Document processing system, document processing device, and document processing program

Abstract

Provided is a document processing system comprising a rule generation unit that generates a document naming rule using a document naming rule generation model, which is a machine learning model for generating the document naming rule for giving a filename of a document image, which is an image of a document, wherein: the document naming rule determines the filename of the document image in accordance with metadata of the document image; the document naming rule generation model outputs a candidate for the document naming rule when the filename of the document image and the metadata are input; and the rule generation unit inputs, into the document naming rule generation model, the filename of the document image and the metadata that are stored in a storage unit, thereby generating the document naming rule (S123).

Description

Document Processing System, Document Processing Apparatus, and Document Processing Program

[0001] The present invention relates to a document processing system, a document processing apparatus, and a document processing program that generate a document naming rule serving as a rule for assigning a file name to a document image that is an image of a document.

[0002] Conventionally, there has been known an image processing system that assigns a file name to a document image in accordance with a document naming rule serving as a rule for assigning a file name to a document image (see, for example, Patent Document 1).

[0003] Japanese Unexamined Patent Publication No. 2018-124656

[0004] However, conventional image processing systems have a problem in that a user bears a heavy burden for generating a document naming rule.

[0005] Accordingly, an object of the present invention is to provide a document processing system, a document processing apparatus, and a document processing program that can reduce a user's burden for generating a document naming rule.

[0006] A document processing system according to one aspect of the present invention generates the document naming rule by using a document naming rule generation model that is a machine learning model for generating a document naming rule serving as a rule for assigning a file name to a document image that is an image of a document, and includes a rule generation unit that generates the document naming rule. The document naming rule determines the file name of the document image in accordance with metadata of the document image. The document naming rule generation model outputs a candidate for the document naming rule when the file name and the metadata of the document image are input thereto. The rule generation unit generates the document naming rule by inputting the file name and the metadata of the document image stored in a document storage unit that stores the document image into the document naming rule generation model.

[0007] With this configuration, the document processing system according to one aspect of the present invention generates document naming rules by inputting the file names and metadata of document images stored in the document storage unit into the document naming rule generation model, thereby reducing the burden on the user in generating document naming rules.

[0008] In a document processing system according to one aspect of the present invention, the document naming rule generation model may output a confidence level for each candidate when outputting candidates, and the rule generation unit may determine the candidate with the highest confidence level among the candidates output by the document naming rule generation model as the document naming rule.

[0009] With this configuration, the document processing system according to one aspect of the present invention determines the candidate with the highest confidence level among the document naming rule candidates output by the document naming rule generation model as the document naming rule. As a result, the number of operations required by the user to generate document naming rules can be reduced, and consequently, the burden on the user to generate document naming rules can be reduced.

[0010] In a document processing system according to one aspect of the present invention, the rule generation unit may generate the document naming rules using the document naming rule generation model through a dialogue via a dialogue UI, which is a chatbot, that performs a dialogue with the user regarding the generation of the document naming rules.

[0011] With this configuration, the document processing system according to one aspect of the present invention generates document naming rules using a document naming rule generation model through interaction with the user via an interactive UI. As a result, the convenience of user operation for generating document naming rules can be improved, and consequently, the burden on the user for generating document naming rules can be reduced.

[0012] A document processing device according to one aspect of the present invention includes a rule generation unit that generates a document naming rule using a document naming rule generation model as a machine learning model for generating a document naming rule as a rule for assigning file names to document images as images of documents, wherein the document naming rule determines the file name of the document image according to the metadata of the document image, the document naming rule generation model outputs a candidate for the document naming rule when the file name of the document image and the metadata are input to the document naming rule generation unit that generates the document naming rule by inputting the file name of the document image and the metadata stored in a document storage unit that stores the document image to the document naming rule generation model.

[0013] With this configuration, the document processing device according to one aspect of the present invention generates document naming rules by inputting the file names and metadata of document images stored in the document storage unit into a document naming rule generation model, thereby reducing the burden on the user in generating document naming rules.

[0014] A document processing program according to one aspect of the present invention implements a rule generation unit in a computer that generates document naming rules, which are rules for assigning file names to document images as images of documents, using a document naming rule generation model as a machine learning model for generating said document naming rules, the document naming rules determine the file name of the document image according to the metadata of the document image, the document naming rule generation model outputs candidate document naming rules when the file name of the document image and the metadata are input to the document naming rule generation unit that generates the document naming rules by inputting the file name of the document image and the metadata stored in a document storage unit that stores the document image to the document naming rule generation model.

[0015] With this configuration, a computer executing a document processing program according to one aspect of the present invention generates document naming rules by inputting the file names and metadata of document images stored in the document storage unit into a document naming rule generation model, thereby reducing the burden on the user in generating document naming rules.

[0016] A document processing system, document processing device, and document processing program according to one aspect of the present invention can reduce the burden on users in generating document naming rules.

[0017] This is a block diagram of an example system according to one embodiment of the present invention. This is a block diagram of an example document processing system shown in Figure 1 when it is configured with one computer. This is a diagram showing an example of user management information shown in Figure 2. This is a diagram showing an example of document naming rules shown in Figure 2. This is a diagram showing an example of rule management information shown in Figure 2. This is a diagram showing an example of metadata for a document image shown in Figure 2. This is a block diagram of an example user terminal shown in Figure 1 when it is configured with a PC. This is a block diagram of an example image reader shown in Figure 1 when it is configured with an MFP. This is a sequence diagram of the operation of the system shown in Figure 1 when a user terminal logs into the document processing system. This is a flowchart of the operation of the document processing system shown in Figure 2 when document naming rules are automatically generated. This is a flowchart of the operation of the document processing system shown in Figure 2 when document naming rules are generated interactively. This is a diagram showing an example of an interactive screen displayed in the operation shown in Figure 11. This is a sequence diagram of the operation of the system shown in Figure 1 when document naming rules are displayed. This is a sequence diagram of the operation of the system shown in Figure 1 when a user terminal inputs a document image into the document processing system. This is a sequence diagram of the operation of the system shown in Figure 1 when an image reader logs into the document processing system. Figure 8 is a flowchart illustrating the operation of the image reader when a document image is input into the document processing system. Figure 2 is a flowchart illustrating the operation of the document processing system when a document image has been input.

[0018] Hereinafter, embodiments of the present invention will be described with reference to the drawings.

[0019] First, the configuration of a system according to one embodiment of the present invention will be described.

[0020] Figure 1 is a block diagram of an example of the system 10 according to this embodiment.

[0021] As shown in Figure 1, system 10 includes a document processing system 20 that processes document images as images of documents. The document processing system (an example of a document processing device) 20 may be configured as a single computer, such as a PC (Personal Computer), or as multiple computers. In this case, the document processing system 20 is an example of a document processing device. The document processing system 20 may also be configured on the cloud.

[0022] System 10 includes a user terminal 30 used by the user. The user terminal 30 may consist of, for example, a PC, tablet, or smartphone.

[0023] System 10 includes an image reading device 40 that reads an image from a document. The image reading device 40 may be configured as, for example, a dedicated scanner, or as an MFP (Multifunction Peripheral).

[0024] Figure 2 is a block diagram of an example of a document processing system 20 that is configured with a single computer.

[0025] As shown in Figure 2, the document processing system 20 includes an operation unit 21, which is an operation device such as a keyboard or mouse into which various operations are input; a display unit 22, which is a display device such as an LCD (Liquid Crystal Display) that displays various information; a communication unit 23, which is a communication device that communicates with external devices via a network such as a LAN (Local Area Network) or the Internet, or directly via wired or wireless connection without going through a network; a storage unit 24, which is a non-volatile storage device such as a semiconductor memory or HDD (Hard Disk Drive) that stores various information; and a control unit 25 that controls the entire document processing system 20.

[0026] The storage unit 24 can store a document processing program 24a for processing document images. The document processing program 24a may be installed in the document processing system 20 during the manufacturing stage, or it may be additionally installed in the document processing system 20 from an external storage medium such as a USB (Universal Serial Bus) memory, or it may be additionally installed in the document processing system 20 from a network.

[0027] The storage unit 24 is capable of storing user management information 24b for managing users of the document processing system 20.

[0028] Figure 3 shows an example of user management information 24b.

[0029] As shown in Figure 3, the user management information 24b includes, for each user, a user ID as user identification information, a user password, a user name as user name, and a user email address. The user management information 24b shown in Figure 3 is depicted with some information omitted.

[0030] As shown in Figure 2, the storage unit 24 can store a document naming rule 24c, which is a rule for assigning file names to document images. The storage unit 24 can store at least one other document naming rule besides the document naming rule 24c. The document naming rule determines the file name of the document image according to the metadata of the document image.

[0031] Figure 4 shows an example of document naming rule 24c.

[0032] In the document naming rule 24c shown in Figure 4, {receipt date yyyyMMdd} indicates that the date on which the original document of the document image was received (hereinafter referred to as the "receipt date") is represented in the format "yyyyMMdd". "yyyy" indicates that the year of the receipt date is represented by four digits in the Western calendar. "MM" indicates that the month of the receipt date is represented by two digits. "dd" indicates that the day of the receipt date is represented by two digits.

[0033] In the document naming rule 24c shown in Figure 4, {Issuer Name} indicates the name of the issuing authority of the original document on which the document image is based (hereinafter referred to as "Issuer Name").

[0034] In the document naming rule 24c shown in Figure 4, {Document Type} indicates the type of document on which the document image is based (hereinafter referred to as "Document Type"). The Document Type is, for example, the type of document such as a receipt or an invoice.

[0035] For example, if the receipt date, issuer name, and document type are "February 28, 2025", "ABC", and "Receipt", respectively, the file name generated by the document naming rule 24c shown in Figure 4 will be "20250228_ABC_Receipt".

[0036] As shown in Figure 2, the storage unit 24 can store rule management information 24d for managing document naming rules.

[0037] Figure 5 shows an example of rule management information 24d.

[0038] The rule management information 24d shown in Figure 5 includes a rule ID as identification information for the document naming rule and the conditions under which the document naming rule applies (hereinafter referred to as "rule application conditions") for each document naming rule. The rule application conditions include the file size of the document image to which the document naming rule applies, the file extension of the document image to which the document naming rule applies, the document type of the original document to which the document naming rule applies, and the output destination of the file of the document image to which the document naming rule applies. Some information is omitted in the depiction of the rule management information 24d shown in Figure 5.

[0039] The output destination may specify a particular folder on a particular storage device.

[0040] As shown in Figure 2, the storage unit 24 is capable of storing document images 24e. The storage unit 24 is capable of storing at least one document image other than document images 24e. In other words, the storage unit 24 constitutes a document storage unit that stores document images.

[0041] Figure 6 shows an example of metadata 24f for a document image 24e.

[0042] As shown in Figure 6, the metadata 24f includes, for example, the file size of the document image 24e, the file extension of the document image 24e, the name of the person who uploaded the document image 24e to the document processing system 20 (hereinafter, the name of the person who uploaded the document image to the document processing system 20 is referred to as the "creator name"), the receipt date of the original document of the document image 24e, the name of the issuer of the original document of the document image 24e, and the document type of the original document of the document image 24e.

[0043] The control unit 25 shown in Figure 2 includes, for example, a CPU (Central Processing Unit), a ROM (Read Only Memory) that stores programs and various data, and a RAM (Random Access Memory) used as a working area for the CPU of the control unit 25. The CPU of the control unit 25 executes programs stored in the storage unit 24 or the ROM of the control unit 25.

[0044] By executing a document processing program 24a, the control unit 25 implements: an authentication execution unit 25a that performs user authentication; a management UI 25b serving as a UI (User Interface) for managing document naming rules; a rule generation unit 25c that generates document naming rules; a rule management unit 25d that manages document naming rules; a dialogue UI 25e which is a chatbot that conducts dialogue regarding the generation of document naming rules with a user; a document upload interface 25f that receives document images; and a document analysis unit 25g that analyzes document images received by the document upload interface 25f.

[0045] The rule generation unit 25c includes a machine learning model (hereinafter referred to as a "document naming rule generation model") 25h for generating document naming rules. The document naming rule generation model 25h outputs at least one candidate document naming rule when a combination of a file name and metadata of at least one document image is input. The document naming rule generation model 25h can also output a confidence degree for each output candidate. The confidence degree is an index indicating how confident the machine learning model is in a specific prediction. The confidence degree is expressed as a probability value in a range from 0 to 1, where a value closer to 1 means that the machine learning model has higher confidence in the prediction. The confidence degree can be obtained from the output layer of the machine learning model.

[0046] FIG. 7 is a block diagram of an example user terminal 30 configured by a PC.

[0047] As shown in FIG. 7, the user terminal 30 includes: an operation unit 31 which is an operation device such as a keyboard or a mouse for inputting various operations; a display unit 32 which is a display device such as an LCD for displaying various information; a communication unit 33 which is a communication device that communicates with external devices via a network such as a LAN or the Internet, or directly by wire or wireless without going through a network; a storage unit 34 which is a non-volatile storage device such as a semiconductor memory or an HDD for storing various information; and a control unit 35 that controls the entire user terminal 30.

[0048] The storage unit 34 can store a program for a web browser (hereinafter referred to as "web browser program") 34a. For example, the web browser program 34a may be installed in the user terminal 30 at the manufacturing stage of the user terminal 30, may be additionally installed in the user terminal 30 from an external storage medium such as a USB memory, or may be additionally installed in the user terminal 30 from a network.

[0049] The control unit 35 includes, for example, a CPU, a ROM that stores programs and various types of data, and a RAM used as a memory serving as a work area for the CPU of the control unit 35. The CPU of the control unit 35 executes programs stored in the storage unit 34 or the ROM of the control unit 35.

[0050] The control unit 35 implements the web browser 35a by executing the web browser program 34a.

[0051] Fig. 8 is a block diagram of an example of the image reading apparatus 40 when configured by an MFP.

[0052] As shown in Fig. 8, the image reading apparatus 40 is a computer including: an operation unit 41 which is an operation device such as a button for inputting various operations; a display unit 42 which is a display device such as an LCD (Liquid Crystal Display) for displaying various information; a printer 43 which is a printing device that prints images on a recording medium such as paper; a scanner 44 which is a reading device that reads images from a document; a communication unit 45 which is a communication device that communicates with an external device via a network such as a LAN or the Internet, or directly by wire or wireless without going through a network; a fax communication unit 46 which is a fax device that performs fax communication with an external facsimile apparatus (not shown) via a communication line such as a public telephone line; a storage unit 47 which is a non-volatile storage device such as a semiconductor memory or an HDD (Hard Disk Drive) that stores various types of information; and a control unit 48 that controls the entire image reading apparatus 40.

[0053] The memory unit 47 is capable of storing an image reading program 47a for reading images from a document. The image reading program 47a may, for example, be installed in the image reading device 40 during the manufacturing stage, or it may be additionally installed in the image reading device 40 from an external storage medium such as a USB memory, or it may be additionally installed in the image reading device 40 from a network.

[0054] The control unit 48 includes, for example, a CPU, a ROM that stores programs and various data, and RAM as memory used as a working area for the CPU of the control unit 48. The CPU of the control unit 48 executes programs stored in the storage unit 47 or the ROM of the control unit 48.

[0055] The control unit 48 implements an image reading unit 48a that reads an image from a document by executing an image reading program 47a.

[0056] Next, we will describe the operation of system 10.

[0057] First, we will explain the operation of system 10 when a user terminal 30 logs in to the document processing system 20.

[0058] Figure 9 is a sequence diagram of the operation of system 10 when a user terminal 30 logs in to the document processing system 20.

[0059] The user can instruct the user terminal 30 via the operation unit 31 of the user terminal 30 to access the web page for logging into the document processing system 20 (hereinafter referred to as the "login page") provided by the document processing system 20. When the user terminal 30 is instructed to access the login page, the web browser 35a of the user terminal 30 accesses the login page (S101).

[0060] When the authentication execution unit 25a of the document processing system 20 receives an access request in S101, it sends the login page data to the web browser 35a (S102).

[0061] When the web browser 35a of the user terminal 30 receives the transmitted data in S102, it displays a login page on the display unit 32 based on the received data (S103). The login page can either allow the user to enter their user ID and password, or, after the user ID and password have been entered, instruct the user to log in to the document processing system 20.

[0062] When the web browser 35a is instructed on the login page to log in to the document processing system 20, it requests the document processing system 20 to log in using the user ID and password combination that was entered on the login page at the time the login instruction was given (S104).

[0063] When the authentication execution unit 25a of the document processing system 20 receives the request in S104, it performs authentication of the user who is the target of the request in S104 (S105). Specifically, the authentication execution unit 25a determines that user authentication was successful if the combination of user ID and password included in the request in S104 is included in the user management information 24b. On the other hand, the authentication execution unit 25a determines that user authentication failed if the combination of user ID and password included in the request in S104 is not included in the user management information 24b.

[0064] If the authentication execution unit 25a of the document processing system 20 determines that the authentication performed in S105 has failed, it notifies the user terminal 30 that login is not permitted (S106).

[0065] When the web browser 35a of the user terminal 30 receives the notification in S106, it displays "Login denied" on the display unit 32 (S107).

[0066] When the authentication execution unit 25a of the document processing system 20 determines that the authentication performed in S105 was successful, it notifies the user terminal 30 that login is permitted (S108).

[0067] When the web browser 35a of the user terminal 30 receives the notification in S108, it displays permission to log in on the display unit 32 (S109).

[0068] Next, we will explain the operation of the document processing system 20 when automatically generating document naming rules.

[0069] Figure 10 is a flowchart showing the operation of the document processing system 20 when automatically generating document naming rules.

[0070] If a user is logged into the document processing system 20 from the user terminal 30, they can input an instruction to automatically generate document naming rules (hereinafter referred to as "automatic rule generation instruction") to the user terminal 30 via the operation unit 31 of the user terminal 30. The automatic rule generation instruction includes the specification of at least one document image from among the document images stored in the storage unit 24 of the document processing system 20. There are several methods for specifying document images, for example, specifying document images one by one, specifying any folder so that all document images contained in the specified folder are specified together, and specifying the value of at least one item in the metadata of the document image so that all document images in which the specified value is included in the metadata are specified together. When the user terminal 30's web browser 35a receives the automatic rule generation instruction, it sends the input instruction to the document processing system 20. When the document processing system 20's management UI 25b receives the automatic rule generation instruction sent from the user terminal 30, it executes the operation shown in Figure 10.

[0071] As shown in Figure 10, the management UI 25b of the document processing system 20 notifies the rule generation unit 25c of the automatic rule generation instruction that triggered the operation shown in Figure 10 (hereinafter referred to as the "target rule automatic generation instruction" in the description of the operation shown in Figure 10) (S121).

[0072] When the rule generation unit 25c receives the notification in S121, it retrieves the file names and metadata of all document images specified in the automatic rule generation instruction notified in S121 from the storage unit 24 (S122).

[0073] When the processing in S122 is completed, the rule generation unit 25c generates a document naming rule using the document naming rule generation model 25h based on the combination of file name and metadata for each document image acquired in S122 (S123). Here, the document naming rule generation model 25h is capable of generating multiple candidate document naming rules. However, in S123, the rule generation unit 25c selects the document naming rule with the highest confidence level from among the candidate document naming rules generated by the document naming rule generation model 25h as the document naming rule.

[0074] When the processing in S123 is completed, the rule generation unit 25c notifies the rule management unit 25d of the document naming rules generated in S123 (S124).

[0075] When the rule management unit 25d receives the notification in S124, it saves the document naming rule notified in S124 to the storage unit 24 (S125).

[0076] When the processing in S125 is completed, the rule management unit 25d saves the document naming rule information stored in the storage unit 24 in S125 to the rule management information 24d (S126). If, in S125, the user has entered the rule application conditions for the document naming rule stored in the storage unit 24 to the document processing system 20 via the user terminal 30 before or after the start of the operation shown in Figure 10, the rule management unit 25d saves the rule application conditions entered via the user terminal 30 to the rule management information 24d in S126 as the rule application conditions for the document naming rule stored in the storage unit 24 in S125.

[0077] When the process in S126 is executed, the rule management unit 25d terminates the operation shown in Figure 10.

[0078] Next, we will describe the operation of the document processing system 20 when generating document naming rules interactively.

[0079] Figure 11 is a flowchart showing the operation of the document processing system 20 when generating document naming rules interactively.

[0080] If a user is logged into the document processing system 20 from the user terminal 30, they can input instructions to generate document naming rules interactively (hereinafter referred to as "rule dialogue generation instructions") into the user terminal 30 via the operation unit 31 of the user terminal 30. When the user terminal 30's web browser 35a receives the rule dialogue generation instructions, it transmits the input instructions to the document processing system 20. When the document processing system 20's management UI 25b receives the rule dialogue generation instructions transmitted from the user terminal 30, it executes the actions shown in Figure 11.

[0081] As shown in Figure 11, the management UI 25b instructs the dialogue UI 25e to perform a process (S141).

[0082] When the interactive UI 25e receives the instruction in S141, it displays a screen for generating document naming rules while interacting with the user (hereinafter referred to as the "interaction screen") 300 (see, for example, Figure 12) on the user terminal 30 (S142).

[0083] Figure 12 shows an example of the dialogue screen 300.

[0084] In the dialogue screen 300 shown in Figure 12, the speech bubble on the left shows the content entered by the user, and the speech bubble on the right shows the content output by the document processing system 20.

[0085] As shown in Figure 11, once the processing in S142 is complete, the interactive UI 25e determines whether the desired file name content and the target document image have been specified until it determines that the content of the file name that the user wishes to be assigned to the document image (hereinafter referred to as the "desired file name content") and at least one document image stored in the storage unit 24 to which the user wishes to be assigned a file name corresponding to the desired file name content (hereinafter referred to as the "target document image" in the explanation of the operation shown in Figure 11) have been specified (S143). The user can specify the desired file name content and the target document image on the interactive screen 300. In the example shown in Figure 12, the desired file name content is "20240725Thr_ABC_Receipt" in the message "I want to name it 20240725Thr_ABC_Receipt". In the example shown in Figure 12, the target document image is a document image with the file name "July 25th_abc.pdf". The desired file name content can be a complete and specific file name, such as "20240725Thr_ABC_Receipt" as shown in Figure 12, or it can be a specification of only the items corresponding to the values ​​to be included in the file name, such as "I want a name that includes the receipt date and the name of the issuer."

[0086] When the rule generation unit 25c determines in S143 that the desired file name content and the target document image have been specified, it uses the document naming rule generation model 25h to obtain candidate document naming rules and the confidence level for each candidate based on the combination of the file name corresponding to the desired file name content specified in the dialogue screen 300 and the metadata of the target document image specified in the dialogue screen 300 (S144). The dialogue UI 25e and the rule generation unit 25c may also recognize that if the desired file name content specified in the dialogue screen 300 is "20240725Thr_ABC_Receipt" as shown in Figure 12, then "Thr" in "20240725Thr_ABC_Receipt" indicates Thursday.

[0087] When the processing in S144 is completed, the rule generation unit 25c generates filenames to be assigned to the target document image for each of the top number of candidates with the highest confidence levels among the candidates obtained in S144 (S145).

[0088] When the processing in S145 is completed, the interactive UI 25e displays the file name generated in S145 on the interactive screen 300 (S146). In the example shown in Figure 12, the file names "07252024_07_ABC_Receipt", "240725_ABC_Receipt", and "0725_ABC_Receipt" are displayed as file names assigned to the target document image according to each of the three candidate document naming rules. The user can specify any of the file names displayed on the interactive screen 300. The user can also request a file name other than those displayed on the interactive screen 300. When requesting a file name other than those displayed on the interactive screen 300, the user can also add new conditions to the file name. In the example shown in Figure 12, the message "Do you have any other suggestions? The day of the week is not necessary" indicates that the user is requesting a file name other than those displayed on the interactive screen 300. Furthermore, in the example shown in Figure 12, the message "Do you have any other suggestions? You don't need to specify the day of the week" indicates a new condition added to the file name.

[0089] When the processing in S146 is completed, the rule generation unit 25c determines whether or not one of the file names displayed on the interactive screen 300 has been specified (S147).

[0090] If the rule generation unit 25c determines in S147 that none of the file names displayed on the interactive screen 300 have been specified, it determines whether a file name other than those displayed on the interactive screen 300 has been requested (S148).

[0091] If the rule generation unit 25c determines in S148 that no file name other than the file name displayed on the interactive screen 300 has been requested, it executes the process in S147.

[0092] If the rule generation unit 25c determines in S148 that a file name other than the file name displayed on the interactive screen 300 has been requested, it determines whether or not a new condition has been added to the file name (S149).

[0093] If the rule generation unit 25c determines in S149 that a new condition has been added regarding the file name, it uses the document naming rule generation model 25h to obtain candidate document naming rules and a confidence level for each candidate based on the combination of the desired file name content specified in the interactive screen 300, the file name corresponding to all the conditions added regarding the file name after the desired file name content was specified in the interactive screen 300, and the metadata of the target document image specified in the interactive screen 300 (S150).

[0094] If the rule generation unit 25c determines in S149 that no new conditions have been added regarding the file name, or if the process in S150 is completed, it generates a file name to be assigned to the target document image based on a specific number of top candidates, in descending order of confidence level obtained in the most recent process of S144 and S150, that have not yet been displayed on the interactive screen 300 in the operation shown in Figure 11 (S151).

[0095] When the processing in S151 is completed, the interactive UI 25e displays the file name generated by S151 on the interactive screen 300 (S146). In the example shown in Figure 12, one file name, "20240725_ABC_Receipt", is displayed.

[0096] If the rule generation unit 25c determines in S147 that one of the filenames displayed on the interactive screen 300 has been specified, it notifies the rule management unit 25d of a candidate document naming rule corresponding to the filename specified on the interactive screen 300 as a new document naming rule (S152). In the example shown in Figure 12, the user has specified the filename "20240725_ABC_Receipt" by sending the message "OK".

[0097] When the rule management unit 25d receives the notification in S152, it saves the document naming rule notified in S152 to the storage unit 24 (S153).

[0098] When the processing in S153 is completed, the rule management unit 25d saves the document naming rule information stored in the storage unit 24 in S153 to the rule management information 24d (S154). If, in S153, the user has entered the rule application conditions for the document naming rule stored in the storage unit 24 to the document processing system 20 via the user terminal 30 before or after the start of the operation shown in Figure 11, the rule management unit 25d saves the rule application conditions entered via the user terminal 30 to the rule management information 24d in S154 as the rule application conditions for the document naming rule stored in the storage unit 24 in S153.

[0099] When the process in S154 is executed, the rule management unit 25d terminates the operation shown in Figure 11.

[0100] Next, we will describe the operation of system 10 when displaying document naming rules.

[0101] Figure 13 is a sequence diagram of the operation of system 10 when displaying document naming rules.

[0102] If a user is logged into the document processing system 20 from the user terminal 30, they can input a command to display document naming rules (hereinafter referred to as "rule display command") into the user terminal 30 via the operation unit 31 of the user terminal 30. When a rule display command is input, the web browser 35a of the user terminal 30 transmits the input rule display command to the document processing system 20, as shown in Figure 13 (S161).

[0103] When the management UI 25b of the document processing system 20 receives the rule display instruction transmitted in S161, it requests the rule management unit 25d to provide the document naming rule specified by the received rule display instruction (S162).

[0104] When the rule management unit 25d receives a request in S162, it retrieves the document naming rule requested in S162 and the rule application conditions associated with this document naming rule in the rule management information 24d from the storage unit 24 (S163).

[0105] When the processing in S163 is completed, the management UI 25b transmits the data of the web page (hereinafter referred to as the "rule display page") that shows the document naming rules and rule application conditions acquired in S163 to the user terminal 30 (S164).

[0106] When the web browser 35a of the user terminal 30 receives the data transmitted in S164, it displays a rule display page on the display unit 32 based on this data (S165). The user can input instructions for editing document naming rules and rule application conditions displayed on the rule display page (hereinafter referred to as "rule editing instructions") and instructions for deleting document naming rules displayed on the rule display page (hereinafter referred to as "rule deletion instructions").

[0107] When a rule editing instruction is entered on the rule display page, the web browser 35a sends this rule editing instruction to the document processing system 20 (S166).

[0108] When the management UI 25b of the document processing system 20 receives the rule editing instruction sent in S166, it requests the rule management unit 25d to change the document naming rule and rule application conditions in accordance with this rule editing instruction (S167).

[0109] When the rule management unit 25d receives the request in S167, it modifies the document naming rules and rule application conditions in accordance with the request in S167 (S168).

[0110] When a rule deletion instruction is entered on the rule display page, the web browser 35a sends this rule deletion instruction to the document processing system 20 (S169).

[0111] When the management UI 25b of the document processing system 20 receives a rule deletion instruction sent in S169, it requests the rule management unit 25d to delete the document naming rule and rule application conditions corresponding to this rule deletion instruction (S170).

[0112] When the rule management unit 25d receives the request in S170, it deletes the document naming rule and rule application conditions from the storage unit 24 in accordance with the request in S170 (S171). In addition, the rule management unit 25d deletes the rule application conditions for the document naming rule from the storage unit 24 by deleting the record of the document naming rule from the rule management information 24d in accordance with the request in S170.

[0113] Next, we will explain the operation of system 10 when the user terminal 30 inputs a document image into the document processing system 20.

[0114] Figure 14 is a sequence diagram of the operation of system 10 when a user terminal 30 inputs a document image into the document processing system 20.

[0115] If a user is logged into the document processing system 20 from the user terminal 30, the user terminal 30 can be instructed via the operation unit 31 to access the web page for uploading document images provided by the document processing system 20 (hereinafter referred to as the "document upload page"). When the user terminal 30 is instructed to access the document upload page, the web browser 35a of the user terminal 30 accesses the document upload page as shown in Figure 14 (S181).

[0116] When the document upload interface 25f of the document processing system 20 receives an access request in S181, it sends the data for the document upload page to the web browser 35a (S182).

[0117] When the web browser 35a of the user terminal 30 receives the transmitted data in S182, it displays a document upload page on the display unit 32 based on the received data (S183). The user can instruct the upload of a document image via the operation unit 31 of the user terminal 30 on the document upload page. Here, when the user instructs the upload of a document image on the document upload page, the user specifies the output destination of the document image to be uploaded on the document upload page. The document image to be uploaded on the document upload page may be, for example, a document image read from a document as an original by the image reading device 40, or a document image generated by taking a picture of the document with a camera (not shown).

[0118] When the web browser 35a is instructed to upload a document image on the document upload page, it transmits the document image that has been instructed to be uploaded, along with information indicating the output destination specified on the document upload page (hereinafter, the information indicating the output destination is referred to as "output destination information"), to the document processing system 20 (S184).

[0119] Next, we will explain the operation of system 10 when logging in to document processing system 20 from image reading device 40.

[0120] Figure 15 is a sequence diagram of the operation of system 10 when logging in from image reading device 40 to document processing system 20.

[0121] Users of the image reading device 40 can instruct the image reading device 40 via the operation unit 41 to display a login screen for the document processing system 20 (hereinafter referred to as the "login screen"). When the image reading unit 48a of the image reading device 40 is instructed to display the login screen, it displays the login screen on the display unit 42 (S201). The login screen can either allow the user to enter their user ID and password, or, after the user ID and password have been entered, instruct the user to log in to the document processing system 20.

[0122] When the image reading unit 48a is instructed to log in to the document processing system 20 on the login screen, it requests the document processing system 20 to log in using the combination of user ID and password that was entered on the login screen at the time the login to the document processing system 20 was instructed (S202).

[0123] When the authentication execution unit 25a of the document processing system 20 receives the request in S202, it performs authentication of the user who is the target of the request in S202 (S203). Specifically, the authentication execution unit 25a determines that user authentication has been successful if the combination of user ID and password included in the request in S202 is included in the user management information 24b. On the other hand, the authentication execution unit 25a determines that user authentication has failed if the combination of user ID and password included in the request in S202 is not included in the user management information 24b.

[0124] If the authentication execution unit 25a of the document processing system 20 determines that the authentication performed in S203 has failed, it notifies the image reading device 40 that login is not permitted (S204).

[0125] When the image reading unit 48a of the image reading device 40 receives the notification in S204, it displays "Login denied" on the display unit 42 (S205).

[0126] When the authentication execution unit 25a of the document processing system 20 determines that the authentication performed in S203 was successful, it notifies the image reading device 40 that login is permitted (S206).

[0127] When the image reading unit 48a of the image reading device 40 receives the notification in S206, it displays "Login permission" on the display unit 42 (S207).

[0128] Next, we will explain the operation of the image reading device 40 when inputting a document image into the document processing system 20.

[0129] Figure 16 is a flowchart showing the operation of the image reading device 40 when a document image is input into the document processing system 20.

[0130] If the user is logged into the document processing system 20 from the image reading device 40, they can input instructions to the image reading device 40 via the operation unit 41 of the image reading device 40 to execute a process that includes reading an image from a document and sending this image to the document processing system 20 (hereinafter referred to as "scan transmission processing instructions"). Here, the user includes the output destination of the image read from the document in the scan transmission processing instructions. When the scan transmission processing instructions are input, the image reading unit 48a of the image reading device 40 executes the operation shown in Figure 16.

[0131] As shown in Figure 16, the image reading unit 48a reads an image, that is, a document image, from the document set in the scanner 44 (S221).

[0132] When the processing in S221 is completed, the image reading unit 48a transmits the document image read in S221, along with the output destination information indicating the output destination included in the scan transmission processing instruction, to the document processing system 20 (S222), thereby ending the operation shown in Figure 16.

[0133] Next, we will describe the operation of the document processing system 20 when a document image is input.

[0134] Figure 17 is a flowchart showing the operation of the document processing system 20 when a document image is input.

[0135] The document processing system 20 performs the operation shown in Figure 17 when, for example, a document image is transmitted from a user terminal 30 or an image reader 40 as described above.

[0136] As shown in Figure 17, the document upload interface 25f of the document processing system 20 receives a document image (hereinafter referred to as "target document image" in the explanation shown in Figure 17) transmitted from the user terminal 30 or the image reader 40 (S241). Note that the metadata of the target document image may already contain some information, such as the file size and extension of the target document image.

[0137] When the processing in S241 is completed, the document analysis unit 25g includes the user name in the user management information 24b of the user who uploaded the target document image as the creator name in the metadata of the target document image (S242).

[0138] When the processing in S242 is completed, the document analysis unit 25g performs OCR (Optical Character Recognition) on the target document image to extract strings representing the values ​​of specific items from the target document image for each item (S243). Various methods can be used to extract the values ​​of specific items from the target document image. For example, the document analysis unit 25g may use a template that indicates the location where the string representing the value of a specific item is placed to extract the string representing the value of a specific item from the target document image.

[0139] When the processing in S243 is completed, the document analysis unit 25g includes the strings for each item extracted in S243 in the metadata of the target document image (S244). For example, the document analysis unit 25g includes information such as the date of receipt, the name of the issuer, and the document type in the metadata in S244.

[0140] When the processing in S244 is completed, the rule management unit 25d identifies the document naming rule to be applied to the target document image based on the metadata of the target document image, the output destination indicated in the output destination information received along with the target document image in S241, and the rule management information 24d (S245).

[0141] When the processing in S245 is completed, the rule management unit 25d determines the file name of the target document image based on the metadata of the target document image and the document naming rule identified in S245 (S246).

[0142] When the processing in S246 is completed, the rule generation unit 25c determines whether or not a document image that satisfies the rule application conditions associated in the rule management information 24d with the document naming rule identified in S245 exists at the output destination indicated in the output destination information received along with the target document image in S241 (S247).

[0143] If the rule generation unit 25c determines in S247 that a document image that satisfies the rule application conditions associated in the rule management information 24d with the document naming rule identified in S245 exists at the output destination indicated in the output destination information received in S241 along with the target document image, it inputs the combination of the file name determined in S246 and the metadata of the target document image into the document naming rule generation model 25h, thereby obtaining candidate document naming rules and the confidence level for each candidate (S248).

[0144] When the processing in S248 is completed, the rule generation unit 25c determines whether the candidate with the same document naming rule as identified in S245 has the highest confidence level among the results obtained in S248 (S249).

[0145] If the rule generation unit 25c determines in S249 that the result obtained in S248 has the highest confidence level for a candidate identical to the document naming rule identified in S245, it determines in S250 whether the confidence level of the candidate identical to the document naming rule identified in S245 is higher than a certain threshold in the result obtained in S248.

[0146] If the rule generation unit 25c determines in S250 that the confidence level of a candidate identical to the document naming rule identified in S245 is higher than a certain threshold based on the results obtained in S248, it determines in S251 whether the difference between the confidence level of the candidate identical to the document naming rule identified in S245 and the confidence levels of other document naming rule candidates is greater than a certain threshold based on the results obtained in S248.

[0147] If, based on the results obtained in S248, the rule generation unit 25c determines in S249 that the confidence level of the candidate identical to the document naming rule identified in S245 is not the highest, or if, based on the results obtained in S248, the confidence level of the candidate identical to the document naming rule identified in S245 is not higher than a certain threshold, or if, based on the results obtained in S248, the difference between the confidence level of the candidate identical to the document naming rule identified in S245 and the confidence levels of other document naming rule candidates is not greater than a certain threshold, then, based on the results obtained in S248, the rule generation unit 25c determines in S251 that it will update the document naming rule, for example, by saying, "The current document naming rule does not suit the current document image, so a new document naming rule will be generated." (S252). In S252, there are two methods of notification: one is to send information to the source of the input for displaying the target document image, and the other is to notify the user who uploaded the target document image, i.e., the currently logged-in user, via email to the email address associated with the user management information 24b.

[0148] When the processing in S252 is completed, the rule generation unit 25c retrieves from the storage unit 24 the file names and metadata of all document images that exist at the output destination indicated in the output destination information received along with the target document image in S241, and that satisfy the rule application conditions associated in the rule management information 24d with the document naming rule identified in S245 (S253).

[0149] When the processing in S253 is completed, the rule generation unit 25c generates a document naming rule using the document naming rule generation model 25h based on the combination of file name and metadata for each document image acquired in S253 (S254). Here, the document naming rule generation model 25h is capable of generating multiple candidate document naming rules. However, in S254, the rule generation unit 25c identifies the one with the highest confidence level among the candidate document naming rules generated by the document naming rule generation model 25h as the document naming rule.

[0150] For example, if the file name of the document image obtained in S253 is "20240725_ABC_Receipt", and the values ​​of the items "Date Received", "Recipient Name", and "Document Type" in the metadata of this document image are "July 25, 2024", "ABC", and "Receipt", the rule generation unit 25c may generate, for example, the document naming rule 24c shown in Figure 4 in S254.

[0151] When the processing in S254 is completed, the rule generation unit 25c notifies the rule management unit 25d of the document naming rules generated in S254 (S255).

[0152] When the rule management unit 25d receives the notification in S255, it saves the document naming rule notified in S255 to the storage unit 24 in place of the document naming rule identified in S245 (S256). Here, the rule management unit 25d assigns the same rule ID to the document naming rule to be saved in S256 as the rule ID of the document naming rule identified in S245.

[0153] When the processing in S256 is completed, the rule management unit 25d determines the file name of the target document image based on the metadata of the target document image and the document naming rules generated in S254 (S257).

[0154] The rule management unit 25d determines in S247 that no document image that satisfies the rule application conditions associated in the rule management information 24d with the document naming rule identified in S245 exists at the output destination indicated in the output destination information received along with the target document image in S241, or determines in S251 that the difference between the confidence level of the document naming rule identified in S245 and the confidence level of other candidate document naming rules is greater than a certain threshold based on the results obtained in S248, or when the processing in S257 is completed, it assigns the file name determined in the most recent of the processing in S246 and S257 to the target document image (S258).

[0155] When the processing in S258 is completed, the rule management unit 25d saves the target document image to the output destination indicated in the output destination information received along with the target document image in S241 (S259), and then terminates the operation shown in Figure 17.

[0156] As described above, the document processing system 20 generates document naming rules by inputting the file names and metadata of document images stored in the storage unit 24 into the document naming rule generation model 25h (S123, "S144, S150 and S152" or S254), thereby reducing the burden on the user in generating document naming rules.

[0157] The document processing system 20 selects the candidate with the highest confidence level from among the document naming rule candidates output by the document naming rule generation model 25h as the document naming rule (S123 or S254). This reduces the number of user operations required to generate document naming rules, and as a result, reduces the burden on the user in generating document naming rules.

[0158] The document processing system 20 generates document naming rules using the document naming rule generation model 25h through interaction with the user via an interactive UI (S142-S152). This improves the user's convenience in generating document naming rules, and as a result, reduces the user's burden in generating document naming rules.

[0159] The document processing system 20 inputs the file name determined in S246 based on a specific document naming rule for the input document image, along with the metadata of the input document image, into the document naming rule generation model 25h to obtain candidate document naming rules and the confidence level for each candidate (S248). If the confidence level of a candidate identical to the specific document naming rule does not meet certain conditions (NO in S249, NO in S250, or NO in S251), the system generates a document naming rule by inputting the file name and metadata of the document image stored in the storage unit 24 into the document naming rule generation model 25h (S254). The system then updates the specific document naming rule with the generated document naming rule (S255). As a result, the specific document naming rule can be updated automatically, and the burden on the user for generating document naming rules can be reduced.

[0160] The document processing system 20 can improve convenience by notifying the user (S152) that it will update the specific document naming rule if the confidence level of a candidate identical to the specific document naming rule does not meet certain conditions (NO in S249, NO in S250, or NO in S251).

[0161] In this embodiment, the specific conditions are those that satisfy all of the following: a first condition that the candidate identical to the specific document naming rule has the highest confidence level; a second condition that the confidence level of the candidate identical to the specific document naming rule is higher than a certain threshold; and a third condition that the difference between the confidence level of the candidate identical to the specific document naming rule and the confidence levels of the candidates for other document naming rules is greater than a certain threshold. However, the specific conditions do not necessarily have to include at least one of the first, second, and third conditions.

[0162] The document processing system 20 generates document naming rules using the receipt date in the metadata of the document image stored in the storage unit 24, so it can generate document naming rules that determine the file name according to the receipt date in the metadata of the document image.

[0163] The document processing system 20 generates document naming rules using the issuer name in the metadata of the document image stored in the storage unit 24, so it can generate document naming rules that determine the file name according to the issuer name in the metadata of the document image.

[0164] The document processing system 20 generates document naming rules using the document type in the metadata of the document image stored in the storage unit 24, so it can generate document naming rules that determine the file name according to the document type in the metadata of the document image.

[0165] The document processing system 20 displays the document naming rules generated using the document naming rule generation model 25h via a management UI (S164 and S165), thereby improving usability.

[0166] The document processing system 20 edits the document naming rules generated using the document naming rule generation model 25h according to instructions entered into the management UI (S168), thereby improving usability.

[0167] The document processing system 20 deletes the document naming rules generated using the document naming rule generation model 25h according to instructions entered into the management UI (S171), thereby improving usability.

[0168] In this embodiment, the rule generation unit 25c generates document naming rules using the document naming rule generation model 25h that it itself possesses. However, the rule generation unit 25c may also generate document naming rules using an external document naming rule generation model, such as an external document naming rule generation model of the document processing system 20.

[0169] In this embodiment, the document processing system 20 stores the document naming rules and rule management information 24d in the storage unit 24. However, the document processing system 20 may store at least one of the document naming rules and rule management information outside of the document processing system 20.

[0170] In this embodiment, the document processing system 20 stores document images in the storage unit 24. However, the document processing system 20 may also store document images outside of the document processing system 20.

Claims

1. A document processing system comprising: a rule generation unit that generates document naming rules using a document naming rule generation model as a machine learning model for generating document naming rules as rules for assigning file names to document images as images of documents; the document naming rules determine the file name of the document image according to the metadata of the document image; the document naming rule generation model outputs candidate document naming rules when the file name and metadata of the document image are input; and the rule generation unit generates the document naming rules by inputting the file name and metadata of the document image stored in a document storage unit that stores the document image to the document naming rule generation model.

2. The document processing system according to claim 1, wherein the document naming rule generation model outputs a confidence level for each candidate when outputting candidates, and the rule generation unit determines the candidate with the highest confidence level among the candidates output by the document naming rule generation model as the document naming rule.

3. The document processing system according to claim 1, wherein the rule generation unit generates the document naming rules using the document naming rule generation model through the dialogue via a dialogue UI which is a chatbot that performs a dialogue with the user regarding the generation of the document naming rules.

4. The document processing system according to claim 3, wherein, when a user requests a file name other than the candidate file name via the dialogue UI, the rule generation unit obtains a new candidate and confidence level that reflects the additional conditions included in the request using the document naming rule generation model.

5. When a document image is input, the system determines the file name of the input document image based on the metadata of the input document image and a specific document naming rule; when the document naming rule generation model outputs a candidate, it outputs a confidence level for each candidate; the rule generation unit obtains the candidate and the confidence level for each candidate by inputting the file name determined for the input document image and the metadata of the input document image into the document naming rule generation model; when the confidence level of the candidate that is the same as the specific document naming rule does not satisfy a specific condition, it generates the document naming rule by inputting the file name and metadata of the document image stored in the document storage unit into the document naming rule generation model; and updates the specific document naming rule with the generated document naming rule, the system according to claim 1.

6. The document processing system according to claim 5, wherein the rule generation unit notifies the user that it will update the specific document naming rule if the confidence level of the candidate identical to the specific document naming rule does not satisfy the specific conditions.

7. The document processing system according to claim 1, wherein the metadata includes a receipt date as the date on which the original document of the document image subject to the metadata was received, and the document naming rule generation model uses the receipt date in the output of the candidates.

8. The document processing system according to claim 1, wherein the metadata includes the name of the publisher of the original document to which the document image of the metadata is subject, and the document naming rule generation model uses the publisher name in the output of the candidates.

9. The document processing system according to claim 1, wherein the metadata includes a document type as the type of the original document of the document image to which the metadata is subject, and the document naming rule generation model uses the document type for the output of the candidates.

10. The document processing system according to claim 1, comprising a rule management unit for managing the document naming rules, wherein the rule management unit displays the document naming rules generated by the rule generation unit via a management UI as a UI for managing the document naming rules.

11. The document processing system according to claim 10, wherein the rule management unit edits the document naming rules generated by the rule generation unit in accordance with instructions entered into the management UI.

12. The document processing system according to claim 10, wherein the rule management unit deletes the document naming rules generated by the rule generation unit in accordance with instructions entered into the management UI.

13. A document processing device comprising: a rule generation unit that generates a document naming rule using a document naming rule generation model as a machine learning model for generating a document naming rule as a rule for assigning file names to document images as images of documents; the document naming rule determines the file name of the document image according to the metadata of the document image; the document naming rule generation model outputs a candidate for the document naming rule when the file name of the document image and the metadata are input to the document naming rule generation unit; and the rule generation unit generates the document naming rule by inputting the file name of the document image and the metadata stored in a document storage unit that stores the document image to the document naming rule generation model.

14. A document processing program that operates a computer as a rule generation unit to generate document naming rules, which are rules for assigning filenames to document images as images of documents; the document naming rules determine the filename of the document image according to the metadata of the document image; the document naming rule generation model outputs candidates for the document naming rules when the filename of the document image and the metadata of the document image are input; and the computer is further operated so that the rule generation unit generates the document naming rules by inputting the filename of the document image and the metadata of the document image stored in a document storage unit that stores the document image into the document naming rule generation model.

Images