JP2022139564A - Information processing device, information processing device control method, and program - Google Patents

Abstract

To simplify and facilitate a document processing flow and reduce user's work effort.SOLUTION: A business form template is learned in advance in order to determine the type of a business form. After that, it is determined whether a scanned document matches the learned business form template. When the matching with the learned business form template is confirmed, a document's recognition confidence factor is calculated based on a confidence factor of the business form and a confidence factor of character recognition results. The recognition confidence factor is classified into "very high," "high," and "medium-low." When the document's recognition confidence factor is "very high," the document is registered without user's confirmation of matching between an OCR area image and a recognition result character string.SELECTED DRAWING: Figure 3

Description

The present invention relates to an information processing device, a control method for an information processing device, and a program.

2. Description of the Related Art Conventionally, a technique for extracting character strings from scanned or photographed images using optical character recognition (OCR) is known. As an example of using such a technology, there is an application that distinguishes scanned documents by form type such as an order form or an invoice, and inputs the necessary business data for each form type from the recognition result character string of OCR processing. be. At that time, there is a use case in which the recognition result character string extracted from the scanned document is stored and managed as metadata and used as business data.
In image recognition by an OCR engine, since the image quality of scanned documents and the types of forms vary, it is difficult to achieve 100% accuracy in both form type determination and OCR character recognition. Therefore, instead of calculating a certainty value that indicates the certainty of recognition for each scanned document, this certainty can be effectively used to automate some or all of the scan job processing operations on behalf of the user. If it is possible to do so, it is hoped that work efficiency will be improved.
Japanese Patent Application Laid-Open No. 2002-200003 discloses a technique for improving the accuracy of processing for sorting data into folders by using feature amounts for each document.

JP 2014-149743 A

However, with the technology disclosed in Patent Document 1, it is not possible to reduce the time and effort of matching confirmation by the user, and improvement in work efficiency cannot be sufficiently expected.

SUMMARY OF THE INVENTION An object of the present invention is to simplify and facilitate the flow of document processing and to reduce user's labor.

An information processing apparatus according to the present invention includes an extracting unit for extracting a character string by executing character recognition on a scanned document containing the character string, and a device for determining whether or not the scanned document matches a pre-stored learned form. a calculating means for calculating the certainty factor of the scanned document; and a case where it is judged by the judging means that the scanned document matches the learned form, and the certainty calculated by the calculating means and control means for registering the character string extracted by the extraction means as metadata when the degree is equal to or greater than a first threshold.

According to the present invention, the document processing flow can be simplified and facilitated, and the work effort of the user can be reduced.

It is a figure which shows the system configuration|structure of embodiment, and a network configuration example. It is a block diagram showing an example of hardware composition of an embodiment. It is a block diagram showing an example of software and hardware configuration of an embodiment. FIG. 4 is a diagram for explaining a UI of a client application; FIG. FIG. 10 is a diagram for explaining a UI for associating an OCR area with metadata; FIG. 10 is a diagram for explaining a UI for saving a form template; FIG. FIG. 5 is a diagram for explaining a scan job list UI of a client application; FIG. 10 is a diagram for explaining a message notifying the arrival of a new job; FIG. FIG. 10 is a diagram for explaining a method of displaying new job notifications and automatic transmission job notifications; FIG. 10 is a diagram for explaining a notification center UI; 4 is a flow chart showing an example of a scanned document processing procedure; 10 is a flow chart showing an example of a processing procedure for an expired scanned document;

BEST MODE FOR CARRYING OUT THE INVENTION The best mode for carrying out the present invention will be described below with reference to the drawings.
First, the system configuration and network configuration of an information processing system 1 according to the present embodiment will be described using FIG.
The information processing system 1 according to this embodiment has a scanned document processing server 111 , a client terminal 121 , and a business server 131 . The scanned document processing server 111, client terminal 121, and business server 131 are communicably connected via a network 101 such as the Internet or an intranet.
The scanned document processing server 111 is a server for processing scanned documents.
The client terminal 121 is composed of, for example, a personal computer, a laptop computer, a tablet computer, a smart phone, or the like.
The business server 131 is a server for receiving data from the scanned document processing server 111 and performing various processes.

Next, a hardware configuration example of the scanned document processing server 111, the client terminal 121, and the business server 131 will be described with reference to FIG.
Scan document processing server 111 , client terminal 121 , and business server 131 have user I/F 201 , network I/F 202 , CPU 203 , ROM 204 , RAM 205 , secondary storage device 206 , and internal bus 207 . is doing. Each unit is connected via an internal bus 207 .
A network I/F 202 communicates with other computers and network devices via the network 101 . The communication method may be wired or wireless.

The ROM 204 records embedded programs and data. RAM 205 is a temporary memory area. A secondary storage device 206 is a storage device such as an HDD or a flash memory.
The CPU 203 executes programs read from the ROM 204, the RAM 205, the secondary storage device 206, and the like. A user I/F 201 inputs and outputs information, signals, and the like using a display, keyboard, mouse, button, touch panel, and the like.
Computers without such hardware may be connected and operated from other computers by means of remote desktop, remote shell, or the like.

Next, a software configuration example of the information processing system 1 according to this embodiment will be described with reference to FIG. The software installed in each piece of hardware is executed by the CPU 203, and is configured to be able to communicate with each other, as indicated by network connection arrows.

First, the scanned document processing server 111 has a scanned document processing application 311 , a data store 321 and a backend application 331 .
A scanned document processing application 311 is an application installed in the scanned document processing server 111 . Although this embodiment is described as operating as a web application server, it may operate as another application implementation form.
The scanned document processing application 311 has an API (Application Programming Interface) 312 and a Web UI (User Interface) 313 .

The data store 321 is a data store that saves and stores data used by the scanned document processing application 311 or a backend application 331 described later. The data store 321 has a document storage section 322 , a job queue 323 , a management section 324 and a result storage section 325 .
The document storage unit 322 holds the file of the scanned document itself as an image file such as JPEG or a document file such as PDF (Portable Document Format).
The job queue 323 holds a queue for managing jobs waiting for metadata input processing, which will be described later.

The management unit 324 holds a list of metadata that needs to be added to each scanned document, a name for each metadata, a value format (character string, number, etc.), color information to be described later, and the like.
The result storage unit 325 stores OCR processing results and form determination results. The result storage unit 325 also holds OCR area information associated with metadata for each scanned document, input metadata values, and the like.

The backend application 331 is in charge of the following processes that may be executed sequentially in the background. The backend application 331 has an OCR processing section 332 , a form processing section 333 , a communication section 334 and a calculation section 335 .
The OCR processing unit 332 acquires the scanned document from the document storage unit 322 and performs OCR. In the OCR processing, the OCR processing unit 332 extracts the starting point coordinates, width, and height of the region recognized as the character string, and the recognition result character string of the OCR processing that has been recognized.
The form processing unit 333 determines the type of the form using information such as the OCR-processed scanned document, the OCR-processed result area pattern, and the OCR-processed recognition result character string. Any technique such as pattern recognition or machine learning may be used for the determination process.

The calculating unit 335 calculates the recognition certainty for the scanned document based on the certainty of the form discrimination processing by the form processing unit 333 and the certainty of the recognition result character string of the OCR processing by the OCR processing unit 332 . Here, the reliability of the form discrimination process is, for example, a similarity score (0% to 100%) between the learned form template used for the form type discrimination process and the scanned document. The degree of certainty of the recognition result character string of the OCR processing is, for example, the degree of certainty (0% to 100%) that the character string recognized from the OCR area matches the correct answer character string.
Recognition certainty may be a score in percentage units with an upper limit of 100%, a level divided from the score such as high, medium, or low, or may be assigned in any manner. Further, the method of calculating the recognition certainty factor may be to take the average of the certainty factors of the recognition result character strings of the form discrimination processing and all the OCR areas, or another method may be used. For example, a high score can be obtained by calculating the average confidence factor, but if some OCR regions have a significantly low confidence factor, a calculation method that lowers the score as a whole may be adopted. In any case, since the purpose of the present embodiment is to process using the recognition certainty, any calculation method may be used.
The communication unit 334 transmits the scanned document and its processing result to the external business server 131 . Note that the communication unit 334 may be omitted if there is no need to transmit the scanned document and its processing result to the outside.

Next, the client terminal 121 has a client application 351 . In this embodiment, the web application of the scanned document processing application 311 is executed. As one form of providing the client application 351, there is a method of displaying the web UI 313 on the browser, sending and receiving necessary data to and from the API 312, and executing the web application. Alternatively, it may be a computer or smart phone application created to transmit and receive necessary data to and from the API 312 .

Next, the business server 131 has a business application 361 and a storage 362 .
The business application 361 is an application executed by the business server 131 . The storage 362 stores data used by the business application 361 . The business application 361 may be any business application such as file management, document management, order receiving, and accounting. The business application 361 is required when receiving, business processing, and storing the results processed by the scanned document processing server 111 . Otherwise, it may be omitted.

Next, the process of learning a form template used in the form type determination process will be described with reference to FIGS. 4 to 6. FIG. Specifically, in the form type determination process to be described later, the form processing unit 333 determines whether or not the scanned document matches the learned form template. The following processing is for registering a new learned form template when the scanned document does not match the learned form template, that is, when the document is an unlearned form.
First, the UI of the client application 351 will be described with reference to FIG.
A preview pane 401 has a preview page image 402 . In the preview pane 401, the preview page image 402 can be scrolled or zoomed to display any position. A metadata pane 411 is a pane for displaying and editing a list of metadata to be assigned to each form type for which a scanned document has been determined. The example of FIG. 4 shows a state in which the scanned document is determined to be an unlearned form and there are three pieces of metadata that need to be input.

An area 421 is an area for explaining an orthogonal coordinate system indicating the relationship between the preview page image 402 and the OCR area. In the example shown in FIG. 4, there are three OCR areas 422, 423, and 424, which are recognition results of OCR processing. The preview page image 402 has three or more OCR areas, which are omitted here. The OCR regions are each identified by their starting point coordinates, width and height, like their respective shaded rectangles. For example, the area 424 is expressed as starting point coordinates (1200, 700), width 720, height 120, etc. with the unit being pixels.

FIG. 5 is a diagram for explaining a UI for associating an OCR area of a preview page image with metadata. Area 502 is the OCR area associated with the metadata. In the area 502, a colored rectangle is transparently displayed so that the user can easily identify the area as the OCR area. It should be noted that the coloring means may be any other means such as colored border display.
An area 512 is an area for displaying the color assigned to the metadata name 511 “Customer Name”. An area 513 is an area for displaying a clipped image of the area 502 . If the association with the OCR area is not specified, a "+" button is displayed in area 513 as shown in FIG. A control 514 is a control for displaying and editing the extracted character string in the area 502 .
Once all the required metadata has been entered, the register button 515 is activated. When the user selects the registration button 515, the screen transitions to the final confirmation dialog shown in FIG.

FIG. 6 is a diagram for explaining a UI for naming and saving a form template used for form type determination. Dialog 601 is the final confirmation dialog. A control 602 is a text control for entering a form template name. A control 603 is a selection control for designating the destination of the scanned document. A save/send button 604 is a button for saving the form template name and sending the scan job to the destination.

After the form template is learned as described above, next time a scanned document of the same form type arrives, the form processing unit 333 determines that the scanned document matches the learned form template in the form type determination process. be done. After that, a character string is input to each metadata from the OCR area already defined in the learned form template, and the same transmission destination is selected, so that the scan job can be transmitted simply and easily.

Next, a processing flow of scanned document processing using document recognition certainty in the scanned document processing server 111 will be described with reference to the flowchart of FIG.
This processing starts when the user uploads a scanned document file from the client terminal 121 or the like to the document storage unit 322 . As for the upload method, the upload may be performed from the client application 351 via the API 312, or may be directly uploaded from a device such as a scanner.
Here, there are simple cases where a file contains only one page image, and cases where it contains multiple page images. In the latter case, each page may be divided into a plurality of scanned documents. It is also possible to detect a bar code, partition paper, or the like, and divide the document into a plurality of scanned documents each having an arbitrary number of pages. In both the single and plural cases, the processing flow will be described below on the premise that N scanned documents are processed.

First, as described above, in step S1101, the backend application 331 divides one file uploaded to the document storage unit 322 into N scanned documents. Thereafter, in steps S1102 to S1114, loop processing from n=1 to N is executed.
Next, in step S1102, the backend application 331 stores the scan job of document [n] in the job queue 323, and the OCR processing unit 332, form processing unit 333, and calculation unit 335 process it.
Specifically, the OCR processing unit 332 first performs OCR processing on the document [n], and stores the recognition result character string of the OCR processing in the result storage unit 325 . Next, the form processing unit 333 determines the form type of document [n] based on the result of the OCR processing, and searches for a learned form template that matches the form type of document [n]. If a learned form template of the same form exists, the form processing unit 333 stores the learned form template in the result storage unit 325 . On the other hand, if a learned form template that is the same form does not exist, the result storage unit 325 stores a form recognition result that does not match.
Next, when it matches the learned form template, the calculation unit 335 calculates the recognition certainty of document [n], and the calculation unit 335 stores the calculated recognition certainty in the result storage unit 325 as well.
In step S1103, the back-end application 331 determines whether or not the form recognition result in step S1102 matches the learned form template. If the back-end application 331 determines that it matches the learned form template, the process advances to step S1104. On the other hand, if the back-end application 331 determines that it does not match the learned form template, the process advances to step S1111.

Next, in step S<b>1104 , the backend application 331 acquires the recognition confidence of document [n] from the result storage unit 325 .
Next, in step S1105, the backend application 331 determines the recognition confidence of document[n]. In this embodiment, for example, the recognition certainty is classified into three levels of "very high", "high", and "moderately low". That is, if the recognition certainty is greater than or equal to the first threshold, it is "very high"; if the recognition certainty is smaller than the first threshold and greater than or equal to the second threshold, it is "high"; If it is smaller, it is classified as "medium-low".
When the back-end application 331 determines that the recognition certainty is "pretty high", it does not require the user to review the form recognition result and the recognition result character string of the OCR process, and the process proceeds to step S1106.
Further, when the back-end application 331 determines that the recognition confidence is "high" or "medium-low", it requires the user to review the form recognition result and the recognition result character string of the OCR processing, and the process proceeds to step S1110. proceed. When the recognition confidence is determined to be "high", processing is performed so that the user's operation is as simple and easy as possible compared to when the recognition confidence is determined to be "middle-low".

Next, in step S1106, the backend application 331 extracts the recognition result character string of the OCR process as metadata using the defined OCR area coordinates and size in the learned form template that matches the document [n]. ,sign up.
In step S1107, the back-end application 331 transmits the document [n] and metadata via the communication unit 334 to the destination defined in the learned form template.

Next, in step S 1108 , the backend application 331 dequeues document[n] from the job queue 323 . Then, in step S1109, the backend application 331 generates mail notification content for notifying that the transmission of document [n] has been completed.

On the other hand, in step S1110, the backend application 331 extracts the recognition result character string of the OCR processing as metadata, as in step S1106. However, since the recognition result character string of the OCR process is not necessarily the correct recognition result, it is extracted as temporary metadata. Next, in step S1111, the backend application 331 updates the status of the job in the job queue 323 to recognition completed. Here, if the recognition confidence is determined to be "high", the process proceeds to step S1112. If so, the process proceeds to step S1113.

Next, in step S1112, the backend application 331 generates mail notification content notifying that the job is available for review. Here, the mail notification content generated in step S1112 will be described with reference to FIG.

FIG. 8 is an example of message content 801 of an e-mail notifying that a new scan job has arrived. When one notification message is created for one scan job and an email notification is sent in step S1114, which will be described later, the user can confirm the details of the contents shown in FIG. Thumbnail 810 is a thumbnail of the scanned image. A template name 811 is a learned form template name that matches the scanned document.
Confidence 812 is the recognition confidence of the scanned document. A destination 813 is a transmission destination of the scanned document. File name 814 is the file name of the scanned document. The item name 815 is the item name of metadata. Image 816 is an image of the OCR area. A character string 817 is a recognition result character string of OCR processing. That is, by displaying the content of the thumbnail 810 to the character string 817 in the message content 801 on the client terminal 121, the user can check the OCR recognition result on the spot.

A registration button 821 is a URL link that performs the same processing as the transmission processing of the save/send button 604 described above. The user selects the registration button 821 when judging that the result is correct as a result of matching confirmation. When the registration button 821 is selected, the browser is activated and a request is sent to the scanned document processing application 311 indicated by the URL. The scanned document processing application 311 processes scan jobs in the job queue 323 and registers scanned documents and metadata in the business application 361 via the communication unit 334 . The metadata registered here is the confirmed character string 817 .
A calibration button 822 is a link button to a calibration screen. If the user finds a typo in matching between the image 816 and the character string 817, and selects the calibration button 822, the client application 351 is activated and the scan job list UI is displayed on the browser. The user corrects the character string of the metadata as necessary on the scan job list UI.

Here, the scan job list UI of the client application 351 will be described using FIG. The scan job list UI 701 can display scan jobs assigned to the user in one or more columns in a list format.
A column 711 is a column for displaying thumbnails of scan jobs. A column 712 is a column for displaying the file name of the scan job. A column 713 is a column for displaying the creation date and time of the scan job. A column 714 is a column for displaying a learned form template name that matches as a result of the form type determination of the scanned document. A column 715 is a column for displaying the transmission destination of the scan job. An update button 716 is a button for updating the scan job list.

A metadata pane 721 displays a list of metadata required for each form type. A field 722 is a text field that displays the item name of metadata. Image 723 is an image of the OCR area associated with the metadata.
Control 724 is a text edit control for entering and modifying metadata values. Area 725 is an area that displays colored rectangles so that the user can easily identify each piece of metadata, similar to area 512 in FIG.

A registration button 726 is a button for registering a scanned document as a transmission destination.
When the calibration button 822 is selected from the screen of FIG. 8, the line of the scan job to be processed is first selected from the scan job list on the scan job list UI 701 . In the metadata pane 721 on the right side, based on the learned form template that matches the scanned document, an image 723 of the OCR area defined for each metadata and a control 724 in which the recognition result character string of the OCR processing is described. Is displayed. The user can modify the metadata values in control 724 . When the correction is completed and the user selects the registration button 726 , the scanned document is registered at the destination displayed in column 715 .

Meanwhile, in step S1113, back-end application 331 generates mail notification content notifying that recognition is complete and the user can start processing document [n]. This content includes only the proofreading button 822 in FIG. 8, and the user can confirm the details of the content when an email notification is sent in step S1114, which will be described later. Here, if the recognition confidence level is determined to be "medium-low" in step S1105, the calibration button 822 is selected to transition to the scan job list UI of the client application 351 shown in FIG. On the other hand, if it is determined in step S1103 that the document template does not match the learned form template, the user selects the calibration button 822 in FIG. 8 to transition to the screen in FIG.
Then, in step S1114, the backend application 331 transmits the email notification generated in steps S1112 and S1113.

After the processing of steps S1102 to S1114 has been performed on the N scanned documents, a notification of the N scanned documents is sent to the client terminal 121 in the subsequent steps.
First, in step S1115, the back-end application 331 notifies the scan job list UI of the scanned document for which the recognition confidence level is "high" or "medium-low" in step S1105 or "No" in step S1103 as a new job. .
Also, in step S1116, the back-end application 331 notifies the scan job list UI 901 of the scanned document for which the degree of recognition certainty was determined to be "very high" in step S1105 as an automatic transmission job.

Here, a method of displaying a new job notification and an automatic transmission job notification on the scan job list UI will be described with reference to FIG. Note that the screen of the scan job list UI 901 shown in FIG. 9 is displayed when the client application 351 is activated in the client terminal 121 by the user's operation.
First, the number of new jobs notified in step S1115 is displayed in the update button 911. FIG. When the user selects the update button 911 , the client application 351 transmits a request to acquire the latest scan job list to the scanned document processing application 311 . Upon receiving the response, the client application 351 updates the scan job list UI 901 to the latest state. In the example of FIG. 9, two new jobs have arrived in the job queue 323, indicating that there are additional messages (two new jobs) in the update button 911. FIG. The client application 351 checks the number of arrivals of new jobs in the background with the scanned document processing application 311 . If there is a change in the number of arrivals of new jobs, the message of the update button 911 is updated.

A message 912 is a message for displaying the number of automatic transmission jobs notified in step S1116. The client application 351 confirms the number of automatic transmission jobs in the background with the scanned document processing application 311 . If there is a change in the number of automatic transmission jobs, the contents of the message 912 are updated.
A link 913 is a link to a history of automatic transmission jobs. If the recognition confidence of the scanned document is sufficiently high, it will be sent to the destination without user matching confirmation. In this case, since the scan job submitted by the user is not displayed in the list of the scan job list UI 901, the user can confirm it with a message 912 instead. If necessary, the processing completion status of the scan job can be checked from the history link 913 .
A button 914 is a button for deleting a message. After confirming the number and completion status of automatic transmission jobs from message 912 or link 913, the user selects button 914 to delete the automatic transmission job notification message.

Next, in step S1117, the back-end application 331 uses the OS (operating system) of the client terminal 121 for the scanned document whose recognition certainty is "high" or "medium-low" in step S1105 or "No" in step S1103. The notification center provided with such as is notified as a new job.
Next, in step S1118, the back-end application 331 notifies the notification center of the scanned document for which the degree of recognition certainty was determined to be "very high" in step S1105 as an automatic transmission job.

Here, a method for displaying a new job notification and an automatic transmission job notification on the notification center provided by the OS of the client terminal 121 will be described with reference to FIG.
In FIG. 10, the notification center UI 1000 has a new job notification 1001 and an automatic transmission job notification 1002 . In steps S1117 and S1118, new job notification 1001 and automatic transmission job notification 1002 are displayed when backend application 331 notifies notification center of new jobs and automatic transmission jobs.
The notification message described above with reference to FIG. 9 can only be confirmed within the client application 351, but by displaying the notification message in the notification center provided in the OS, etc., a so-called toast notification can be displayed to the user to alert the user. . Specifically, for example, after the user submits 5 scan jobs, the user can clearly see that 3 new jobs have arrived and 2 jobs have been automatically sent, and the visibility of the automatically sent jobs improve sexuality.
The new job notification 1001 and the automatic transmission job notification 1002 can be closed by the user selecting the close button after confirmation. Also, when the user selects the automatic transmission job notification 1002, it operates as a link to the history of automatic transmission jobs, similar to the link 913 in FIG.

Next, the processing when the retention period of the scan job has passed will be described with reference to the flowchart of FIG. 12 . Note that this process is performed by the backend application 331 at a predetermined cycle.
First, in step S<b>1201 , the backend application 331 acquires the creation date of the scan job of unprocessed document [n] stored in the job queue 323 .
Next, in step S1202, the back-end application 331 compares the creation date acquired in step S1201 with the current date and time, and determines whether the job retention period (eg, 7 days) has passed. If the back-end application 331 determines that the job retention period has elapsed, the process advances to step S1203. On the other hand, if the back-end application 331 determines that the job holding period has not elapsed, the processing ends.

Next, in step S<b>1203 , the backend application 331 acquires the recognition confidence of document [n] from the result storage unit 325 .
Next, in step S1204, the backend application 331 determines the recognition confidence. If the back-end application 331 determines that the degree of recognition certainty is "high", the process proceeds to step S1205. On the other hand, when the back-end application 331 determines that the recognition certainty is "medium-low", the process proceeds to step S1206.
In step S<b>1205 , the backend application 331 transmits document [n] and metadata to the destination defined in the learned form template via the communication unit 334 . In this process, since the temporary metadata has already been extracted in step S1110, the temporary metadata is registered as regular metadata, and the document [n] and the metadata are sent to the destination.
In step S<b>1206 , the backend application 331 deletes document [n] from the document storage unit 322 .
In step S1207, the backend application 331 dequeues the document [n] from the job queue 323, and ends this process.

In this embodiment, as an example of message notification means, message notification by e-mail has been described, but notification may be sent to other applications such as business chat and in-house SNS. With markup languages such as HTML and Markdown, which are supported by business chats and in-house SNS, it is possible to provide an OCR result review and match confirmation UI. Therefore, any application, protocol, markup language, or the like may be used as a message notification means.

As described above, according to the information processing system of the present embodiment, by branching the document processing flow using the recognition confidence in OCR and form recognition, user operations can be simplified and facilitated. Work efficiency can be improved.

(Other embodiments)
The present invention supplies a program that implements one or more functions of the above-described embodiments to a system or device via a network or a storage medium, and one or more processors in a computer of the system or device reads and executes the program. It can also be realized by processing to It can also be implemented by a circuit (for example, ASIC) that implements one or more functions.

331 backend application, 332 OCR processing unit, 333 form processing unit, 335 calculation unit

Claims (18)

an extracting means for extracting a character string by performing character recognition on a scanned document containing the character string;
determining means for determining whether or not the scanned document matches a pre-stored learned form;
a calculating means for calculating a certainty factor of the scanned document;
a character string extracted by the extraction means when the determination means determines that the scanned document matches the learned form and the certainty calculated by the calculation means is equal to or greater than a first threshold; a control means for registering as metadata;
An information processing device comprising: When the determination means determines that the scanned document matches the learned form, the control means determines that the degree of certainty calculated by the calculation means is smaller than the first threshold and a second threshold. If the above, notify the client terminal of information about the result of character recognition;
The information processing apparatus according to claim 1, characterized by: When the determination means determines that the scanned document matches the learned form, and the certainty factor calculated by the calculation means is smaller than the first threshold value and equal to or greater than the second threshold value. the information on the result of character recognition notified by the control means is notified by e-mail;
3. The information processing apparatus according to claim 2, characterized by: When the determination means determines that the scanned document matches the learned form, and the certainty factor calculated by the calculation means is smaller than the first threshold value and equal to or greater than the second threshold value. the information on the result of character recognition notified by the control means is notified by a business chat or an in-house SNS;
3. The information processing apparatus according to claim 2, characterized by: The notified information about the result of character recognition includes information on the character string extracted by the extracting means, a link for registering the character string extracted by the extracting means as metadata, and the extracting means. contains a link for proofreading the string extracted by
The information processing apparatus according to any one of claims 2 to 4, characterized by: notifying the client terminal of a link for proofreading the character string extracted by the extraction means when the confidence calculated by the calculation means is smaller than the second threshold;
6. The information processing apparatus according to any one of claims 2 to 5, characterized by: When the certainty calculated by the calculating means is smaller than the first threshold and equal to or larger than the second threshold, and when a predetermined period of time has passed while the scanned document is held, the control is performed. means for registering the character string extracted by the extraction means as metadata;
7. The information processing apparatus according to any one of claims 2 to 6, characterized by: The calculating means calculates the certainty factor of the scanned document based on the certainty factor of the result of the character recognition by the extracting means;
The information processing apparatus according to any one of claims 1 to 7, characterized by: The calculating means calculates the certainty of the scanned document based on the certainty of the result of the character recognition by the extracting means and the similarity between the learned form determined by the determining means;
The information processing apparatus according to any one of claims 1 to 7, characterized by: notifying a client terminal of information for learning the scanned document as a form when the determination means determines that the scanned document does not match a pre-stored learned form;
The information processing apparatus according to any one of claims 1 to 9, characterized by: the control means notifying the client terminal of information about the scanned document whose certainty is equal to or greater than the first threshold;
The information processing apparatus according to any one of claims 1 to 10, characterized by: the control means notifying the client terminal of information for displaying a link to information on the scanned document whose certainty is equal to or greater than the first threshold;
The information processing apparatus according to any one of claims 1 to 10, characterized by: the control means notifying a notification center of the client terminal of information for displaying, on the client terminal, information relating to the scanned document whose certainty is equal to or greater than the first threshold;
The information processing apparatus according to any one of claims 1 to 10, characterized by: the information about the scanned document includes a link to the information about the scanned document;
14. The information processing apparatus according to claim 13, characterized by: When it is determined that the scanned document matches the learned form and the certainty calculated by the calculation means is equal to or greater than the first threshold, the control means controls the Registering a character string as metadata and transmitting the scanned document and the metadata to a destination registered in the learned form;
The information processing apparatus according to any one of claims 1 to 14, characterized by: an extracting means for extracting a character string by performing character recognition on a scanned document containing the character string;
Information on the character string extracted by the extraction means, a link for registering the character string extracted by the extraction means as metadata, and a link for proofreading the character string extracted by the extraction means. Notification means for performing notification including by at least one of e-mail, business chat, and in-house SNS;
2. The information processing apparatus according to claim 1, further comprising: an extraction step of extracting the character string by performing character recognition on the scanned document containing the character string;
a determination step of determining whether or not the scanned document matches a pre-stored learned form;
a calculation step of calculating a certainty factor of the scanned document;
When the determination step determines that the scanned document matches the learned form and the certainty calculated in the calculation step is equal to or greater than a first threshold, the character string extracted in the extraction step a control step of registering as metadata;
A control method for an information processing device, comprising: A program for causing a computer to function as each means of the information processing apparatus according to any one of claims 1 to 16.

Images