliver 10 cans of XX 350ml canned juice and 10 cans of XX 350ml canned beer."

[0024] Referring to Figure 3, customer T6 places an order by sending an input message (a message entered by the inputter) related to the order to the designated user using a chat application. The input message is information that replaces the purchase order and contains information related to the form. A chat application is an application that provides a platform for exchanging messages in a chat format. When an order is placed via the chat application, the content of the input message is recorded as text data stored in the terminal storage unit 17 of the user terminal 3 through processing by the designated user's system or other systems (this processing may include human processing). Hereinafter, the text data in which the input message is recorded will be referred to as "purchase order text data". The data type of the purchase order text data is the text type corresponding to the text data. There is no format for the input message recorded in the purchase order text data, and therefore the manner in which information related to the form is recorded differs for each input message. An example of the content of an input message can be found by referring to the example of the content of a spoken message described above.

[0025] As described above, based on orders from business partners, the system stores data corresponding to purchase order-related information for the user of interest, including purchase order PDF data, purchase order CSV data, purchase order audio data, or purchase order text data. Hereinafter, these data will be collectively referred to as "initial purchase order-related data." Initial purchase order-related data corresponds to "document data." Due to the diverse methods of providing documents, the initial purchase order-related data has various data types (in this embodiment, PDF data type, CSV data type, audio data type, and text data type), and even if the data type is the same, the manner in which information about the document is recorded varies.

[0026] Now, in the following explanation of the operation of control system 1, it is assumed that the following initial purchase order corresponding data is stored in the terminal storage unit 17 of the user terminal 3 of the user of interest. • Purchase order PDF data D1 related to supplier T1 • Purchase order PDF data D2 related to supplier T2 • Purchase order CSV data D3 related to supplier T3 • Purchase order CSV data D4 related to supplier T4 • Audio data D5 of the purchase order related to supplier T5 • Purchase order text data D6 related to supplier T6 The user in question intends to use this service to generate unified format data 22 (described later) for data D1 to D6.

[0027] The user in question launches the browser on user terminal 3 and instructs it to access a predetermined URL on information processing server 2. This predetermined URL is the URL of the main screen of the service. In response to this instruction, terminal information processing unit 13 accesses the predetermined URL using the browser's functions. When accessing the URL, the user in question logs in to the service using an appropriate method. When the server information processing unit 10 of information processing server 2 receives the predetermined URL, it performs the necessary authentication and responds to terminal information processing unit 13 with display data (HTML file and data attached to the HTML file) to display the main screen. The terminal information processing unit 13 receives this display data and displays the main screen on terminal display unit 15 based on the display data using the browser's functions.

[0028] In the following, it is assumed that HTTP communication with the server information processing unit 10 naturally occurs when the terminal information processing unit 13 displays various screens, and therefore, a detailed explanation of such HTTP communication will be omitted. Furthermore, the detection of operations on various screens displayed based on the display data sent from the server information processing unit 10 to the terminal information processing unit 13, the acquisition of information entered into various screens, the dynamic modification of various screens, and other screen-related processing will be appropriately executed by the server information processing unit 10 and the terminal information processing unit 13 using CGI, JavaScript (registered trademark), Ajax, DHTML, and other technologies, and a detailed explanation will not be provided. In the following, the display of various screens on the terminal display unit 15 by the terminal information processing unit 13 using the browser's functions will be expressed as "the terminal information processing unit 13 works with the server information processing unit 10 to display the ○○ screen on the terminal display unit 15" or simply "the terminal information processing unit 13 displays the ○○ screen."

[0029] After the main screen is displayed, the user in question performs a predetermined operation on the main screen to instruct the display of the upload screen (not shown). The upload screen is for uploading desired data from the data stored in the terminal storage unit 17 to the information processing server 2. It is equipped with a user interface for specifying the data (files) to be uploaded and instructing the upload of the specified data. In this example, the user in question uses the upload screen to upload each of the data D1 to D6 to the information processing server 2.

[0030] When data D1 to D are uploaded, the server information processing unit 10 assigns a unique data ID (identification information) to each of the data D1 to D6. Next, the server information processing unit 10 registers data D1 to D6 in the user management database 18, associating each data ID with the record corresponding to the user of interest. Note that the method of sending data from the user to the information processing server 2 is not limited to uploading. It may also be done by attaching it to an email or by other methods.

[0031] After uploading data D1 to D6, the user of interest performs a predetermined operation to instruct the display of the candidate list screen 23. In response to this instruction, the terminal information processing unit 13 works in cooperation with the server information processing unit 10 to display the candidate list screen 23 on the terminal display unit 15. When displaying this screen, the server information processing unit 10 sends the information registered in the record corresponding to the user of interest in the user management database 18, and other necessary information to the terminal information processing unit 13. Figure 6 shows an example of the candidate list screen 23. The candidate list screen 23 is a screen for selecting the target to be converted to unified format data 22 (described later) (hereinafter referred to as "conversion target"). The candidate list screen 23 displays a list of information regarding the data uploaded for the corresponding user that has not been converted to unified format data 22 (hereinafter referred to as "candidate data"). As shown in Figure 6, the candidate list screen 23 is provided with a block-shaped candidate data block 24 for each candidate data. In the candidate list screen 23 of Figure 6, the candidate data blocks 24 related to data D1 to D6 are displayed. The candidate data block 24 displays a thumbnail image of the preview image of the corresponding candidate data. However, for candidate data of the audio type, information indicating that the data type is audio is displayed instead of a thumbnail image. The candidate data block 24 also displays information indicating the upload date and time, the file name, the data ID, and the data type for the corresponding candidate data. The server information processing unit 10 identifies the data type based on the extension of the data (file). The thumbnail image is selectable, and when a thumbnail image is selected, the terminal information processing unit 13 displays a preview of the candidate data using the viewer corresponding to the data type.

[0032] On the candidate list screen 23, each candidate data block 24 is provided with a check button 25 for selecting candidate data. The screen also has a confirm button 26 for confirming the selection. In this example, the user checks the corresponding check button for each of the data D1 to D6 and selects the confirm button 26. In other words, the user selects data D1 to D6 as the data to be converted. Hereinafter, the candidate data selected by the user on the candidate list screen 23 (= the initial purchase order corresponding data specified by the user as the data to be converted) will be referred to as "selected data". When the confirm button 26 is selected, the terminal information processing unit 13 sends notification information to the server information processing unit 10 that includes at least the data ID of each of the selected data (each of the data D1 to D6 in this example) and the user ID of the user of interest.

[0033] Upon receiving the notification information, the server information processing unit 10 executes the conversion process. The conversion process is described in detail below. In Figure 7, the symbol FA is a flowchart showing the information processing method by the server information processing unit 10 when executing the conversion process. As shown in Figure 7, the server information processing unit 10 recognizes each of the data IDs of the selected data included in the notification information received from the terminal information processing unit 13 (step SA1). Next, the server information processing unit 10 refers to the user management database 18 and obtains each of the initial purchase order corresponding data (=each of the selected data) that corresponds to each of the data IDs recognized in step SA1 (step SA2). In this example, in step SA2, the server information processing unit 10 obtains data D1 to D6. Thus, the server information processing unit 10 has a function to "acquire report data, which is data relating to a report, the data type is not limited to a specific type, and even if the data type is the same, the manner in which information relating to the report is recorded may differ."

[0034] Next, the server information processing unit 10 determines whether or not there is any unprocessed data among the acquired initial purchase order corresponding data (step SA3). If there is no unprocessed data (step SA3: NO), the server information processing unit 10 terminates the conversion process. If there is unprocessed data (step SA3: YES), the server information processing unit 10 identifies one data item from the unprocessed data to be processed (step SA4). Hereinafter, the data identified in step SA4 will be referred to as the "data to be processed". Next, the server information processing unit 10 identifies the data type of the data to be processed (step SA5). Then, the server information processing unit 10 executes the following processes according to the identified data type. • Data type: PDF type → PDF type conversion process (Step SA6) • Data type: CSV type → CSV type conversion process (Step SA7) • Data type: Voice type → Voice type conversion process (Step SA8) • Data type: Text type → Text type conversion process (Step SA9) Each process in steps SA6 to SA9 corresponds to "information processing according to the data type." After each process in steps SA6 to SA9 is completed, the server information processing unit 10 returns the processing procedure to step SA3. The processes in steps SA6 to SA9 are described in detail below.

[0035] ● PDF type conversion process First, let's explain the PDF type conversion process in step SA6. The symbol FB in Figure 8 is a flowchart showing the information processing method by the server information processing unit 10 when executing the PDF type conversion process. As shown in flowchart FB, the server information processing unit 10 executes an intermediate data generation process (step SB1). The intermediate data generation process processes the data to be processed.<PDFタイプ> This process converts the data into intermediate data 28 (described later). The intermediate data generation process is described in detail below.

[0036] Here, the server memory unit 12 stores the AI-OCR model 29. AI-OCR is a machine learning model. "PDF-type form data" is input to the AI-OCR model 29. The AI-OCR model 29 performs character recognition on the input form data using optical character recognition technology and has the function of analyzing the results of character recognition to extract combinations of purchase order items and item values ​​recorded in the form. Hereinafter, the combinations of purchase order items and item values ​​extracted by the AI-OCR model 29 will be called "form-related combinations". Furthermore, the AI-OCR model 29 has the function of deriving an accuracy score that indicates the accuracy of the item value (accuracy of character recognition) for each extracted form-related combination. In this embodiment, the accuracy score takes a value between 0 and 100pt (pt is a unit assigned for convenience), and the larger the value, the higher the accuracy. Furthermore, the AI-OCR model 29 has the function of outputting intermediate data 28 in JSON format in which the extracted form-related combinations and accuracy are recorded in a structured state.

[0037] In the intermediate data generation process, the server information processing unit 10 inputs the data to be processed (PDF type) into the AI-OCR model 29 and obtains the intermediate data 28 output by the AI-OCR model 29. Specifically, the server information processing unit 10 has a function that "performs character recognition on the PDF type report data using optical character recognition technology based on the functions of a machine learning model, analyzes the results of the character recognition to extract combinations of items and item values ​​recorded in the report data, and generates intermediate data in which the extracted combinations are recorded in a structured state." The server information processing unit 10 also has a function that "derives the accuracy of the item values ​​for each extracted combination from the PDF type report data using the functions of the model, and generates intermediate data in which the derived accuracy is recorded."

[0038] For example, suppose the initial purchase order data related to purchase order H1 in Figure 4(A) is the data to be processed. In this case, the server information processing unit 10 processes the data to be processed using the functions of the AI-OCR model 29, and generates and outputs intermediate data 28 in JSON format as shown in Figure 9. Part B1 of the intermediate data 28 in Figure 9 shows that the item value of purchase order date and time <purchase order item> is "2024-11-1-13:00" and its accuracy score is 89pt. Also, part B2 of the intermediate data 28 in Figure 9 shows that there are two detail lines 21 in purchase order H1, and the item value and accuracy score are recorded for each detail line 21. The above is a detailed explanation of the intermediate data generation process.

[0039] After step SB1, the server information processing unit 10 executes the PDF-compatible unified format data generation process (step SB2). The PDF-compatible unified format data generation process generates unified format data 22 based on the intermediate data 28. The PDF-compatible unified format data generation process is described in detail below.

[0040] First, let's explain the unified format data 22. Unified format data 22 is structured data that holds item values ​​for "specific items related to a form" in a specific data structure. In this embodiment, the unified format data 22 is configured as a table that holds a record "for each detail line 21" of the form. Figure 10 shows the contents of one record of the unified format data 22 in this example. Hereinafter, the items that make up the unified format data will be collectively referred to as "unified items". Unified items correspond to "specific items". As shown in Figure 10, the unified format data 22 in this example includes customer name <unified item>, delivery destination name <unified item>, delivery destination address <unified item>, desired delivery date <unified item>, overall subtotal <unified item>, overall total <unified item>, product number <unified item>, product name <unified item>, quantity <unified item>, unit price <unified item>, subtotal <unified item>, and remarks <unified item>. The "Customer Name" field indicates the name of the ordering party, the "Recipient Name" field indicates the name of the recipient (delivery destination), and the "Recipient Address" field indicates the address of the recipient. The "Desired Delivery Date" field indicates the desired delivery date, the "Overall Subtotal" field indicates the overall subtotal (calculated excluding tax) in the report, and the "Overall Total" field indicates the overall total (calculated including tax) in the report. The "Product Number" field, "Product Name" field, "Quantity" field, "Unit Price" field, "Line Subtotal" field, and "Remarks" field are all fields related to line 21.

[0041] Furthermore, the data structure of the standardized format data (especially the types and order of standardized items) should vary depending on the type of report. Also, the data structure of the standardized format data may differ for each user and may be configured by the user.

[0042] In step SB2, the server information processing unit 10 first refers to the intermediate data 28 and identifies the ordering party related to the data to be processed (PDF type). For example, if the intermediate data 28 is the data shown in Figure 9, the server information processing unit 10 recognizes that the ordering party is "T1 Corporation" based on the item value of ordering party <order form item>. Note that the method for identifying the ordering party is not limited to the method exemplified in this embodiment. For example, the server information processing unit 10 and the terminal information processing unit 13 may be configured to allow the user to specify the ordering party via a predetermined user interface and recognize it. Alternatively, for example, the server information processing unit 10 may be configured to refer to predetermined data when the initial order form data is uploaded, or after the upload, and register it in the database. The same points apply to the CSV-compatible unified format data generation process described later.

[0043] Next, the server information processing unit 10 refers to the PDF-related correspondence table corresponding to the identified ordering party. The PDF-related correspondence table for a particular ordering party is data that records the correspondence between the order form items of the order form provided by that ordering party and the unified items of the unified format data 22. For example, the desired delivery date <order form item> in order form H1 in Figure 4 (A) (= desired delivery date <order form item> in Figure 9) and the desired delivery date <unified item> in the unified format data 22 shown in Figure 10 are identical items. Therefore, the PDF-related correspondence table corresponding to the ordering party (T1 Corporation) related to order form H1 records the desired delivery date <unified item> and the desired delivery date <order form item> in association. Note that the identity of the order form items and the unified items means that the item values ​​of each item have the same meaning. The PDF-related correspondence table is prepared in advance and stored in a predetermined database of the server storage unit 12 in association with information that identifies the ordering party. The server information processing unit 10 and the terminal information processing unit 13 provide the user with a predetermined user interface that enables the creation and editing of PDF-related correspondence tables. The user can create and edit the said tables through this interface.

[0044] After referring to the PDF-related correspondence table, the server information processing unit 10 generates unified format data 22 from the intermediate data 28, taking into account the correspondence between purchase order items and unified items recorded in the PDF-related correspondence table. That is, the server information processing unit 10 generates the unified format data 22 by setting the item value of the purchase order item that is identical to the unified item for each unified item. If the item value of the purchase order item is a null value, the server information processing unit 10 sets the null value as is for the unified item. It is also possible that there are no purchase order items in the intermediate data 28 that are identical to the unified item. For example, referring to Figures 9 and 10, the unified format data 22 in this example includes "product number" as a unified item, but the intermediate data 28 in Figure 9 does not include "product number" as a purchase order item. In this case, the server information processing unit 10 sets a null value for the unified item. Figure 11 shows an example of unified format data generated based on the intermediate data illustrated in Figure 9.

[0045] Referring to the flowchart FB in Figure 8, after step SB2, the server information processing unit 10 registers the intermediate data 28 generated in step SB1 and the unified format data 22 generated in step SB2 in the record corresponding to the user of interest in the user management database 18, associating them with the data ID of the data to be processed. This completes the PDF type conversion process. In this example, the PDF type conversion process is performed on the purchase order PDF data D1 related to business partner T1 and the purchase order PDF data D2 related to business partner T2, and unified format data 22 corresponding to each of these data is generated. The unified format data 22 corresponding to the purchase order PDF data D1 is called unified format data X1, and the unified format data 22 corresponding to the purchase order PDF data D2 is called unified format data X2.

[0046] ●CSV type conversion process Next, the CSV type conversion process in step SA7 will be described. The reference numeral FC in FIG. 12 is a flowchart showing the information processing method by the server information processing unit 10 when executing the CSV type conversion process. As shown in the flowchart FC, the server information processing unit 10 executes a CSV-compatible unified format data generation process (step SC1). This will be described in detail below. First, the server information processing unit 10 refers to the data to be processed (CSV type) and identifies the orderer. For example, when the data to be processed is the data shown in (A) of FIG. 5, the server information processing unit 10 recognizes that the orderer is "T3 Shokai" based on the item value of the <CSV item> of the orderer. However, as described above, the method of identifying the orderer is not limited to the exemplified method.

[0047] Next, the server information processing unit 10 refers to the CSV-related correspondence table corresponding to the identified orderer. The CSV-related correspondence table regarding one orderer is data in which the correspondence between the CSV items of the order form CSV data provided by the one orderer and the unified items of the unified format data 22 is recorded. Next, the server information processing unit 10 generates the unified format data 22 from the data to be processed (CSV type) based on the correspondence between the CSV items and the unified items recorded in the CSV-related correspondence table. That is, the server information processing unit 10 generates the unified format data 22 by setting the item values of the CSV items having the same identity as the unified item for each of the unified items. The above is the detail of step SC1.

[0048] After step SC1, the server information processing unit 10 registers the unified format data 22 generated in step SC1 in the record corresponding to the user of interest in the user management database 18, associating it with the data ID of the data to be processed. This completes the CSV type conversion process. In this example, the CSV type conversion process is performed on the purchase order CSV data D3 related to business partner T3 and the purchase order CSV data D4 related to business partner T4, and unified format data 22 corresponding to each of these data is generated. The unified format data 22 corresponding to the purchase order CSV data D3 is called unified format data X3, and the unified format data 22 corresponding to the purchase order CSV data D3 is called unified format data X4.

[0049] ● Voice type conversion processing Next, the speech type conversion process in step SA8 will be described. The symbol FD in Figure 13 is a flowchart showing the information processing method by the server information processing unit 10 when executing the speech type conversion process. As shown in flowchart FD, the server information processing unit 10 performs speech recognition processing (step SD1). More specifically, the server information processing unit 10 performs speech recognition on the data to be processed for speech type using speech recognition technology and generates speech message data in which the spoken message is recorded as text.

[0050] Next, the server information processing unit 10 executes the prompt generation process (step SD2). More specifically, the server information processing unit 10 generates a speech-related prompt, which is a prompt for the large-scale language model 5. The speech-related prompt includes the speech message recorded in the speech message data. The speech-related prompt also includes the following requests. • Extract the item values ​​of standardized items included in spoken messages. • Derive an accuracy score that indicates the accuracy of the extracted item values. Accuracy refers to the degree to which the extracted item value for a particular item is accurate as an item value for that item. Below is an example of a voice-related prompt. However, the following example is intended to provide a clear overview of voice-related prompts and disregards its validity as a prompt.

[0051] ***Start of voice-related prompts*** #Prerequisites You work for a company that supplies beverages and are responsible for processing purchase orders. You received the following voicemail message requesting an order using the answering machine's recording function. From the following voicemail message, identify all the items to be delivered, and for each item, extract the item value for each item listed in the following item list. Also, derive an accuracy score for each extracted item value. Accuracy refers to the degree to which the extracted item value is accurate as the item value for that item. The accuracy score should be a number in the range of "0 to 100". Note that the item list may contain items whose item values ​​are not included in the voicemail message. <Speech message> This is T3 Co., Ltd. Tomorrow, please deliver 10 cans of [product name] beer (350ml each) and 10 cans of [product name] juice (350ml each). The delivery address is the T3 office. The T3 office address is in Saitama Prefecture... <Item List> • Customer name ·Delivery destination name • Delivery address • Desired delivery date ·Overall subtotal • Total Product number ·Product name ·quantity ·unit price ·subtotal ·remarks #Result output • Please record the extraction results for each product in your response. • For each item in the item list, please record the extracted item value and the derived accuracy score in the above extraction results, following the format exemplified below. For items from which no item value could be extracted, please set both the item value and the accuracy score to "None". • If the spoken message is as shown in <Example Spoken Message>, the response will be as shown in <Example Response>. Please refer to this when generating your response. <Format> • Item: Item Value: Accuracy Score <Example of spoken message> Hello. This is ○○ Store... <Example of a response> Product 1 • Customer name: ○○ Store: 98 • Delivery destination name: XX Company: 88 • Delivery address: None • Desired delivery date: Not specified ... Product 2 • Customer name: ○○ Store: 98 • Delivery destination name: XX Company: 88 • Delivery address: None • Desired delivery date: Not specified ... ***End of voice-related prompts***

[0052] The server information processing unit 10 according to this embodiment generates voice-related prompts based on the functions of a program stored in the server storage unit 12. This program has the function of generating voice-related prompts by embedding the speech message recorded in the speech message data into a pre-prepared voice-related prompt format and performing necessary modifications. However, the method by which the server information processing unit 10 generates voice-related prompts is not limited to the method illustrated in this embodiment. For example, the server information processing unit 10 may request the large-scale language model 5 to generate a voice-related prompt and obtain the voice-related prompt as a response to that request. The above points regarding the generation of voice-related prompts also apply to the acquisition of other prompts.

[0053] Now, referring to the flowchart FD in Figure 13, after generating the voice-related prompt, the server information processing unit 10 sends the voice-related prompt to the large-scale language model server 4 (step SD3). The large-scale language model server 4 receives the voice-related prompt and generates a response sentence to the voice-related prompt using the functions of the large-scale language model 5. Figure 14 shows an example of a response sentence obtained as a response to the voice-related prompt described above. As shown in Figure 14, the response sentence records the item values ​​of the extracted unified items and the accuracy score for each item value of the extracted unified items. The large-scale language model server 4 sends the response sentence to the server information processing unit 10 (responds). The server information processing unit 10 receives and acquires the response sentence (step SD4).

[0054] After step SD4, the server information processing unit 10 executes the voice-related unified format data generation process (step SD5). More specifically, the server information processing unit 10 generates unified format data 22 based on the response text received from the large-scale language model server 4. The response text contains the necessary information to generate the unified format data 22, as it records item values ​​for each unified item (however, there are cases where the value indicates that the item does not exist in the spoken message) for each product (i.e., each detail line 21).

[0055] Next, the server information processing unit 10 registers the unified format data 22 generated in step SD5 and the response text obtained in step SD4 in the record corresponding to the user of interest in the user management database 18, associating it with the data ID of the data to be processed (step SD6). The above is a detailed description of the voice type conversion process. In this example, in this conversion process, the voice type conversion process is performed on the purchase order voice data D5 related to business partner T5, and unified format data 22 corresponding to this data is generated. The unified format data 22 corresponding to the purchase order voice data D5 is called unified format data X5.

[0056] ● Text type conversion process Next, the text type conversion process in step SA9 will be described. The symbol FE in Figure 15 is a flowchart showing the information processing method by the server information processing unit 10 when executing the text type conversion process. As shown in flowchart FE, the server information processing unit 10 executes a prompt generation process (step SE1). In step SE1, the server information processing unit 10 generates a text-related prompt based on the data to be processed (text type). The content of the text-related prompt is the same as the content of the voice-related prompt. That is, the text-related prompt includes the input message recorded in the data to be processed (text type). The text-related prompt also includes the following requests. • Extract the field values ​​of standardized items included in the input message. • Derive an accuracy score that indicates the accuracy of the extracted item values.

[0057] Next, the server information processing unit 10 sends a text-related prompt to the large-scale language model server 4 (step SE2). The large-scale language model server 4 receives the text-related prompt and generates a response sentence to the text-related prompt using the functions of the large-scale language model 5. The large-scale language model server 4 sends the response sentence to the server information processing unit 10 (responds). The server information processing unit 10 receives and acquires the response sentence (step SE3).

[0058] Next, the server information processing unit 10 executes a text-compatible unified format data generation process (step SE4). More specifically, the server information processing unit 10 generates unified format data 22 based on the response text received from the large-scale language model server 4. Next, the server information processing unit 10 registers the unified format data 22 generated in step SE4 and the response text obtained in step SE3 in the record corresponding to the user of interest in the user management database 18, associating it with the data ID of the data to be processed (step SE5). The above is a detailed description of the text type conversion process. In this example, in this conversion process, the text type conversion process is executed on the purchase order text data D6 related to business partner T6, and unified format data 22 corresponding to this data is generated. The unified format data 22 corresponding to the purchase order text data D6 is called unified format data X6.

[0059] The above is a detailed explanation of the conversion process. In this example, once the conversion process is complete, unified format data X1 to X6 are generated for each of the data D1 to D6, and these are registered in the records corresponding to the featured users in the user management database 18.

[0060] Once the conversion process is complete, the terminal information processing unit 13 works in cooperation with the server information processing unit 10 to display the processing result screen 31 on the terminal display unit 15. In displaying the processing result screen 31, the server information processing unit 10 appropriately provides the terminal information processing unit 13 with predetermined information (including at least information based on the unified format data 22 of the selected data, information based on the intermediate data 28, and information based on the response text) and other information registered in the record corresponding to the user of interest in the user management database 18. Figures 16, 17, and 18 show an example of the processing result screen 31. At the top of the processing result screen 31, a data selection area 32 is provided for selecting one or more desired selected data from the selected data. In the data selection area 32, the data ID of each selected data is shown, and the desired selected data (= initial purchase order corresponding data for which the unified format data 22 was generated by the conversion process) can be selected by checking the checkboxes. In the examples shown in Figures 16, 17, and 18, the data selection area 32 selects the following: purchase order PDF data D1 (data ID: ID01), purchase order CSV data D3 (data ID: ID03), and purchase order audio data D5 (data ID: ID05). Hereafter, the selected data selected in the data selection area 32 will be referred to as "specific data".

[0061] A preview area 33 is provided on the left side of the processing result screen 31, and a preview image 34 of the specific data specified by the user is displayed in this preview area 33. The specific data can be specified (switched) using a tab 35 that displays the data ID. Figure 16 shows the processing result screen 31 when PDF type purchase order PDF data D1 is specified, Figure 17 shows the processing result screen 31 when CSV type purchase order CSV data D3 is specified, and Figure 18 shows the processing result screen 31 when voice type purchase order voice data D5 is specified. As shown in Figure 18, if the data type of the specified specific data is voice type, the server information processing unit 10 displays a preview image 34 of the spoken message data (data generated by speech recognition processing).

[0062] A data display area 36 is provided on the right side of the processing result screen 31. In this data display area 36, ​​the unified format data 22 of each specific data is displayed together in a single common table, the data integration table 37. Because the unified format data 22 has a specific data structure, it is possible to display them together in a single table regardless of the data type of the initial purchase order corresponding data that formed the basis of the unified format data. In Figure 16, a portion of the data integration table 37 is displayed, but the desired portion of the data integration table 37 can be viewed by scrolling with the scroll bar.

[0063] Figure 19 is a diagram illustrating the data integration table 37 from Figures 16, 17, and 18. As mentioned above, the data integration table 37 in Figure 19 displays the unified format data X1 for the purchase order PDF data D1, the unified format data X3 for the purchase order CSV data D3, and the unified format data X5 for the purchase order audio data D5. As shown in Figure 19, the data integration table 37 is a matrix table in which a record is created for each record of the unified format data. In the data integration table 37, the data ID is clearly indicated for each specific data, and the unified format data 22 is displayed in association with the data ID. In the data integration table 37, a cell 38 (square) is provided for each item value, and the item value is shown in each square. If the item value is a null value, the cell is left blank, clearly indicating that the item value is null. Each cell is selectable, and when a cell is selected, the item value of that cell becomes editable. Editing values ​​includes setting new values ​​for null values. The server information processing unit 10 works in cooperation with the terminal information processing unit 13 to update the contents of the appropriate unified format data 22 registered in the user management database 18 when an item value of a certain unified format data 22 is edited.

[0064] In the data integration table 37, the background of cell 38 is white for unified item values ​​with an accuracy score of 70pt or higher, while the background of cell 38 is yellow for unified item values ​​with an accuracy score of 70pt or lower. The "yellow background of cell 38" corresponds to an accuracy-based warning. In other words, in the data integration table 37, the unified format data 22 is displayed on the terminal display unit 15 with an accuracy-based warning indicated. The user can intuitively and accurately recognize item values ​​with low accuracy based on the background color of cell 38, and can check the preview image 34 and correct the value as needed. As described above, the terminal information processing unit 13 has a function to "display the unified format data on the display unit with the item values ​​of specific items shown in an editable format along with a preview image of the report data for each data type, and with the accuracy-based warning indicated for the item values ​​of those specific items."

[0065] Referring to Figure 16, a CSV generation button 40 is provided below the data integration table 37 in the data display area 36 to instruct the generation of CSV data for the data integration table 37. When the CSV generation button 40 is selected, the server information processing unit 10 works in cooperation with the terminal information processing unit 13 to generate a single CSV data file (structured data) containing all the unified format data 22 of the data integration table 37, and the terminal information processing unit 13 downloads the generated CSV data to the terminal storage unit 17. Note that the process by which the server information processing unit 10 generates structured data and allows it to be downloaded to the user terminal 3 corresponds to the "process of outputting structured data". To the right of the CSV generation button 40, a JSON generation button 41 is provided to instruct the generation of JSON data for the data integration table 37. When the JSON generation button 41 is selected, the server information processing unit 10 works in cooperation with the terminal information processing unit 13 to generate a single JSON data file containing all the unified format data 22 of the data integration table 37, and the terminal information processing unit 13 downloads the generated JSON data to the terminal storage unit 17.

[0066] In addition to the functions related to the processing result screen 31, the terminal information processing unit 13 has the following functions. Specifically, the terminal information processing unit 13 works in cooperation with the server information processing unit 10 to provide the user with a user interface that allows the user to select one or more data from all initial purchase order corresponding data for which unified format data 22 has already been generated. When one or more initial purchase order corresponding data are selected, the terminal information processing unit 13 displays a screen on the left showing a preview image 34 of the specified initial purchase order corresponding data, and on the right showing a data integration table 37, a CSV generation button 40, and a JSON generation button 41. By selecting the CSV generation button 40 or the JSON generation button 41, the user can obtain CSV data or JSON data containing the unified format data 22 generated for the specified initial purchase order corresponding data.

[0067] Thus, the server information processing unit 10 has a function to display a single table on the display unit that shows all of the unified format data for one or more of the aforementioned report data specified by the user together. The server information processing unit 10 also has a function to display a preview image of the specified report data on the display unit along with the table, which is configured to allow editing of arbitrary item values. The server information processing unit 10 also has a function to output the table as predetermined structured data in accordance with the user's instructions. As a result, the user can refer to one or more desired initial purchase order corresponding data as a single table and obtain it as a single structured data. For example, a user can group initial purchase order corresponding data from their own perspective, refer to a table that contains only the unified format data 22 of the initial purchase order corresponding data belonging to a specific group, and easily obtain the structured data corresponding to that table.

[0068] As described above, the information processing system according to this embodiment includes an information processing unit that has the function of acquiring report data relating to a report, wherein the data type is not limited to a specific type, and even if the data type is the same, the manner in which information about the report is recorded may differ, identifying the data type of the report data, performing information processing on the report data according to the data type, and generating unified format data that holds item values ​​for one or more specific items relating to the report in a specific data structure.

[0069] This configuration provides the following benefits. Specifically, due to the diverse methods of providing reports, the report data has various data types, and even with the same data type, the way in which information about the report is recorded varies. With this configuration, regardless of the differences in data types, and regardless of the differences in how information about the report is recorded within the same data type, the information processing unit's automatic information processing means generates unified format data from the report data, holding item values ​​for specific items related to the report in a specific data structure. Here, the unified format data has a unified specific data structure for all reports, regardless of the method of providing the report, and therefore can be managed and processed using a unified method regardless of the method of providing the report. In other words, the data is easy to manage and easy to process by subsequent systems. With this configuration, since such data that is easy to manage and easy to process by subsequent systems is generated by the information processing unit's automatic information processing means, the increase in human effort required for data generation is suppressed. In other words, with the above configuration, it is possible to generate data that is easy to manage and easy to process by subsequent systems, while suppressing an increase in manual labor, reflecting the characteristic that trading partners have various ways of providing information in the form or in lieu of a form.

[0070] <Variation> Although one embodiment of the present invention has been described above, the above embodiment is merely an example of how the present invention can be implemented, and the technical scope of the present invention should not be interpreted as being limited by it. That is, the present invention can be implemented in various forms without departing from its gist or its main features. The following are modifications of the above embodiment. The following modifications may be applied in combination where possible.

[0071] Regarding the data integration table 37, the method for reflecting the accuracy score is not limited to the method exemplified in the above embodiment. For example, the terminal information processing unit 13 may be configured to display information indicating the accuracy score in association with cell 38, information indicating the level derived from the accuracy score, and other information reflecting accuracy. In the above embodiment, in the PDF-compatible unified format data generation process, the server information processing unit 10 generated unified format data 22 based on the intermediate data 28 using the PDF-related correspondence table. However, the method for generating unified format data 22 from the intermediate data 28 is not limited to the method exemplified in the above embodiment. For example, the server information processing unit 10 may analyze the intermediate data 28, estimate the correspondence between the order form items recorded in the intermediate data 28 and the unified items, and use the estimation results to generate unified format data 22 from the intermediate data 28. In this configuration, the estimation can be performed by the functions of a machine learning model. The same points apply to the CSV-compatible unified format data generation process.

[0072] In the above embodiment, several specific examples of methods for providing purchase orders were given, but the method of providing purchase orders is not limited to the examples given. For example, the business partner may send PDF data via email. Alternatively, the business partner may send an email containing information that serves as a substitute for a purchase order in the body of the email. The above points also apply when the document is not a purchase order. In the above embodiment, the document was a purchase order, but of course, the document is not limited to a purchase order. Furthermore, it goes without saying that the information that can substitute for the document is not limited to the information exemplified. For example, when taking an order by telephone, a memo written by the person answering the phone could serve as information that substitutes for the document. In the above embodiment, the server information processing unit 10 was configured to generate intermediate data using the AI-OCR model 29 stored in the server storage unit 12. In this regard, the server information processing unit 10 may also be configured to request the generation of intermediate data from an external device equipped with the AI-OCR model 29. The unified format data 22 can be any data that holds item values ​​for specific items related to a form in a specific data structure. For example, the unified format data can be JSON data or a file in a specified spreadsheet software. In the above embodiment, the large-scale language model 5 was installed on the large-scale language model server 4 connected to the network N. In this regard, it is also possible to configure the large-scale language model 5 to be installed on the server storage unit 12 of the information processing server 2.

[0073] In the above embodiment, it is also possible to configure the server information processing unit 10 to store some or all of the data that the server storage unit 12 is supposed to store in a storage unit of an external device that is accessible to the server information processing unit 10. • In the above embodiment, it is also possible to configure the system so that some or all of the processes described as being executed by the browser are executed by an application other than the browser. The functional blocks shown in the above embodiments can be implemented using any hardware, or through the collaboration of any hardware and any software. In other words, these functional blocks are not limited to specific hardware. Furthermore, the embodiment may include providing a program to be executed by the computer of the information processing server 2 or the user terminal 3. The embodiment may also include providing a recording medium on which the program is recorded in a way that allows it to be read by a computer. The recording medium can be a magnetic, optical, or semiconductor memory device. Specifically, examples include portable or fixed recording media such as flexible disks, HDDs (Hard Disk Drives), CD-ROMs (Compact Disk Read Only Memory), DVDs (Digital Versatile Disks), Blu-ray® Discs, magneto-optical disks, flash memory, and card-type recording media.

[0074] Furthermore, in the above embodiment, the device having one function is not limited to the device exemplified as the device having one function. For example, consider a function to generate unified format data (referred to as the "first function") and a function to display unified format data on a display unit (referred to as the "second function"). In the above embodiment, as shown in Figure 20(A), the server information processing unit 10 of the information processing server 2 has the first function, and the terminal information processing unit 13 of the user terminal 3 has the second function. In this configuration, the combination of the information processing server 2 and the user terminal 3 functions as an "information processing system". Also, the server information processing unit 10 and the terminal information processing unit 13 each function as an "information processing unit". On the other hand, as shown in Figure 20(B), the terminal information processing unit 13 of the user terminal 3 may have both the first and second functions. In this configuration, the user terminal 3 functions as an "information processing system", and the terminal information processing unit 13 functions as an "information processing unit". Also, as shown in Figure 20(C), the server information processing unit 10 may have both the first and second functions. In this configuration, the information processing server 2 functions as an "information processing system," and the server information processing unit 10 functions as an "information processing unit." However, when both the server information processing unit 10 and the terminal information processing unit 13 function as "information processing units," the functions provided in each information processing unit are not limited. Furthermore, a configuration in which a single function is realized through the cooperation of the server information processing unit 10 and the terminal information processing unit 13 is also possible. Additionally, the device that realizes a certain function may be a device other than the information processing server 2 and the user terminal 3. [Explanation of symbols]

[0075] 1. Control System (Information Processing System) 2. Information Processing Server (Information Processing System) 3. User terminal (information processing system) 4. Large-scale language model server 5. Large-scale language models 10. Server Information Processing Unit (Information Processing Unit) 13 Terminal Information Processing Unit (Information Processing Unit) 22. Unified format data 29 AI-OCR Model (Model) 34 Preview Images 37 Data Integration Table

Claims

1. The information processing unit has the function of acquiring report data relating to a report, where the data type is not limited to a specific type, and even if the data type is the same, the manner in which information about the report is recorded may differ; identifying the data type of the report data; performing information processing on the report data according to the data type; and generating unified format data that holds item values ​​for one or more specific items relating to the report in a specific data structure. An information processing system characterized by the following:

2. The aforementioned data types include PDF types corresponding to PDF data. The aforementioned information processing unit, The aforementioned PDF-type report data is provided with a function that performs character recognition using optical character recognition technology based on the functions of a machine learning model, analyzes the results of the character recognition to extract combinations of items and item values ​​recorded in the report data, and generates intermediate data in which the extracted combinations are recorded in a structured state. The system includes a function to generate the unified format data based on the intermediate data. The information processing system according to feature 1.

3. The aforementioned information processing unit, With respect to the PDF-type report data, the model has a function to derive the accuracy of the item values ​​for each extracted combination and to generate intermediate data in which the derived accuracy is recorded. The system includes a function to display the unified format data on the display unit with a preview image of the document data, along with an editable display of the item values ​​of specific items, and with information reflecting the accuracy of the item values ​​of the specific items or a warning based on the accuracy displayed. The information processing system according to feature 2.

4. The aforementioned data type includes a CSV type corresponding to CSV data. The aforementioned information processing unit, The CSV-type report data is provided with a function to generate the unified format data from the report data by setting the item value of the corresponding item in the unified format data to the specific item in the unified format data. The information processing system according to feature 1.

5. The aforementioned data type includes an audio type corresponding to the audio data. The aforementioned voice-type document data includes a spoken message uttered by the speaker that contains information about the document. The aforementioned information processing unit, If the data type of the acquired report data is the voice type, The aforementioned document data is recognized using speech recognition technology, and speech message data is generated in which the spoken messages are recorded as text. A prompt is generated requesting the extraction of the item value of a specific item contained in the utterance message recorded in the utterance message data; the prompt is sent to the generating AI model server; and a response sentence containing the extracted item value of the specific item is received from the generating AI model server. It includes a function to generate the unified format data based on the received response text. The information processing system according to feature 1.

6. The aforementioned information processing unit, The prompt includes a request to derive the accuracy of the item value of the extracted specific item from the aforementioned voice-type report data, The aforementioned voice-type report data is provided with a function that displays the unified format data on the display unit in a state where the item value of a specific item is shown in an editable manner along with a preview image of the spoken message data, and information reflecting the accuracy of the item value of the specific item or a warning based on the accuracy is shown. The information processing system according to feature 5.

7. The aforementioned data type includes a text type corresponding to text data, The aforementioned text-type report data includes input messages entered by the user that have information about the report, The aforementioned information processing unit, If the data type of the acquired report data is the text type, A prompt is generated requesting the extraction of the item value of the specific item contained in the input message recorded in the report data; the prompt is sent to the generating AI model server; and a response sentence containing the extracted item value of the specific item is received from the generating AI model server. It includes a function to generate the unified format data based on the received response text. The information processing system according to feature 1.

8. The aforementioned information processing unit, The prompt includes a request to derive the accuracy of the item value of the extracted specific item from the aforementioned text-type report data, The system includes a function to display the unified format data on the display unit with the item values ​​of specific items shown in an editable manner along with a preview image of the document data, and with information reflecting the accuracy of the item values ​​of the specific items or a warning based on the accuracy displayed for those items. The information processing system according to feature 7.

9. The aforementioned information processing unit, The system includes a function to display a single table on the display unit that shows all of the unified format data for one or more of the aforementioned report data specified by the user together. The information processing system according to any one of claims 1 to 8.

10. The aforementioned information processing unit, The system includes a function to display a preview image of the specified report data on the display unit, along with the table configured to allow editing of arbitrary item values. The information processing system according to feature 9.

11. The aforementioned information processing unit, The system includes a function to output the table as predetermined structured data in accordance with the user's instructions. The information processing system according to feature 9.

12. The information processing system acquires document data, which is data relating to a document, the data type is not limited to a specific type, and even if the data type is the same, the manner in which information about the document is recorded may differ; The information processing system includes the steps of: identifying the data type of the report data; performing information processing on the report data according to the data type; and generating unified format data that holds item values ​​for one or more specific items relating to the report in a specific data structure. An information processing method characterized by the following:

13. A program executed on a computer in an information processing system, The aforementioned computer, This information processing unit has the function of acquiring report data relating to a report, where the data type is not limited to a specific type, and even if the data type is the same, the manner in which information about the report is recorded may differ; identifying the data type of the report data; performing information processing on the report data according to the data type; and generating unified format data that holds item values ​​for one or more specific items relating to the report in a specific data structure. A program characterized by the following features.

Images